6502/llvm-6502
1
0
Fork 0
mirror of https://github.com/c64scene-ar/llvm-6502.git synced 2025-06-26 07:24:25 +00:00
Code Issues Projects Releases Wiki Activity

Replace the original flex lexer with a hand writen one. This

Browse Source 
drops a dependency on flex and lets us make future progress more 
easily.  Yay for 2 fewer .cvs files to make silly conflicts with.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@44213 91177308-0d34-0410-b5e6-96231b3b80d8
This commit is contained in:
Chris Lattner
2007-11-18 08:46:26 +00:00
parent c1819188b6
commit 8e3a8e0452
9 changed files with 929 additions and 4424 deletions
Download Patch File Download Diff File Expand all files Collapse all files

826
lib/AsmParser/LLLexer.cpp Normal file
View File

@ -0,0 +1,826 @@
//===- LLLexer.cpp - Lexer for .ll Files ----------------------------------===//
//
// The LLVM Compiler Infrastructure
//
// This file was developed by Chris Lattner and is distributed under
// the University of Illinois Open Source License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
// Implement the Lexer for .ll files.
//
//===----------------------------------------------------------------------===//
#include "LLLexer.h"
#include "ParserInternals.h"
#include "llvm/Support/MemoryBuffer.h"
#include <list>
#include "llvmAsmParser.h"
using namespace llvm;
//===----------------------------------------------------------------------===//
// Helper functions.
//===----------------------------------------------------------------------===//
// atoull - Convert an ascii string of decimal digits into the unsigned long
// long representation... this does not have to do input error checking,
// because we know that the input will be matched by a suitable regex...
//
static uint64_t atoull(const char *Buffer, const char *End) {
uint64_t Result = 0;
for (; Buffer != End; Buffer++) {
uint64_t OldRes = Result;
Result *= 10;
Result += *Buffer-'0';
if (Result < OldRes) { // Uh, oh, overflow detected!!!
GenerateError("constant bigger than 64 bits detected!");
return 0;
}
}
return Result;
}
static uint64_t HexIntToVal(const char *Buffer, const char *End) {
uint64_t Result = 0;
for (; Buffer != End; ++Buffer) {
uint64_t OldRes = Result;
Result *= 16;
char C = *Buffer;
if (C >= '0' && C <= '9')
Result += C-'0';
else if (C >= 'A' && C <= 'F')
Result += C-'A'+10;
else if (C >= 'a' && C <= 'f')
Result += C-'a'+10;
if (Result < OldRes) { // Uh, oh, overflow detected!!!
GenerateError("constant bigger than 64 bits detected!");
return 0;
}
}
return Result;
}
// HexToFP - Convert the ascii string in hexadecimal format to the floating
// point representation of it.
//
static double HexToFP(const char *Buffer, const char *End) {
return BitsToDouble(HexIntToVal(Buffer, End)); // Cast Hex constant to double
}
static void HexToIntPair(const char *Buffer, const char *End, uint64_t Pair[2]){
Pair[0] = 0;
for (int i=0; i<16; i++, Buffer++) {
assert(Buffer != End);
Pair[0] *= 16;
char C = *Buffer;
if (C >= '0' && C <= '9')
Pair[0] += C-'0';
else if (C >= 'A' && C <= 'F')
Pair[0] += C-'A'+10;
else if (C >= 'a' && C <= 'f')
Pair[0] += C-'a'+10;
}
Pair[1] = 0;
for (int i=0; i<16 && Buffer != End; i++, Buffer++) {
Pair[1] *= 16;
char C = *Buffer;
if (C >= '0' && C <= '9')
Pair[1] += C-'0';
else if (C >= 'A' && C <= 'F')
Pair[1] += C-'A'+10;
else if (C >= 'a' && C <= 'f')
Pair[1] += C-'a'+10;
}
if (*Buffer)
GenerateError("constant bigger than 128 bits detected!");
}
// UnEscapeLexed - Run through the specified buffer and change \xx codes to the
// appropriate character.
static void UnEscapeLexed(std::string &Str) {
if (Str.empty()) return;
char *Buffer = &Str[0], *EndBuffer = Buffer+Str.size();
char *BOut = Buffer;
for (char *BIn = Buffer; BIn != EndBuffer; ) {
if (BIn[0] == '\\') {
if (BIn < EndBuffer-1 && BIn[1] == '\\') {
*BOut++ = '\\'; // Two \ becomes one
BIn += 2;
} else if (BIn < EndBuffer-2 && isxdigit(BIn[1]) && isxdigit(BIn[2])) {
char Tmp = BIn[3]; BIn[3] = 0; // Terminate string
*BOut = (char)strtol(BIn+1, 0, 16); // Convert to number
BIn[3] = Tmp; // Restore character
BIn += 3; // Skip over handled chars
++BOut;
} else {
*BOut++ = *BIn++;
}
} else {
*BOut++ = *BIn++;
}
}
Str.resize(BOut-Buffer);
}
/// isLabelChar - Return true for [-a-zA-Z$._0-9].
static bool isLabelChar(char C) {
return isalnum(C) || C == '-' || C == '$' || C == '.' || C == '_';
}
/// isLabelTail - Return true if this pointer points to a valid end of a label.
static const char *isLabelTail(const char *CurPtr) {
while (1) {
if (CurPtr[0] == ':') return CurPtr+1;
if (!isLabelChar(CurPtr[0])) return 0;
++CurPtr;
}
}
//===----------------------------------------------------------------------===//
// Lexer definition.
//===----------------------------------------------------------------------===//
// FIXME: REMOVE THIS.
#define YYEOF 0
#define YYERROR -2
LLLexer::LLLexer(MemoryBuffer *StartBuf) : CurLineNo(1), CurBuf(StartBuf) {
CurPtr = CurBuf->getBufferStart();
}
std::string LLLexer::getFilename() const {
return CurBuf->getBufferIdentifier();
}
int LLLexer::getNextChar() {
char CurChar = *CurPtr++;
switch (CurChar) {
default: return (unsigned char)CurChar;
case 0:
// A nul character in the stream is either the end of the current buffer or
// a random nul in the file. Disambiguate that here.
if (CurPtr-1 != CurBuf->getBufferEnd())
return 0; // Just whitespace.
// Otherwise, return end of file.
--CurPtr; // Another call to lex will return EOF again.
return EOF;
case '\n':
case '\r':
// Handle the newline character by ignoring it and incrementing the line
// count. However, be careful about 'dos style' files with \n\r in them.
// Only treat a \n\r or \r\n as a single line.
if ((*CurPtr == '\n' || (*CurPtr == '\r')) &&
*CurPtr != CurChar)
++CurPtr; // Eat the two char newline sequence.
++CurLineNo;
return '\n';
}
}
int LLLexer::LexToken() {
TokStart = CurPtr;
int CurChar = getNextChar();
switch (CurChar) {
default:
// Handle letters: [a-zA-Z_]
if (isalpha(CurChar) || CurChar == '_')
return LexIdentifier();
return CurChar;
case EOF: return YYEOF;
case 0:
case ' ':
case '\t':
case '\n':
case '\r':
// Ignore whitespace.
return LexToken();
case '+': return LexPositive();
case '@': return LexAt();
case '%': return LexPercent();
case '"': return LexQuote();
case '.':
if (const char *Ptr = isLabelTail(CurPtr)) {
CurPtr = Ptr;
llvmAsmlval.StrVal = new std::string(TokStart, CurPtr-1);
return LABELSTR;
}
if (CurPtr[0] == '.' && CurPtr[1] == '.') {
CurPtr += 2;
return DOTDOTDOT;
}
return '.';
case '$':
if (const char *Ptr = isLabelTail(CurPtr)) {
CurPtr = Ptr;
llvmAsmlval.StrVal = new std::string(TokStart, CurPtr-1);
return LABELSTR;
}
return '$';
case ';':
SkipLineComment();
return LexToken();
case '0': case '1': case '2': case '3': case '4':
case '5': case '6': case '7': case '8': case '9':
case '-':
return LexDigitOrNegative();
}
}
void LLLexer::SkipLineComment() {
while (1) {
if (CurPtr[0] == '\n' || CurPtr[0] == '\r' || getNextChar() == EOF)
return;
}
}
/// LexAt - Lex all tokens that start with an @ character:
/// AtStringConstant @\"[^\"]*\"
/// GlobalVarName @[-a-zA-Z$._][-a-zA-Z$._0-9]*
/// GlobalVarID @[0-9]+
int LLLexer::LexAt() {
// Handle AtStringConstant: @\"[^\"]*\"
if (CurPtr[0] == '"') {
++CurPtr;
while (1) {
int CurChar = getNextChar();
if (CurChar == EOF) {
GenerateError("End of file in global variable name");
return YYERROR;
}
if (CurChar == '"') {
llvmAsmlval.StrVal = new std::string(TokStart+2, CurPtr-1);
UnEscapeLexed(*llvmAsmlval.StrVal);
return ATSTRINGCONSTANT;
}
}
}
// Handle GlobalVarName: @[-a-zA-Z$._][-a-zA-Z$._0-9]*
if (isalpha(CurPtr[0]) || CurPtr[0] == '-' || CurPtr[0] == '$' ||
CurPtr[0] == '.' || CurPtr[0] == '_') {
++CurPtr;
while (isalnum(CurPtr[0]) || CurPtr[0] == '-' || CurPtr[0] == '$' ||
CurPtr[0] == '.' || CurPtr[0] == '_')
++CurPtr;
llvmAsmlval.StrVal = new std::string(TokStart+1, CurPtr); // Skip @
return GLOBALVAR;
}
// Handle GlobalVarID: @[0-9]+
if (isdigit(CurPtr[0])) {
for (++CurPtr; isdigit(CurPtr[0]); ++CurPtr);
uint64_t Val = atoull(TokStart+1, CurPtr);
if ((unsigned)Val != Val)
GenerateError("Invalid value number (too large)!");
llvmAsmlval.UIntVal = unsigned(Val);
return GLOBALVAL_ID;
}
return '@';
}
/// LexPercent - Lex all tokens that start with a % character:
/// PctStringConstant %\"[^\"]*\"
/// LocalVarName %[-a-zA-Z$._][-a-zA-Z$._0-9]*
/// LocalVarID %[0-9]+
int LLLexer::LexPercent() {
// Handle PctStringConstant: %\"[^\"]*\"
if (CurPtr[0] == '"') {
++CurPtr;
while (1) {
int CurChar = getNextChar();
if (CurChar == EOF) {
GenerateError("End of file in local variable name");
return YYERROR;
}
if (CurChar == '"') {
llvmAsmlval.StrVal = new std::string(TokStart+2, CurPtr-1);
UnEscapeLexed(*llvmAsmlval.StrVal);
return PCTSTRINGCONSTANT;
}
}
}
// Handle LocalVarName: %[-a-zA-Z$._][-a-zA-Z$._0-9]*
if (isalpha(CurPtr[0]) || CurPtr[0] == '-' || CurPtr[0] == '$' ||
CurPtr[0] == '.' || CurPtr[0] == '_') {
++CurPtr;
while (isalnum(CurPtr[0]) || CurPtr[0] == '-' || CurPtr[0] == '$' ||
CurPtr[0] == '.' || CurPtr[0] == '_')
++CurPtr;
llvmAsmlval.StrVal = new std::string(TokStart+1, CurPtr); // Skip %
return LOCALVAR;
}
// Handle LocalVarID: %[0-9]+
if (isdigit(CurPtr[0])) {
for (++CurPtr; isdigit(CurPtr[0]); ++CurPtr);
uint64_t Val = atoull(TokStart+1, CurPtr);
if ((unsigned)Val != Val)
GenerateError("Invalid value number (too large)!");
llvmAsmlval.UIntVal = unsigned(Val);
return LOCALVAL_ID;
}
return '%';
}
/// LexQuote - Lex all tokens that start with a " character:
/// QuoteLabel "[^"]+":
/// StringConstant "[^"]*"
int LLLexer::LexQuote() {
while (1) {
int CurChar = getNextChar();
if (CurChar == EOF) {
GenerateError("End of file in quoted string");
return YYERROR;
}
if (CurChar != '"') continue;
if (CurPtr[0] != ':') {
llvmAsmlval.StrVal = new std::string(TokStart+1, CurPtr-1);
UnEscapeLexed(*llvmAsmlval.StrVal);
return STRINGCONSTANT;
}
++CurPtr;
llvmAsmlval.StrVal = new std::string(TokStart+1, CurPtr-2);
UnEscapeLexed(*llvmAsmlval.StrVal);
return LABELSTR;
}
}
static bool JustWhitespaceNewLine(const char *&Ptr) {
const char *ThisPtr = Ptr;
while (*ThisPtr == ' ' || *ThisPtr == '\t')
++ThisPtr;
if (*ThisPtr == '\n' || *ThisPtr == '\r') {
Ptr = ThisPtr;
return true;
}
return false;
}
/// LexIdentifier: Handle several related productions:
/// Label [-a-zA-Z$._0-9]+:
/// IntegerType i[0-9]+
/// Keyword sdiv, float, ...
/// HexIntConstant [us]0x[0-9A-Fa-f]+
int LLLexer::LexIdentifier() {
const char *StartChar = CurPtr;
const char *IntEnd = CurPtr[-1] == 'i' ? 0 : StartChar;
const char *KeywordEnd = 0;
for (; isLabelChar(*CurPtr); ++CurPtr) {
// If we decide this is an integer, remember the end of the sequence.
if (!IntEnd && !isdigit(*CurPtr)) IntEnd = CurPtr;
if (!KeywordEnd && !isalnum(*CurPtr) && *CurPtr != '_') KeywordEnd = CurPtr;
}
// If we stopped due to a colon, this really is a label.
if (*CurPtr == ':') {
llvmAsmlval.StrVal = new std::string(StartChar-1, CurPtr++);
return LABELSTR;
}
// Otherwise, this wasn't a label. If this was valid as an integer type,
// return it.
if (IntEnd == 0) IntEnd = CurPtr;
if (IntEnd != StartChar) {
CurPtr = IntEnd;
uint64_t NumBits = atoull(StartChar, CurPtr);
if (NumBits < IntegerType::MIN_INT_BITS ||
NumBits > IntegerType::MAX_INT_BITS) {
GenerateError("Bitwidth for integer type out of range!");
return YYERROR;
}
const Type* Ty = IntegerType::get(NumBits);
llvmAsmlval.PrimType = Ty;
return INTTYPE;
}
// Otherwise, this was a letter sequence. See which keyword this is.
if (KeywordEnd == 0) KeywordEnd = CurPtr;
CurPtr = KeywordEnd;
--StartChar;
unsigned Len = CurPtr-StartChar;
#define KEYWORD(STR, TOK) \
if (Len == strlen(STR) && !memcmp(StartChar, STR, strlen(STR))) return TOK;
KEYWORD("begin", BEGINTOK);
KEYWORD("end", ENDTOK);
KEYWORD("true", TRUETOK);
KEYWORD("false", FALSETOK);
KEYWORD("declare", DECLARE);
KEYWORD("define", DEFINE);
KEYWORD("global", GLOBAL);
KEYWORD("constant", CONSTANT);
KEYWORD("internal", INTERNAL);
KEYWORD("linkonce", LINKONCE);
KEYWORD("weak", WEAK);
KEYWORD("appending", APPENDING);
KEYWORD("dllimport", DLLIMPORT);
KEYWORD("dllexport", DLLEXPORT);
KEYWORD("hidden", HIDDEN);
KEYWORD("protected", PROTECTED);
KEYWORD("extern_weak", EXTERN_WEAK);
KEYWORD("external", EXTERNAL);
KEYWORD("thread_local", THREAD_LOCAL);
KEYWORD("zeroinitializer", ZEROINITIALIZER);
KEYWORD("undef", UNDEF);
KEYWORD("null", NULL_TOK);
KEYWORD("to", TO);
KEYWORD("tail", TAIL);
KEYWORD("target", TARGET);
KEYWORD("triple", TRIPLE);
KEYWORD("deplibs", DEPLIBS);
KEYWORD("datalayout", DATALAYOUT);
KEYWORD("volatile", VOLATILE);
KEYWORD("align", ALIGN);
KEYWORD("section", SECTION);
KEYWORD("alias", ALIAS);
KEYWORD("module", MODULE);
KEYWORD("asm", ASM_TOK);
KEYWORD("sideeffect", SIDEEFFECT);
KEYWORD("cc", CC_TOK);
KEYWORD("ccc", CCC_TOK);
KEYWORD("fastcc", FASTCC_TOK);
KEYWORD("coldcc", COLDCC_TOK);
KEYWORD("x86_stdcallcc", X86_STDCALLCC_TOK);
KEYWORD("x86_fastcallcc", X86_FASTCALLCC_TOK);
KEYWORD("signext", SIGNEXT);
KEYWORD("zeroext", ZEROEXT);
KEYWORD("inreg", INREG);
KEYWORD("sret", SRET);
KEYWORD("nounwind", NOUNWIND);
KEYWORD("noreturn", NORETURN);
KEYWORD("noalias", NOALIAS);
KEYWORD("byval", BYVAL);
KEYWORD("nest", NEST);
KEYWORD("pure", PURE);
KEYWORD("const", CONST);
KEYWORD("type", TYPE);
KEYWORD("opaque", OPAQUE);
KEYWORD("eq" , EQ);
KEYWORD("ne" , NE);
KEYWORD("slt", SLT);
KEYWORD("sgt", SGT);
KEYWORD("sle", SLE);
KEYWORD("sge", SGE);
KEYWORD("ult", ULT);
KEYWORD("ugt", UGT);
KEYWORD("ule", ULE);
KEYWORD("uge", UGE);
KEYWORD("oeq", OEQ);
KEYWORD("one", ONE);
KEYWORD("olt", OLT);
KEYWORD("ogt", OGT);
KEYWORD("ole", OLE);
KEYWORD("oge", OGE);
KEYWORD("ord", ORD);
KEYWORD("uno", UNO);
KEYWORD("ueq", UEQ);
KEYWORD("une", UNE);
#undef KEYWORD
// Keywords for types.
#define TYPEKEYWORD(STR, LLVMTY, TOK) \
if (Len == strlen(STR) && !memcmp(StartChar, STR, strlen(STR))) { \
llvmAsmlval.PrimType = LLVMTY; return TOK; }
TYPEKEYWORD("void", Type::VoidTy, VOID);
TYPEKEYWORD("float", Type::FloatTy, FLOAT);
TYPEKEYWORD("double", Type::DoubleTy, DOUBLE);
TYPEKEYWORD("x86_fp80", Type::X86_FP80Ty, X86_FP80);
TYPEKEYWORD("fp128", Type::FP128Ty, FP128);
TYPEKEYWORD("ppc_fp128", Type::PPC_FP128Ty, PPC_FP128);
TYPEKEYWORD("label", Type::LabelTy, LABEL);
#undef TYPEKEYWORD
// Handle special forms for autoupgrading. Drop these in LLVM 3.0. This is
// to avoid conflicting with the sext/zext instructions, below.
if (Len == 4 && !memcmp(StartChar, "sext", 4)) {
// Scan CurPtr ahead, seeing if there is just whitespace before the newline.
if (JustWhitespaceNewLine(CurPtr))
return SIGNEXT;
} else if (Len == 4 && !memcmp(StartChar, "zext", 4)) {
// Scan CurPtr ahead, seeing if there is just whitespace before the newline.
if (JustWhitespaceNewLine(CurPtr))
return ZEROEXT;
}
// Keywords for instructions.
#define INSTKEYWORD(STR, type, Enum, TOK) \
if (Len == strlen(STR) && !memcmp(StartChar, STR, strlen(STR))) { \
llvmAsmlval.type = Instruction::Enum; return TOK; }
INSTKEYWORD("add", BinaryOpVal, Add, ADD);
INSTKEYWORD("sub", BinaryOpVal, Sub, SUB);
INSTKEYWORD("mul", BinaryOpVal, Mul, MUL);
INSTKEYWORD("udiv", BinaryOpVal, UDiv, UDIV);
INSTKEYWORD("sdiv", BinaryOpVal, SDiv, SDIV);
INSTKEYWORD("fdiv", BinaryOpVal, FDiv, FDIV);
INSTKEYWORD("urem", BinaryOpVal, URem, UREM);
INSTKEYWORD("srem", BinaryOpVal, SRem, SREM);
INSTKEYWORD("frem", BinaryOpVal, FRem, FREM);
INSTKEYWORD("shl", BinaryOpVal, Shl, SHL);
INSTKEYWORD("lshr", BinaryOpVal, LShr, LSHR);
INSTKEYWORD("ashr", BinaryOpVal, AShr, ASHR);
INSTKEYWORD("and", BinaryOpVal, And, AND);
INSTKEYWORD("or", BinaryOpVal, Or , OR );
INSTKEYWORD("xor", BinaryOpVal, Xor, XOR);
INSTKEYWORD("icmp", OtherOpVal, ICmp, ICMP);
INSTKEYWORD("fcmp", OtherOpVal, FCmp, FCMP);
INSTKEYWORD("phi", OtherOpVal, PHI, PHI_TOK);
INSTKEYWORD("call", OtherOpVal, Call, CALL);
INSTKEYWORD("trunc", CastOpVal, Trunc, TRUNC);
INSTKEYWORD("zext", CastOpVal, ZExt, ZEXT);
INSTKEYWORD("sext", CastOpVal, SExt, SEXT);
INSTKEYWORD("fptrunc", CastOpVal, FPTrunc, FPTRUNC);
INSTKEYWORD("fpext", CastOpVal, FPExt, FPEXT);
INSTKEYWORD("uitofp", CastOpVal, UIToFP, UITOFP);
INSTKEYWORD("sitofp", CastOpVal, SIToFP, SITOFP);
INSTKEYWORD("fptoui", CastOpVal, FPToUI, FPTOUI);
INSTKEYWORD("fptosi", CastOpVal, FPToSI, FPTOSI);
INSTKEYWORD("inttoptr", CastOpVal, IntToPtr, INTTOPTR);
INSTKEYWORD("ptrtoint", CastOpVal, PtrToInt, PTRTOINT);
INSTKEYWORD("bitcast", CastOpVal, BitCast, BITCAST);
INSTKEYWORD("select", OtherOpVal, Select, SELECT);
INSTKEYWORD("va_arg", OtherOpVal, VAArg , VAARG);
INSTKEYWORD("ret", TermOpVal, Ret, RET);
INSTKEYWORD("br", TermOpVal, Br, BR);
INSTKEYWORD("switch", TermOpVal, Switch, SWITCH);
INSTKEYWORD("invoke", TermOpVal, Invoke, INVOKE);
INSTKEYWORD("unwind", TermOpVal, Unwind, UNWIND);
INSTKEYWORD("unreachable", TermOpVal, Unreachable, UNREACHABLE);
INSTKEYWORD("malloc", MemOpVal, Malloc, MALLOC);
INSTKEYWORD("alloca", MemOpVal, Alloca, ALLOCA);
INSTKEYWORD("free", MemOpVal, Free, FREE);
INSTKEYWORD("load", MemOpVal, Load, LOAD);
INSTKEYWORD("store", MemOpVal, Store, STORE);
INSTKEYWORD("getelementptr", MemOpVal, GetElementPtr, GETELEMENTPTR);
INSTKEYWORD("extractelement", OtherOpVal, ExtractElement, EXTRACTELEMENT);
INSTKEYWORD("insertelement", OtherOpVal, InsertElement, INSERTELEMENT);
INSTKEYWORD("shufflevector", OtherOpVal, ShuffleVector, SHUFFLEVECTOR);
#undef INSTKEYWORD
// Check for [us]0x[0-9A-Fa-f]+ which are Hexadecimal constant generated by
// the CFE to avoid forcing it to deal with 64-bit numbers.
if ((TokStart[0] == 'u' || TokStart[0] == 's') &&
TokStart[1] == '0' && TokStart[2] == 'x' && isxdigit(TokStart[3])) {
int len = CurPtr-TokStart-3;
uint32_t bits = len * 4;
APInt Tmp(bits, TokStart+3, len, 16);
uint32_t activeBits = Tmp.getActiveBits();
if (activeBits > 0 && activeBits < bits)
Tmp.trunc(activeBits);
if (Tmp.getBitWidth() > 64) {
llvmAsmlval.APIntVal = new APInt(Tmp);
return TokStart[0] == 's' ? ESAPINTVAL : EUAPINTVAL;
} else if (TokStart[0] == 's') {
llvmAsmlval.SInt64Val = Tmp.getSExtValue();
return ESINT64VAL;
} else {
llvmAsmlval.UInt64Val = Tmp.getZExtValue();
return EUINT64VAL;
}
}
// Finally, if this is "cc1234", return this as just "cc".
if (TokStart[0] == 'c' && TokStart[1] == 'c') {
CurPtr = TokStart+2;
return CC_TOK;
}
// Finally, if this isn't known, return just a single character.
CurPtr = TokStart+1;
return TokStart[0];
}
/// Lex0x: Handle productions that start with 0x, knowing that it matches and
/// that this is not a label:
/// HexFPConstant 0x[0-9A-Fa-f]+
/// HexFP80Constant 0xK[0-9A-Fa-f]+
/// HexFP128Constant 0xL[0-9A-Fa-f]+
/// HexPPC128Constant 0xM[0-9A-Fa-f]+
int LLLexer::Lex0x() {
CurPtr = TokStart + 2;
char Kind;
if (CurPtr[0] >= 'K' && CurPtr[0] <= 'M') {
Kind = *CurPtr++;
} else {
Kind = 'J';
}
if (!isxdigit(CurPtr[0])) {
// Bad token, return it as just zero.
CurPtr = TokStart+1;
return '0';
}
while (isxdigit(CurPtr[0]))
++CurPtr;
if (Kind == 'J') {
// HexFPConstant - Floating point constant represented in IEEE format as a
// hexadecimal number for when exponential notation is not precise enough.
// Float and double only.
llvmAsmlval.FPVal = new APFloat(HexToFP(TokStart+2, CurPtr));
return FPVAL;
}
uint64_t Pair[2];
HexToIntPair(TokStart+3, CurPtr, Pair);
switch (Kind) {
default: assert(0 && "Unknown kind!");
case 'K':
// F80HexFPConstant - x87 long double in hexadecimal format (10 bytes)
llvmAsmlval.FPVal = new APFloat(APInt(80, 2, Pair));
return FPVAL;
case 'L':
// F128HexFPConstant - IEEE 128-bit in hexadecimal format (16 bytes)
llvmAsmlval.FPVal = new APFloat(APInt(128, 2, Pair), true);
return FPVAL;
case 'M':
// PPC128HexFPConstant - PowerPC 128-bit in hexadecimal format (16 bytes)
llvmAsmlval.FPVal = new APFloat(APInt(128, 2, Pair));
return FPVAL;
}
}
/// LexIdentifier: Handle several related productions:
/// Label [-a-zA-Z$._0-9]+:
/// NInteger -[0-9]+
/// FPConstant [-+]?[0-9]+[.][0-9]*([eE][-+]?[0-9]+)?
/// PInteger [0-9]+
/// HexFPConstant 0x[0-9A-Fa-f]+
/// HexFP80Constant 0xK[0-9A-Fa-f]+
/// HexFP128Constant 0xL[0-9A-Fa-f]+
/// HexPPC128Constant 0xM[0-9A-Fa-f]+
int LLLexer::LexDigitOrNegative() {
// If the letter after the negative is a number, this is probably a label.
if (!isdigit(TokStart[0]) && !isdigit(CurPtr[0])) {
// Okay, this is not a number after the -, it's probably a label.
if (const char *End = isLabelTail(CurPtr)) {
llvmAsmlval.StrVal = new std::string(TokStart, End-1);
CurPtr = End;
return LABELSTR;
}
return CurPtr[-1];
}
// At this point, it is either a label, int or fp constant.
// Skip digits, we have at least one.
for (; isdigit(CurPtr[0]); ++CurPtr);
// Check to see if this really is a label afterall, e.g. "-1:".
if (isLabelChar(CurPtr[0]) || CurPtr[0] == ':') {
if (const char *End = isLabelTail(CurPtr)) {
llvmAsmlval.StrVal = new std::string(TokStart, End-1);
CurPtr = End;
return LABELSTR;
}
}
// If the next character is a '.', then it is a fp value, otherwise its
// integer.
if (CurPtr[0] != '.') {
if (TokStart[0] == '0' && TokStart[1] == 'x')
return Lex0x();
unsigned Len = CurPtr-TokStart;
uint32_t numBits = ((Len * 64) / 19) + 2;
APInt Tmp(numBits, TokStart, Len, 10);
if (TokStart[0] == '-') {
uint32_t minBits = Tmp.getMinSignedBits();
if (minBits > 0 && minBits < numBits)
Tmp.trunc(minBits);
if (Tmp.getBitWidth() > 64) {
llvmAsmlval.APIntVal = new APInt(Tmp);
return ESAPINTVAL;
} else {
llvmAsmlval.SInt64Val = Tmp.getSExtValue();
return ESINT64VAL;
}
} else {
uint32_t activeBits = Tmp.getActiveBits();
if (activeBits > 0 && activeBits < numBits)
Tmp.trunc(activeBits);
if (Tmp.getBitWidth() > 64) {
llvmAsmlval.APIntVal = new APInt(Tmp);
return EUAPINTVAL;
} else {
llvmAsmlval.UInt64Val = Tmp.getZExtValue();
return EUINT64VAL;
}
}
}
++CurPtr;
// Skip over [0-9]*([eE][-+]?[0-9]+)?
while (isdigit(CurPtr[0])) ++CurPtr;
if (CurPtr[0] == 'e' || CurPtr[0] == 'E') {
if (isdigit(CurPtr[1]) ||
((CurPtr[1] == '-' || CurPtr[1] == '+') && isdigit(CurPtr[2]))) {
CurPtr += 2;
while (isdigit(CurPtr[0])) ++CurPtr;
}
}
llvmAsmlval.FPVal = new APFloat(atof(TokStart));
return FPVAL;
}
/// FPConstant [-+]?[0-9]+[.][0-9]*([eE][-+]?[0-9]+)?
int LLLexer::LexPositive() {
// If the letter after the negative is a number, this is probably not a
// label.
if (!isdigit(CurPtr[0]))
return CurPtr[-1];
// Skip digits.
for (++CurPtr; isdigit(CurPtr[0]); ++CurPtr);
// At this point, we need a '.'.
if (CurPtr[0] != '.') {
CurPtr = TokStart+1;
return TokStart[0];
}
++CurPtr;
// Skip over [0-9]*([eE][-+]?[0-9]+)?
while (isdigit(CurPtr[0])) ++CurPtr;
if (CurPtr[0] == 'e' || CurPtr[0] == 'E') {
if (isdigit(CurPtr[1]) ||
((CurPtr[1] == '-' || CurPtr[1] == '+') && isdigit(CurPtr[2]))) {
CurPtr += 2;
while (isdigit(CurPtr[0])) ++CurPtr;
}
}
llvmAsmlval.FPVal = new APFloat(atof(TokStart));
return FPVAL;
}
//===----------------------------------------------------------------------===//
// Define the interface to this file.
//===----------------------------------------------------------------------===//
static LLLexer *TheLexer;
void InitLLLexer(llvm::MemoryBuffer *MB) {
assert(TheLexer == 0 && "LL Lexer isn't reentrant yet");
TheLexer = new LLLexer(MB);
}
int llvmAsmlex() {
return TheLexer->LexToken();
}
const char *LLLgetTokenStart() { return TheLexer->getTokStart(); }
unsigned LLLgetTokenLength() { return TheLexer->getTokLength(); }
std::string LLLgetFilename() { return TheLexer->getFilename(); }
unsigned LLLgetLineNo() { return TheLexer->getLineNo(); }
void FreeLexer() {
delete TheLexer;
TheLexer = 0;
}

57
lib/AsmParser/LLLexer.h Normal file
View File

@ -0,0 +1,57 @@
//===- LLLexer.h - Lexer for LLVM Assembly Files ----------------*- C++ -*-===//
//
// The LLVM Compiler Infrastructure
//
// This file was developed by Chris Lattner and is distributed under
// the University of Illinois Open Source License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
// This class represents the Lexer for .ll files.
//
//===----------------------------------------------------------------------===//
#ifndef LIB_ASMPARSER_LLLEXER_H
#define LIB_ASMPARSER_LLLEXER_H
#include <vector>
#include <string>
#include <iosfwd>
namespace llvm {
class MemoryBuffer;
class LLLexer {
const char *CurPtr;
unsigned CurLineNo;
MemoryBuffer *CurBuf;
const char *TokStart;
std::string TheError;
public:
LLLexer(MemoryBuffer *StartBuf);
~LLLexer() {}
const char *getTokStart() const { return TokStart; }
unsigned getTokLength() const { return CurPtr-TokStart; }
unsigned getLineNo() const { return CurLineNo; }
std::string getFilename() const;
int LexToken();
const std::string getError() const { return TheError; }
private:
int getNextChar();
void SkipLineComment();
int LexIdentifier();
int LexDigitOrNegative();
int LexPositive();
int LexAt();
int LexPercent();
int LexQuote();
int Lex0x();
};
} // end namespace llvm
#endif

3310
lib/AsmParser/Lexer.cpp.cvs
View File

File diff suppressed because it is too large Load Diff

512
lib/AsmParser/Lexer.l
View File

@ -1,512 +0,0 @@
/*===-- Lexer.l - Scanner for llvm assembly files --------------*- C++ -*--===//
//
// The LLVM Compiler Infrastructure
//
// This file was developed by the LLVM research group and is distributed under
// the University of Illinois Open Source License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
// This file implements the flex scanner for LLVM assembly languages files.
//
//===----------------------------------------------------------------------===*/
%option prefix="llvmAsm"
%option yylineno
%option nostdinit
%option never-interactive
%option batch
%option noyywrap
%option nodefault
%option 8bit
%option outfile="Lexer.cpp"
%option ecs
%option noreject
%option noyymore
%{
#include "ParserInternals.h"
#include "llvm/Module.h"
#include "llvm/Support/MathExtras.h"
#include <list>
#include "llvmAsmParser.h"
#include <cctype>
#include <cstdlib>
void set_scan_file(FILE * F){
yy_switch_to_buffer(yy_create_buffer( F, YY_BUF_SIZE ) );
}
void set_scan_string (const char * str) {
yy_scan_string (str);
}
// Construct a token value for a non-obsolete token
#define RET_TOK(type, Enum, sym) \
llvmAsmlval.type = Instruction::Enum; \
return sym
// Construct a token value for an obsolete token
#define RET_TY(CTYPE, SYM) \
llvmAsmlval.PrimType = CTYPE;\
return SYM
namespace llvm {
// TODO: All of the static identifiers are figured out by the lexer,
// these should be hashed to reduce the lexer size
// atoull - Convert an ascii string of decimal digits into the unsigned long
// long representation... this does not have to do input error checking,
// because we know that the input will be matched by a suitable regex...
//
static uint64_t atoull(const char *Buffer) {
uint64_t Result = 0;
for (; *Buffer; Buffer++) {
uint64_t OldRes = Result;
Result *= 10;
Result += *Buffer-'0';
if (Result < OldRes) // Uh, oh, overflow detected!!!
GenerateError("constant bigger than 64 bits detected!");
}
return Result;
}
static uint64_t HexIntToVal(const char *Buffer) {
uint64_t Result = 0;
for (; *Buffer; ++Buffer) {
uint64_t OldRes = Result;
Result *= 16;
char C = *Buffer;
if (C >= '0' && C <= '9')
Result += C-'0';
else if (C >= 'A' && C <= 'F')
Result += C-'A'+10;
else if (C >= 'a' && C <= 'f')
Result += C-'a'+10;
if (Result < OldRes) // Uh, oh, overflow detected!!!
GenerateError("constant bigger than 64 bits detected!");
}
return Result;
}
// HexToFP - Convert the ascii string in hexadecimal format to the floating
// point representation of it.
//
static double HexToFP(const char *Buffer) {
return BitsToDouble(HexIntToVal(Buffer)); // Cast Hex constant to double
}
static void HexToIntPair(const char *Buffer, uint64_t Pair[2]) {
Pair[0] = 0;
for (int i=0; i<16; i++, Buffer++) {
assert(*Buffer);
Pair[0] *= 16;
char C = *Buffer;
if (C >= '0' && C <= '9')
Pair[0] += C-'0';
else if (C >= 'A' && C <= 'F')
Pair[0] += C-'A'+10;
else if (C >= 'a' && C <= 'f')
Pair[0] += C-'a'+10;
}
Pair[1] = 0;
for (int i=0; i<16 && *Buffer; i++, Buffer++) {
Pair[1] *= 16;
char C = *Buffer;
if (C >= '0' && C <= '9')
Pair[1] += C-'0';
else if (C >= 'A' && C <= 'F')
Pair[1] += C-'A'+10;
else if (C >= 'a' && C <= 'f')
Pair[1] += C-'a'+10;
}
if (*Buffer)
GenerateError("constant bigger than 128 bits detected!");
}
// UnEscapeLexed - Run through the specified buffer and change \xx codes to the
// appropriate character.
char *UnEscapeLexed(char *Buffer, char* EndBuffer) {
char *BOut = Buffer;
for (char *BIn = Buffer; *BIn; ) {
if (BIn[0] == '\\') {
if (BIn < EndBuffer-1 && BIn[1] == '\\') {
*BOut++ = '\\'; // Two \ becomes one
BIn += 2;
} else if (BIn < EndBuffer-2 && isxdigit(BIn[1]) && isxdigit(BIn[2])) {
char Tmp = BIn[3]; BIn[3] = 0; // Terminate string
*BOut = (char)strtol(BIn+1, 0, 16); // Convert to number
BIn[3] = Tmp; // Restore character
BIn += 3; // Skip over handled chars
++BOut;
} else {
*BOut++ = *BIn++;
}
} else {
*BOut++ = *BIn++;
}
}
return BOut;
}
} // End llvm namespace
using namespace llvm;
#define YY_NEVER_INTERACTIVE 1
%}
/* Comments start with a ; and go till end of line */
Comment ;.*
/* Local Values and Type identifiers start with a % sign */
LocalVarName %[-a-zA-Z$._][-a-zA-Z$._0-9]*
/* Global Value identifiers start with an @ sign */
GlobalVarName @[-a-zA-Z$._][-a-zA-Z$._0-9]*
/* Label identifiers end with a colon */
Label [-a-zA-Z$._0-9]+:
QuoteLabel \"[^\"]+\":
/* Quoted names can contain any character except " and \ */
StringConstant \"[^\"]*\"
AtStringConstant @\"[^\"]*\"
PctStringConstant %\"[^\"]*\"
/* LocalVarID/GlobalVarID: match an unnamed local variable slot ID. */
LocalVarID %[0-9]+
GlobalVarID @[0-9]+
/* Integer types are specified with i and a bitwidth */
IntegerType i[0-9]+
/* E[PN]Integer: match positive and negative literal integer values. */
PInteger [0-9]+
NInteger -[0-9]+
/* FPConstant - A Floating point constant. Float and double only.
*/
FPConstant [-+]?[0-9]+[.][0-9]*([eE][-+]?[0-9]+)?
/* HexFPConstant - Floating point constant represented in IEEE format as a
* hexadecimal number for when exponential notation is not precise enough.
* Float and double only.
*/
HexFPConstant 0x[0-9A-Fa-f]+
/* F80HexFPConstant - x87 long double in hexadecimal format (10 bytes)
*/
HexFP80Constant 0xK[0-9A-Fa-f]+
/* F128HexFPConstant - IEEE 128-bit in hexadecimal format (16 bytes)
*/
HexFP128Constant 0xL[0-9A-Fa-f]+
/* PPC128HexFPConstant - PowerPC 128-bit in hexadecimal format (16 bytes)
*/
HexPPC128Constant 0xM[0-9A-Fa-f]+
/* HexIntConstant - Hexadecimal constant generated by the CFE to avoid forcing
* it to deal with 64 bit numbers.
*/
HexIntConstant [us]0x[0-9A-Fa-f]+
/* WSNL - shorthand for whitespace followed by newline */
WSNL [ \r\t]*$
%%
{Comment} { /* Ignore comments for now */ }
begin { return BEGINTOK; }
end { return ENDTOK; }
true { return TRUETOK; }
false { return FALSETOK; }
declare { return DECLARE; }
define { return DEFINE; }
global { return GLOBAL; }
constant { return CONSTANT; }
internal { return INTERNAL; }
linkonce { return LINKONCE; }
weak { return WEAK; }
appending { return APPENDING; }
dllimport { return DLLIMPORT; }
dllexport { return DLLEXPORT; }
hidden { return HIDDEN; }
protected { return PROTECTED; }
extern_weak { return EXTERN_WEAK; }
external { return EXTERNAL; }
thread_local { return THREAD_LOCAL; }
zeroinitializer { return ZEROINITIALIZER; }
\.\.\. { return DOTDOTDOT; }
undef { return UNDEF; }
null { return NULL_TOK; }
to { return TO; }
tail { return TAIL; }
target { return TARGET; }
triple { return TRIPLE; }
deplibs { return DEPLIBS; }
datalayout { return DATALAYOUT; }
volatile { return VOLATILE; }
align { return ALIGN; }
section { return SECTION; }
alias { return ALIAS; }
module { return MODULE; }
asm { return ASM_TOK; }
sideeffect { return SIDEEFFECT; }
cc { return CC_TOK; }
ccc { return CCC_TOK; }
fastcc { return FASTCC_TOK; }
coldcc { return COLDCC_TOK; }
x86_stdcallcc { return X86_STDCALLCC_TOK; }
x86_fastcallcc { return X86_FASTCALLCC_TOK; }
signext { return SIGNEXT; }
zeroext { return ZEROEXT; }
inreg { return INREG; }
sret { return SRET; }
nounwind { return NOUNWIND; }
noreturn { return NORETURN; }
noalias { return NOALIAS; }
byval { return BYVAL; }
nest { return NEST; }
pure { return PURE; }
const { return CONST; }
sext{WSNL} { // For auto-upgrade only, drop in LLVM 3.0
return SIGNEXT; }
zext{WSNL} { // For auto-upgrade only, drop in LLVM 3.0
return ZEROEXT; }
void { RET_TY(Type::VoidTy, VOID); }
float { RET_TY(Type::FloatTy, FLOAT); }
double { RET_TY(Type::DoubleTy,DOUBLE);}
x86_fp80 { RET_TY(Type::X86_FP80Ty, X86_FP80);}
fp128 { RET_TY(Type::FP128Ty, FP128);}
ppc_fp128 { RET_TY(Type::PPC_FP128Ty, PPC_FP128);}
label { RET_TY(Type::LabelTy, LABEL); }
type { return TYPE; }
opaque { return OPAQUE; }
{IntegerType} { uint64_t NumBits = atoull(yytext+1);
if (NumBits < IntegerType::MIN_INT_BITS ||
NumBits > IntegerType::MAX_INT_BITS)
GenerateError("Bitwidth for integer type out of range!");
const Type* Ty = IntegerType::get(NumBits);
RET_TY(Ty, INTTYPE);
}
add { RET_TOK(BinaryOpVal, Add, ADD); }
sub { RET_TOK(BinaryOpVal, Sub, SUB); }
mul { RET_TOK(BinaryOpVal, Mul, MUL); }
udiv { RET_TOK(BinaryOpVal, UDiv, UDIV); }
sdiv { RET_TOK(BinaryOpVal, SDiv, SDIV); }
fdiv { RET_TOK(BinaryOpVal, FDiv, FDIV); }
urem { RET_TOK(BinaryOpVal, URem, UREM); }
srem { RET_TOK(BinaryOpVal, SRem, SREM); }
frem { RET_TOK(BinaryOpVal, FRem, FREM); }
shl { RET_TOK(BinaryOpVal, Shl, SHL); }
lshr { RET_TOK(BinaryOpVal, LShr, LSHR); }
ashr { RET_TOK(BinaryOpVal, AShr, ASHR); }
and { RET_TOK(BinaryOpVal, And, AND); }
or { RET_TOK(BinaryOpVal, Or , OR ); }
xor { RET_TOK(BinaryOpVal, Xor, XOR); }
icmp { RET_TOK(OtherOpVal, ICmp, ICMP); }
fcmp { RET_TOK(OtherOpVal, FCmp, FCMP); }
eq { return EQ; }
ne { return NE; }
slt { return SLT; }
sgt { return SGT; }
sle { return SLE; }
sge { return SGE; }
ult { return ULT; }
ugt { return UGT; }
ule { return ULE; }
uge { return UGE; }
oeq { return OEQ; }
one { return ONE; }
olt { return OLT; }
ogt { return OGT; }
ole { return OLE; }
oge { return OGE; }
ord { return ORD; }
uno { return UNO; }
ueq { return UEQ; }
une { return UNE; }
phi { RET_TOK(OtherOpVal, PHI, PHI_TOK); }
call { RET_TOK(OtherOpVal, Call, CALL); }
trunc { RET_TOK(CastOpVal, Trunc, TRUNC); }
zext { RET_TOK(CastOpVal, ZExt, ZEXT); }
sext { RET_TOK(CastOpVal, SExt, SEXT); }
fptrunc { RET_TOK(CastOpVal, FPTrunc, FPTRUNC); }
fpext { RET_TOK(CastOpVal, FPExt, FPEXT); }
uitofp { RET_TOK(CastOpVal, UIToFP, UITOFP); }
sitofp { RET_TOK(CastOpVal, SIToFP, SITOFP); }
fptoui { RET_TOK(CastOpVal, FPToUI, FPTOUI); }
fptosi { RET_TOK(CastOpVal, FPToSI, FPTOSI); }
inttoptr { RET_TOK(CastOpVal, IntToPtr, INTTOPTR); }
ptrtoint { RET_TOK(CastOpVal, PtrToInt, PTRTOINT); }
bitcast { RET_TOK(CastOpVal, BitCast, BITCAST); }
select { RET_TOK(OtherOpVal, Select, SELECT); }
va_arg { RET_TOK(OtherOpVal, VAArg , VAARG); }
ret { RET_TOK(TermOpVal, Ret, RET); }
br { RET_TOK(TermOpVal, Br, BR); }
switch { RET_TOK(TermOpVal, Switch, SWITCH); }
invoke { RET_TOK(TermOpVal, Invoke, INVOKE); }
unwind { RET_TOK(TermOpVal, Unwind, UNWIND); }
unreachable { RET_TOK(TermOpVal, Unreachable, UNREACHABLE); }
malloc { RET_TOK(MemOpVal, Malloc, MALLOC); }
alloca { RET_TOK(MemOpVal, Alloca, ALLOCA); }
free { RET_TOK(MemOpVal, Free, FREE); }
load { RET_TOK(MemOpVal, Load, LOAD); }
store { RET_TOK(MemOpVal, Store, STORE); }
getelementptr { RET_TOK(MemOpVal, GetElementPtr, GETELEMENTPTR); }
extractelement { RET_TOK(OtherOpVal, ExtractElement, EXTRACTELEMENT); }
insertelement { RET_TOK(OtherOpVal, InsertElement, INSERTELEMENT); }
shufflevector { RET_TOK(OtherOpVal, ShuffleVector, SHUFFLEVECTOR); }
{LocalVarName} {
llvmAsmlval.StrVal = new std::string(yytext+1); // Skip %
return LOCALVAR;
}
{GlobalVarName} {
llvmAsmlval.StrVal = new std::string(yytext+1); // Skip @
return GLOBALVAR;
}
{Label} {
yytext[yyleng-1] = 0; // nuke colon
llvmAsmlval.StrVal = new std::string(yytext);
return LABELSTR;
}
{QuoteLabel} {
yytext[yyleng-2] = 0; // nuke colon, end quote
const char* EndChar = UnEscapeLexed(yytext+1, yytext+yyleng);
llvmAsmlval.StrVal =
new std::string(yytext+1, EndChar - yytext - 1);
return LABELSTR;
}
{StringConstant} { yytext[yyleng-1] = 0; // nuke end quote
const char* EndChar = UnEscapeLexed(yytext+1, yytext+yyleng);
llvmAsmlval.StrVal =
new std::string(yytext+1, EndChar - yytext - 1);
return STRINGCONSTANT;
}
{AtStringConstant} {
yytext[yyleng-1] = 0; // nuke end quote
const char* EndChar =
UnEscapeLexed(yytext+2, yytext+yyleng);
llvmAsmlval.StrVal =
new std::string(yytext+2, EndChar - yytext - 2);
return ATSTRINGCONSTANT;
}
{PctStringConstant} {
yytext[yyleng-1] = 0; // nuke end quote
const char* EndChar =
UnEscapeLexed(yytext+2, yytext+yyleng);
llvmAsmlval.StrVal =
new std::string(yytext+2, EndChar - yytext - 2);
return PCTSTRINGCONSTANT;
}
{PInteger} {
uint32_t numBits = ((yyleng * 64) / 19) + 1;
APInt Tmp(numBits, yytext, yyleng, 10);
uint32_t activeBits = Tmp.getActiveBits();
if (activeBits > 0 && activeBits < numBits)
Tmp.trunc(activeBits);
if (Tmp.getBitWidth() > 64) {
llvmAsmlval.APIntVal = new APInt(Tmp);
return EUAPINTVAL;
} else {
llvmAsmlval.UInt64Val = Tmp.getZExtValue();
return EUINT64VAL;
}
}
{NInteger} {
uint32_t numBits = (((yyleng-1) * 64) / 19) + 2;
APInt Tmp(numBits, yytext, yyleng, 10);
uint32_t minBits = Tmp.getMinSignedBits();
if (minBits > 0 && minBits < numBits)
Tmp.trunc(minBits);
if (Tmp.getBitWidth() > 64) {
llvmAsmlval.APIntVal = new APInt(Tmp);
return ESAPINTVAL;
} else {
llvmAsmlval.SInt64Val = Tmp.getSExtValue();
return ESINT64VAL;
}
}
{HexIntConstant} { int len = yyleng - 3;
uint32_t bits = len * 4;
APInt Tmp(bits, yytext+3, len, 16);
uint32_t activeBits = Tmp.getActiveBits();
if (activeBits > 0 && activeBits < bits)
Tmp.trunc(activeBits);
if (Tmp.getBitWidth() > 64) {
llvmAsmlval.APIntVal = new APInt(Tmp);
return yytext[0] == 's' ? ESAPINTVAL : EUAPINTVAL;
} else if (yytext[0] == 's') {
llvmAsmlval.SInt64Val = Tmp.getSExtValue();
return ESINT64VAL;
} else {
llvmAsmlval.UInt64Val = Tmp.getZExtValue();
return EUINT64VAL;
}
}
{LocalVarID} {
uint64_t Val = atoull(yytext+1);
if ((unsigned)Val != Val)
GenerateError("Invalid value number (too large)!");
llvmAsmlval.UIntVal = unsigned(Val);
return LOCALVAL_ID;
}
{GlobalVarID} {
uint64_t Val = atoull(yytext+1);
if ((unsigned)Val != Val)
GenerateError("Invalid value number (too large)!");
llvmAsmlval.UIntVal = unsigned(Val);
return GLOBALVAL_ID;
}
{FPConstant} { llvmAsmlval.FPVal = new APFloat(atof(yytext)); return FPVAL; }
{HexFPConstant} { llvmAsmlval.FPVal = new APFloat(HexToFP(yytext+2));
return FPVAL;
}
{HexFP80Constant} { uint64_t Pair[2];
HexToIntPair(yytext+3, Pair);
llvmAsmlval.FPVal = new APFloat(APInt(80, 2, Pair));
return FPVAL;
}
{HexFP128Constant} { uint64_t Pair[2];
HexToIntPair(yytext+3, Pair);
llvmAsmlval.FPVal = new APFloat(APInt(128, 2, Pair), true);
return FPVAL;
}
{HexPPC128Constant} { uint64_t Pair[2];
HexToIntPair(yytext+3, Pair);
llvmAsmlval.FPVal = new APFloat(APInt(128, 2, Pair));
return FPVAL;
}
<<EOF>> {
/* Make sure to free the internal buffers for flex when we are
* done reading our input!
*/
yy_delete_buffer(YY_CURRENT_BUFFER);
return EOF;
}
[ \r\t\n] { /* Ignore whitespace */ }
. { return yytext[0]; }
%%

512
lib/AsmParser/Lexer.l.cvs
View File

@ -1,512 +0,0 @@
/*===-- Lexer.l - Scanner for llvm assembly files --------------*- C++ -*--===//
//
// The LLVM Compiler Infrastructure
//
// This file was developed by the LLVM research group and is distributed under
// the University of Illinois Open Source License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
// This file implements the flex scanner for LLVM assembly languages files.
//
//===----------------------------------------------------------------------===*/
%option prefix="llvmAsm"
%option yylineno
%option nostdinit
%option never-interactive
%option batch
%option noyywrap
%option nodefault
%option 8bit
%option outfile="Lexer.cpp"
%option ecs
%option noreject
%option noyymore
%{
#include "ParserInternals.h"
#include "llvm/Module.h"
#include "llvm/Support/MathExtras.h"
#include <list>
#include "llvmAsmParser.h"
#include <cctype>
#include <cstdlib>
void set_scan_file(FILE * F){
yy_switch_to_buffer(yy_create_buffer( F, YY_BUF_SIZE ) );
}
void set_scan_string (const char * str) {
yy_scan_string (str);
}
// Construct a token value for a non-obsolete token
#define RET_TOK(type, Enum, sym) \
llvmAsmlval.type = Instruction::Enum; \
return sym
// Construct a token value for an obsolete token
#define RET_TY(CTYPE, SYM) \
llvmAsmlval.PrimType = CTYPE;\
return SYM
namespace llvm {
// TODO: All of the static identifiers are figured out by the lexer,
// these should be hashed to reduce the lexer size
// atoull - Convert an ascii string of decimal digits into the unsigned long
// long representation... this does not have to do input error checking,
// because we know that the input will be matched by a suitable regex...
//
static uint64_t atoull(const char *Buffer) {
uint64_t Result = 0;
for (; *Buffer; Buffer++) {
uint64_t OldRes = Result;
Result *= 10;
Result += *Buffer-'0';
if (Result < OldRes) // Uh, oh, overflow detected!!!
GenerateError("constant bigger than 64 bits detected!");
}
return Result;
}
static uint64_t HexIntToVal(const char *Buffer) {
uint64_t Result = 0;
for (; *Buffer; ++Buffer) {
uint64_t OldRes = Result;
Result *= 16;
char C = *Buffer;
if (C >= '0' && C <= '9')
Result += C-'0';
else if (C >= 'A' && C <= 'F')
Result += C-'A'+10;
else if (C >= 'a' && C <= 'f')
Result += C-'a'+10;
if (Result < OldRes) // Uh, oh, overflow detected!!!
GenerateError("constant bigger than 64 bits detected!");
}
return Result;
}
// HexToFP - Convert the ascii string in hexadecimal format to the floating
// point representation of it.
//
static double HexToFP(const char *Buffer) {
return BitsToDouble(HexIntToVal(Buffer)); // Cast Hex constant to double
}
static void HexToIntPair(const char *Buffer, uint64_t Pair[2]) {
Pair[0] = 0;
for (int i=0; i<16; i++, Buffer++) {
assert(*Buffer);
Pair[0] *= 16;
char C = *Buffer;
if (C >= '0' && C <= '9')
Pair[0] += C-'0';
else if (C >= 'A' && C <= 'F')
Pair[0] += C-'A'+10;
else if (C >= 'a' && C <= 'f')
Pair[0] += C-'a'+10;
}
Pair[1] = 0;
for (int i=0; i<16 && *Buffer; i++, Buffer++) {
Pair[1] *= 16;
char C = *Buffer;
if (C >= '0' && C <= '9')
Pair[1] += C-'0';
else if (C >= 'A' && C <= 'F')
Pair[1] += C-'A'+10;
else if (C >= 'a' && C <= 'f')
Pair[1] += C-'a'+10;
}
if (*Buffer)
GenerateError("constant bigger than 128 bits detected!");
}
// UnEscapeLexed - Run through the specified buffer and change \xx codes to the
// appropriate character.
char *UnEscapeLexed(char *Buffer, char* EndBuffer) {
char *BOut = Buffer;
for (char *BIn = Buffer; *BIn; ) {
if (BIn[0] == '\\') {
if (BIn < EndBuffer-1 && BIn[1] == '\\') {
*BOut++ = '\\'; // Two \ becomes one
BIn += 2;
} else if (BIn < EndBuffer-2 && isxdigit(BIn[1]) && isxdigit(BIn[2])) {
char Tmp = BIn[3]; BIn[3] = 0; // Terminate string
*BOut = (char)strtol(BIn+1, 0, 16); // Convert to number
BIn[3] = Tmp; // Restore character
BIn += 3; // Skip over handled chars
++BOut;
} else {
*BOut++ = *BIn++;
}
} else {
*BOut++ = *BIn++;
}
}
return BOut;
}
} // End llvm namespace
using namespace llvm;
#define YY_NEVER_INTERACTIVE 1
%}
/* Comments start with a ; and go till end of line */
Comment ;.*
/* Local Values and Type identifiers start with a % sign */
LocalVarName %[-a-zA-Z$._][-a-zA-Z$._0-9]*
/* Global Value identifiers start with an @ sign */
GlobalVarName @[-a-zA-Z$._][-a-zA-Z$._0-9]*
/* Label identifiers end with a colon */
Label [-a-zA-Z$._0-9]+:
QuoteLabel \"[^\"]+\":
/* Quoted names can contain any character except " and \ */
StringConstant \"[^\"]*\"
AtStringConstant @\"[^\"]*\"
PctStringConstant %\"[^\"]*\"
/* LocalVarID/GlobalVarID: match an unnamed local variable slot ID. */
LocalVarID %[0-9]+
GlobalVarID @[0-9]+
/* Integer types are specified with i and a bitwidth */
IntegerType i[0-9]+
/* E[PN]Integer: match positive and negative literal integer values. */
PInteger [0-9]+
NInteger -[0-9]+
/* FPConstant - A Floating point constant. Float and double only.
*/
FPConstant [-+]?[0-9]+[.][0-9]*([eE][-+]?[0-9]+)?
/* HexFPConstant - Floating point constant represented in IEEE format as a
* hexadecimal number for when exponential notation is not precise enough.
* Float and double only.
*/
HexFPConstant 0x[0-9A-Fa-f]+
/* F80HexFPConstant - x87 long double in hexadecimal format (10 bytes)
*/
HexFP80Constant 0xK[0-9A-Fa-f]+
/* F128HexFPConstant - IEEE 128-bit in hexadecimal format (16 bytes)
*/
HexFP128Constant 0xL[0-9A-Fa-f]+
/* PPC128HexFPConstant - PowerPC 128-bit in hexadecimal format (16 bytes)
*/
HexPPC128Constant 0xM[0-9A-Fa-f]+
/* HexIntConstant - Hexadecimal constant generated by the CFE to avoid forcing
* it to deal with 64 bit numbers.
*/
HexIntConstant [us]0x[0-9A-Fa-f]+
/* WSNL - shorthand for whitespace followed by newline */
WSNL [ \r\t]*$
%%
{Comment} { /* Ignore comments for now */ }
begin { return BEGINTOK; }
end { return ENDTOK; }
true { return TRUETOK; }
false { return FALSETOK; }
declare { return DECLARE; }
define { return DEFINE; }
global { return GLOBAL; }
constant { return CONSTANT; }
internal { return INTERNAL; }
linkonce { return LINKONCE; }
weak { return WEAK; }
appending { return APPENDING; }
dllimport { return DLLIMPORT; }
dllexport { return DLLEXPORT; }
hidden { return HIDDEN; }
protected { return PROTECTED; }
extern_weak { return EXTERN_WEAK; }
external { return EXTERNAL; }
thread_local { return THREAD_LOCAL; }
zeroinitializer { return ZEROINITIALIZER; }
\.\.\. { return DOTDOTDOT; }
undef { return UNDEF; }
null { return NULL_TOK; }
to { return TO; }
tail { return TAIL; }
target { return TARGET; }
triple { return TRIPLE; }
deplibs { return DEPLIBS; }
datalayout { return DATALAYOUT; }
volatile { return VOLATILE; }
align { return ALIGN; }
section { return SECTION; }
alias { return ALIAS; }
module { return MODULE; }
asm { return ASM_TOK; }
sideeffect { return SIDEEFFECT; }
cc { return CC_TOK; }
ccc { return CCC_TOK; }
fastcc { return FASTCC_TOK; }
coldcc { return COLDCC_TOK; }
x86_stdcallcc { return X86_STDCALLCC_TOK; }
x86_fastcallcc { return X86_FASTCALLCC_TOK; }
signext { return SIGNEXT; }
zeroext { return ZEROEXT; }
inreg { return INREG; }
sret { return SRET; }
nounwind { return NOUNWIND; }
noreturn { return NORETURN; }
noalias { return NOALIAS; }
byval { return BYVAL; }
nest { return NEST; }
pure { return PURE; }
const { return CONST; }
sext{WSNL} { // For auto-upgrade only, drop in LLVM 3.0
return SIGNEXT; }
zext{WSNL} { // For auto-upgrade only, drop in LLVM 3.0
return ZEROEXT; }
void { RET_TY(Type::VoidTy, VOID); }
float { RET_TY(Type::FloatTy, FLOAT); }
double { RET_TY(Type::DoubleTy,DOUBLE);}
x86_fp80 { RET_TY(Type::X86_FP80Ty, X86_FP80);}
fp128 { RET_TY(Type::FP128Ty, FP128);}
ppc_fp128 { RET_TY(Type::PPC_FP128Ty, PPC_FP128);}
label { RET_TY(Type::LabelTy, LABEL); }
type { return TYPE; }
opaque { return OPAQUE; }
{IntegerType} { uint64_t NumBits = atoull(yytext+1);
if (NumBits < IntegerType::MIN_INT_BITS ||
NumBits > IntegerType::MAX_INT_BITS)
GenerateError("Bitwidth for integer type out of range!");
const Type* Ty = IntegerType::get(NumBits);
RET_TY(Ty, INTTYPE);
}
add { RET_TOK(BinaryOpVal, Add, ADD); }
sub { RET_TOK(BinaryOpVal, Sub, SUB); }
mul { RET_TOK(BinaryOpVal, Mul, MUL); }
udiv { RET_TOK(BinaryOpVal, UDiv, UDIV); }
sdiv { RET_TOK(BinaryOpVal, SDiv, SDIV); }
fdiv { RET_TOK(BinaryOpVal, FDiv, FDIV); }
urem { RET_TOK(BinaryOpVal, URem, UREM); }
srem { RET_TOK(BinaryOpVal, SRem, SREM); }
frem { RET_TOK(BinaryOpVal, FRem, FREM); }
shl { RET_TOK(BinaryOpVal, Shl, SHL); }
lshr { RET_TOK(BinaryOpVal, LShr, LSHR); }
ashr { RET_TOK(BinaryOpVal, AShr, ASHR); }
and { RET_TOK(BinaryOpVal, And, AND); }
or { RET_TOK(BinaryOpVal, Or , OR ); }
xor { RET_TOK(BinaryOpVal, Xor, XOR); }
icmp { RET_TOK(OtherOpVal, ICmp, ICMP); }
fcmp { RET_TOK(OtherOpVal, FCmp, FCMP); }
eq { return EQ; }
ne { return NE; }
slt { return SLT; }
sgt { return SGT; }
sle { return SLE; }
sge { return SGE; }
ult { return ULT; }
ugt { return UGT; }
ule { return ULE; }
uge { return UGE; }
oeq { return OEQ; }
one { return ONE; }
olt { return OLT; }
ogt { return OGT; }
ole { return OLE; }
oge { return OGE; }
ord { return ORD; }
uno { return UNO; }
ueq { return UEQ; }
une { return UNE; }
phi { RET_TOK(OtherOpVal, PHI, PHI_TOK); }
call { RET_TOK(OtherOpVal, Call, CALL); }
trunc { RET_TOK(CastOpVal, Trunc, TRUNC); }
zext { RET_TOK(CastOpVal, ZExt, ZEXT); }
sext { RET_TOK(CastOpVal, SExt, SEXT); }
fptrunc { RET_TOK(CastOpVal, FPTrunc, FPTRUNC); }
fpext { RET_TOK(CastOpVal, FPExt, FPEXT); }
uitofp { RET_TOK(CastOpVal, UIToFP, UITOFP); }
sitofp { RET_TOK(CastOpVal, SIToFP, SITOFP); }
fptoui { RET_TOK(CastOpVal, FPToUI, FPTOUI); }
fptosi { RET_TOK(CastOpVal, FPToSI, FPTOSI); }
inttoptr { RET_TOK(CastOpVal, IntToPtr, INTTOPTR); }
ptrtoint { RET_TOK(CastOpVal, PtrToInt, PTRTOINT); }
bitcast { RET_TOK(CastOpVal, BitCast, BITCAST); }
select { RET_TOK(OtherOpVal, Select, SELECT); }
va_arg { RET_TOK(OtherOpVal, VAArg , VAARG); }
ret { RET_TOK(TermOpVal, Ret, RET); }
br { RET_TOK(TermOpVal, Br, BR); }
switch { RET_TOK(TermOpVal, Switch, SWITCH); }
invoke { RET_TOK(TermOpVal, Invoke, INVOKE); }
unwind { RET_TOK(TermOpVal, Unwind, UNWIND); }
unreachable { RET_TOK(TermOpVal, Unreachable, UNREACHABLE); }
malloc { RET_TOK(MemOpVal, Malloc, MALLOC); }
alloca { RET_TOK(MemOpVal, Alloca, ALLOCA); }
free { RET_TOK(MemOpVal, Free, FREE); }
load { RET_TOK(MemOpVal, Load, LOAD); }
store { RET_TOK(MemOpVal, Store, STORE); }
getelementptr { RET_TOK(MemOpVal, GetElementPtr, GETELEMENTPTR); }
extractelement { RET_TOK(OtherOpVal, ExtractElement, EXTRACTELEMENT); }
insertelement { RET_TOK(OtherOpVal, InsertElement, INSERTELEMENT); }
shufflevector { RET_TOK(OtherOpVal, ShuffleVector, SHUFFLEVECTOR); }
{LocalVarName} {
llvmAsmlval.StrVal = new std::string(yytext+1); // Skip %
return LOCALVAR;
}
{GlobalVarName} {
llvmAsmlval.StrVal = new std::string(yytext+1); // Skip @
return GLOBALVAR;
}
{Label} {
yytext[yyleng-1] = 0; // nuke colon
llvmAsmlval.StrVal = new std::string(yytext);
return LABELSTR;
}
{QuoteLabel} {
yytext[yyleng-2] = 0; // nuke colon, end quote
const char* EndChar = UnEscapeLexed(yytext+1, yytext+yyleng);
llvmAsmlval.StrVal =
new std::string(yytext+1, EndChar - yytext - 1);
return LABELSTR;
}
{StringConstant} { yytext[yyleng-1] = 0; // nuke end quote
const char* EndChar = UnEscapeLexed(yytext+1, yytext+yyleng);
llvmAsmlval.StrVal =
new std::string(yytext+1, EndChar - yytext - 1);
return STRINGCONSTANT;
}
{AtStringConstant} {
yytext[yyleng-1] = 0; // nuke end quote
const char* EndChar =
UnEscapeLexed(yytext+2, yytext+yyleng);
llvmAsmlval.StrVal =
new std::string(yytext+2, EndChar - yytext - 2);
return ATSTRINGCONSTANT;
}
{PctStringConstant} {
yytext[yyleng-1] = 0; // nuke end quote
const char* EndChar =
UnEscapeLexed(yytext+2, yytext+yyleng);
llvmAsmlval.StrVal =
new std::string(yytext+2, EndChar - yytext - 2);
return PCTSTRINGCONSTANT;
}
{PInteger} {
uint32_t numBits = ((yyleng * 64) / 19) + 1;
APInt Tmp(numBits, yytext, yyleng, 10);
uint32_t activeBits = Tmp.getActiveBits();
if (activeBits > 0 && activeBits < numBits)
Tmp.trunc(activeBits);
if (Tmp.getBitWidth() > 64) {
llvmAsmlval.APIntVal = new APInt(Tmp);
return EUAPINTVAL;
} else {
llvmAsmlval.UInt64Val = Tmp.getZExtValue();
return EUINT64VAL;
}
}
{NInteger} {
uint32_t numBits = (((yyleng-1) * 64) / 19) + 2;
APInt Tmp(numBits, yytext, yyleng, 10);
uint32_t minBits = Tmp.getMinSignedBits();
if (minBits > 0 && minBits < numBits)
Tmp.trunc(minBits);
if (Tmp.getBitWidth() > 64) {
llvmAsmlval.APIntVal = new APInt(Tmp);
return ESAPINTVAL;
} else {
llvmAsmlval.SInt64Val = Tmp.getSExtValue();
return ESINT64VAL;
}
}
{HexIntConstant} { int len = yyleng - 3;
uint32_t bits = len * 4;
APInt Tmp(bits, yytext+3, len, 16);
uint32_t activeBits = Tmp.getActiveBits();
if (activeBits > 0 && activeBits < bits)
Tmp.trunc(activeBits);
if (Tmp.getBitWidth() > 64) {
llvmAsmlval.APIntVal = new APInt(Tmp);
return yytext[0] == 's' ? ESAPINTVAL : EUAPINTVAL;
} else if (yytext[0] == 's') {
llvmAsmlval.SInt64Val = Tmp.getSExtValue();
return ESINT64VAL;
} else {
llvmAsmlval.UInt64Val = Tmp.getZExtValue();
return EUINT64VAL;
}
}
{LocalVarID} {
uint64_t Val = atoull(yytext+1);
if ((unsigned)Val != Val)
GenerateError("Invalid value number (too large)!");
llvmAsmlval.UIntVal = unsigned(Val);
return LOCALVAL_ID;
}
{GlobalVarID} {
uint64_t Val = atoull(yytext+1);
if ((unsigned)Val != Val)
GenerateError("Invalid value number (too large)!");
llvmAsmlval.UIntVal = unsigned(Val);
return GLOBALVAL_ID;
}
{FPConstant} { llvmAsmlval.FPVal = new APFloat(atof(yytext)); return FPVAL; }
{HexFPConstant} { llvmAsmlval.FPVal = new APFloat(HexToFP(yytext+2));
return FPVAL;
}
{HexFP80Constant} { uint64_t Pair[2];
HexToIntPair(yytext+3, Pair);
llvmAsmlval.FPVal = new APFloat(APInt(80, 2, Pair));
return FPVAL;
}
{HexFP128Constant} { uint64_t Pair[2];
HexToIntPair(yytext+3, Pair);
llvmAsmlval.FPVal = new APFloat(APInt(128, 2, Pair), true);
return FPVAL;
}
{HexPPC128Constant} { uint64_t Pair[2];
HexToIntPair(yytext+3, Pair);
llvmAsmlval.FPVal = new APFloat(APInt(128, 2, Pair));
return FPVAL;
}
<<EOF>> {
/* Make sure to free the internal buffers for flex when we are
* done reading our input!
*/
yy_delete_buffer(YY_CURRENT_BUFFER);
return EOF;
}
[ \r\t\n] { /* Ignore whitespace */ }
. { return yytext[0]; }
%%

5
lib/AsmParser/Makefile
View File

@ -10,8 +10,7 @@
LEVEL = ../.. LEVEL = ../..
LIBRARYNAME := LLVMAsmParser LIBRARYNAME := LLVMAsmParser
BUILD_ARCHIVE = 1 BUILD_ARCHIVE = 1
EXTRA_DIST := Lexer.cpp.cvs Lexer.l.cvs \ EXTRA_DIST := llvmAsmParser.cpp.cvs llvmAsmParser.h.cvs llvmAsmParser.y.cvs
llvmAsmParser.cpp.cvs llvmAsmParser.h.cvs llvmAsmParser.y.cvs
include $(LEVEL)/Makefile.common include $(LEVEL)/Makefile.common
@ -24,4 +23,4 @@ CompileCommonOpts := $(filter-out -Wno-long-long,$(CompileCommonOpts))
# Make the object code file for the lexer depend upon the header file generated # Make the object code file for the lexer depend upon the header file generated
# by the Bison parser. This prevents the Lexer from being compiled before the # by the Bison parser. This prevents the Lexer from being compiled before the
# header file it needs is built. # header file it needs is built.
$(ObjDir)/Lexer.o: $(PROJ_SRC_DIR)/llvmAsmParser.h $(ObjDir)/LLLexer.o: $(PROJ_SRC_DIR)/llvmAsmParser.h

44
lib/AsmParser/Parser.cpp
View File

@ -13,38 +13,37 @@
#include "ParserInternals.h" #include "ParserInternals.h"
#include "llvm/Module.h" #include "llvm/Module.h"
#include "llvm/Support/MemoryBuffer.h"
using namespace llvm; using namespace llvm;
ParseError* TheParseError = 0; /// FIXME: Not threading friendly ParseError* TheParseError = 0; /// FIXME: Not threading friendly
Module *llvm::ParseAssemblyFile(const std::string &Filename, ParseError* Err) { Module *llvm::ParseAssemblyFile(const std::string &Filename, ParseError* Err) {
FILE *F = stdin; std::string ErrorStr;
MemoryBuffer *F = MemoryBuffer::getFileOrSTDIN(&Filename[0], Filename.size(),
if (Filename != "-") { &ErrorStr);
F = fopen(Filename.c_str(), "r"); if (F == 0) {
if (Err)
if (F == 0) { Err->setError(Filename, "Could not open input file '" + Filename + "'");
if (Err) return 0;
Err->setError(Filename,"Could not open file '" + Filename + "'");
return 0;
}
} }
TheParseError = Err; TheParseError = Err;
Module *Result = RunVMAsmParser(Filename, F); Module *Result = RunVMAsmParser(F);
delete F;
if (F != stdin)
fclose(F);
return Result; return Result;
} }
Module *llvm::ParseAssemblyString( Module *llvm::ParseAssemblyString(const char *AsmString, Module *M,
const char * AsmString, Module * M, ParseError* Err) ParseError *Err) {
{
TheParseError = Err; TheParseError = Err;
return RunVMAsmParser(AsmString, M); MemoryBuffer *F = MemoryBuffer::getMemBuffer(AsmString,
AsmString+strlen(AsmString),
"<string>");
Module *Result = RunVMAsmParser(F);
delete F;
return Result;
} }
@ -54,9 +53,8 @@ Module *llvm::ParseAssemblyString(
void ParseError::setError(const std::string &filename, void ParseError::setError(const std::string &filename,
const std::string &message, const std::string &message,
int lineNo, int colNo) int lineNo, int colNo) {
{
Filename = filename; Filename = filename;
Message = message; Message = message;
LineNo = lineNo; LineNo = lineNo;

30
lib/AsmParser/ParserInternals.h
View File

@ -23,37 +23,25 @@
#include "llvm/Assembly/Parser.h" #include "llvm/Assembly/Parser.h"
#include "llvm/ADT/StringExtras.h" #include "llvm/ADT/StringExtras.h"
#include "llvm/ADT/APFloat.h" #include "llvm/ADT/APFloat.h"
namespace llvm { class MemoryBuffer; }
// Global variables exported from the lexer... // Global variables exported from the lexer...
extern int llvmAsmlineno; /// FIXME: Not threading friendly
extern llvm::ParseError* TheParseError; /// FIXME: Not threading friendly extern llvm::ParseError* TheParseError; /// FIXME: Not threading friendly
extern std::string &llvmAsmTextin;
// functions exported from the lexer // functions exported from the lexer
void set_scan_file(FILE * F); void InitLLLexer(llvm::MemoryBuffer *MB);
void set_scan_string (const char * str); const char *LLLgetTokenStart();
unsigned LLLgetTokenLength();
// Globals exported by the parser... std::string LLLgetFilename();
extern char* llvmAsmtext; unsigned LLLgetLineNo();
extern int llvmAsmleng; void FreeLexer();
namespace llvm { namespace llvm {
class Module; class Module;
// Globals exported by the parser... // RunVMAsmParser - Parse a buffer and return Module
extern std::string CurFilename; /// FIXME: Not threading friendly Module *RunVMAsmParser(llvm::MemoryBuffer *MB);
// RunVMAsmParser - Parse a file and return Module
Module *RunVMAsmParser(const std::string &Filename, FILE *F);
// Parse a string directly
Module *RunVMAsmParser(const char * AsmString, Module * M);
// UnEscapeLexed - Run through the specified buffer and change \xx codes to the
// appropriate character.
char *UnEscapeLexed(char *Buffer);
// GenerateError - Wrapper around the ParseException class that automatically // GenerateError - Wrapper around the ParseException class that automatically
// fills in file line number and column number and options info. // fills in file line number and column number and options info.

57
lib/AsmParser/llvmAsmParser.y
View File

@ -29,9 +29,6 @@
#include <list> #include <list>
#include <map> #include <map>
#include <utility> #include <utility>
#ifndef NDEBUG
#define YYDEBUG 1
#endif
// The following is a gross hack. In order to rid the libAsmParser library of // The following is a gross hack. In order to rid the libAsmParser library of
// exceptions, we have to have a way of getting the yyparse function to go into // exceptions, we have to have a way of getting the yyparse function to go into
@ -51,15 +48,6 @@ static bool TriggerError = false;
int yyerror(const char *ErrorMsg); // Forward declarations to prevent "implicit int yyerror(const char *ErrorMsg); // Forward declarations to prevent "implicit
int yylex(); // declaration" of xxx warnings. int yylex(); // declaration" of xxx warnings.
int yyparse(); int yyparse();
namespace llvm {
std::string CurFilename;
#if YYDEBUG
static cl::opt<bool>
Debug("debug-yacc", cl::desc("Print yacc debug state changes"),
cl::Hidden, cl::init(false));
#endif
}
using namespace llvm; using namespace llvm;
static Module *ParserResult; static Module *ParserResult;
@ -513,7 +501,7 @@ static Value *getVal(const Type *Ty, const ValID &ID) {
// Remember where this forward reference came from. FIXME, shouldn't we try // Remember where this forward reference came from. FIXME, shouldn't we try
// to recycle these things?? // to recycle these things??
CurModule.PlaceHolderInfo.insert(std::make_pair(V, std::make_pair(ID, CurModule.PlaceHolderInfo.insert(std::make_pair(V, std::make_pair(ID,
llvmAsmlineno))); LLLgetLineNo())));
if (inFunctionScope()) if (inFunctionScope())
InsertValue(V, CurFun.LateResolveValues); InsertValue(V, CurFun.LateResolveValues);
@ -945,22 +933,11 @@ static PATypeHolder HandleUpRefs(const Type *ty) {
// //
static Module* RunParser(Module * M); static Module* RunParser(Module * M);
Module *llvm::RunVMAsmParser(const std::string &Filename, FILE *F) { Module *llvm::RunVMAsmParser(llvm::MemoryBuffer *MB) {
set_scan_file(F); InitLLLexer(MB);
Module *M = RunParser(new Module(LLLgetFilename()));
CurFilename = Filename; FreeLexer();
return RunParser(new Module(CurFilename)); return M;
}
Module *llvm::RunVMAsmParser(const char * AsmString, Module * M) {
set_scan_string(AsmString);
CurFilename = "from_memory";
if (M == NULL) {
return RunParser(new Module (CurFilename));
} else {
return RunParser(M);
}
} }
%} %}
@ -3118,13 +3095,7 @@ MemoryInst : MALLOC Types OptCAlign {
// common code from the two 'RunVMAsmParser' functions // common code from the two 'RunVMAsmParser' functions
static Module* RunParser(Module * M) { static Module* RunParser(Module * M) {
llvmAsmlineno = 1; // Reset the current line number...
CurModule.CurrentModule = M; CurModule.CurrentModule = M;
#if YYDEBUG
yydebug = Debug;
#endif
// Check to make sure the parser succeeded // Check to make sure the parser succeeded
if (yyparse()) { if (yyparse()) {
if (ParserResult) if (ParserResult)
@ -3176,21 +3147,21 @@ static Module* RunParser(Module * M) {
} }
void llvm::GenerateError(const std::string &message, int LineNo) { void llvm::GenerateError(const std::string &message, int LineNo) {
if (LineNo == -1) LineNo = llvmAsmlineno; if (LineNo == -1) LineNo = LLLgetLineNo();
// TODO: column number in exception // TODO: column number in exception
if (TheParseError) if (TheParseError)
TheParseError->setError(CurFilename, message, LineNo); TheParseError->setError(LLLgetFilename(), message, LineNo);
TriggerError = 1; TriggerError = 1;
} }
int yyerror(const char *ErrorMsg) { int yyerror(const char *ErrorMsg) {
std::string where std::string where = LLLgetFilename() + ":" + utostr(LLLgetLineNo()) + ": ";
= std::string((CurFilename == "-") ? std::string("<stdin>") : CurFilename)
+ ":" + utostr((unsigned) llvmAsmlineno) + ": ";
std::string errMsg = where + "error: " + std::string(ErrorMsg); std::string errMsg = where + "error: " + std::string(ErrorMsg);
if (yychar != YYEMPTY && yychar != 0) if (yychar != YYEMPTY && yychar != 0) {
errMsg += " while reading token: '" + std::string(llvmAsmtext, llvmAsmleng)+ errMsg += " while reading token: '";
"'"; errMsg += std::string(LLLgetTokenStart(),
LLLgetTokenStart()+LLLgetTokenLength()) + "'";
}
GenerateError(errMsg); GenerateError(errMsg);
return 0; return 0;
} }