/*===-- 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.
//
//===----------------------------------------------------------------------===*/
#include "llvm/Module.h"
#include <list>
#include "llvmAsmParser.h"
-#include <ctype.h>
-#include <stdlib.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
+ llvmAsmlval.type = Instruction::Enum; \
+ return sym
+// Construct a token value for an obsolete token
+#define RET_TY(CTYPE, SYM) \
+ llvmAsmlval.PrimType = CTYPE;\
+ return SYM
-// TODO: All of the static identifiers are figured out by the lexer,
+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,
+// 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) {
Result *= 10;
Result += *Buffer-'0';
if (Result < OldRes) // Uh, oh, overflow detected!!!
- ThrowException("constant bigger than 64 bits detected!");
+ GenerateError("constant bigger than 64 bits detected!");
}
return Result;
}
Result += C-'a'+10;
if (Result < OldRes) // Uh, oh, overflow detected!!!
- ThrowException("constant bigger than 64 bits detected!");
+ GenerateError("constant bigger than 64 bits detected!");
}
return Result;
}
for (char *BIn = Buffer; *BIn; ) {
if (BIn[0] == '\\' && isxdigit(BIn[1]) && isxdigit(BIn[2])) {
char Tmp = BIn[3]; BIn[3] = 0; // Terminate string
- *BOut = strtol(BIn+1, 0, 16); // Convert to number
+ *BOut = (char)strtol(BIn+1, 0, 16); // Convert to number
if (!AllowNull && !*BOut)
- ThrowException("String literal cannot accept \\00 escape!");
-
+ GenerateError("String literal cannot accept \\00 escape!");
+
BIn[3] = Tmp; // Restore character
BIn += 3; // Skip over handled chars
++BOut;
return BOut;
}
+} // End llvm namespace
+
+using namespace llvm;
+
#define YY_NEVER_INTERACTIVE 1
%}
/* Comments start with a ; and go till end of line */
Comment ;.*
-/* Variable(Value) identifiers start with a % sign */
-VarID %[-a-zA-Z$._][-a-zA-Z$._0-9]*
+/* 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 \"[^\"]+\"
-
-
-/* [PN]Integer: match positive and negative literal integer values that
- * are preceeded by a '%' character. These represent unnamed variable slots.
- */
-EPInteger %[0-9]+
-ENInteger %-[0-9]+
+StringConstant \"[^\"]*\"
+AtStringConstant @\"[^\"]*\"
+
+/* 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 */
+/* E[PN]Integer: match positive and negative literal integer values. */
PInteger [0-9]+
NInteger -[0-9]+
* it to deal with 64 bit numbers.
*/
HexIntConstant [us]0x[0-9A-Fa-f]+
+
%%
{Comment} { /* Ignore comments for now */ }
begin { return BEGINTOK; }
end { return ENDTOK; }
-true { return TRUE; }
-false { return FALSE; }
+true { return TRUETOK; }
+false { return FALSETOK; }
declare { return DECLARE; }
+define { return DEFINE; }
global { return GLOBAL; }
constant { return CONSTANT; }
-const { return CONST; }
internal { return INTERNAL; }
linkonce { return LINKONCE; }
+weak { return WEAK; }
appending { return APPENDING; }
-uninitialized { return EXTERNAL; } /* Deprecated, turn into external */
+dllimport { return DLLIMPORT; }
+dllexport { return DLLEXPORT; }
+hidden { return HIDDEN; }
+extern_weak { return EXTERN_WEAK; }
external { return EXTERNAL; }
-implementation { return IMPLEMENTATION; }
+thread_local { return THREAD_LOCAL; }
+zeroinitializer { return ZEROINITIALIZER; }
\.\.\. { return DOTDOTDOT; }
+undef { return UNDEF; }
null { return NULL_TOK; }
to { return TO; }
-except { return EXCEPT; }
-not { return NOT; } /* Deprecated, turned into XOR */
+tail { return TAIL; }
target { return TARGET; }
-endian { return ENDIAN; }
-pointersize { return POINTERSIZE; }
-little { return LITTLE; }
-big { return BIG; }
-
-void { llvmAsmlval.PrimType = Type::VoidTy ; return VOID; }
-bool { llvmAsmlval.PrimType = Type::BoolTy ; return BOOL; }
-sbyte { llvmAsmlval.PrimType = Type::SByteTy ; return SBYTE; }
-ubyte { llvmAsmlval.PrimType = Type::UByteTy ; return UBYTE; }
-short { llvmAsmlval.PrimType = Type::ShortTy ; return SHORT; }
-ushort { llvmAsmlval.PrimType = Type::UShortTy; return USHORT; }
-int { llvmAsmlval.PrimType = Type::IntTy ; return INT; }
-uint { llvmAsmlval.PrimType = Type::UIntTy ; return UINT; }
-long { llvmAsmlval.PrimType = Type::LongTy ; return LONG; }
-ulong { llvmAsmlval.PrimType = Type::ULongTy ; return ULONG; }
-float { llvmAsmlval.PrimType = Type::FloatTy ; return FLOAT; }
-double { llvmAsmlval.PrimType = Type::DoubleTy; return DOUBLE; }
-type { llvmAsmlval.PrimType = Type::TypeTy ; return TYPE; }
-label { llvmAsmlval.PrimType = Type::LabelTy ; return LABEL; }
+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; }
+
+inreg { return INREG; }
+sret { return SRET; }
+nounwind { return NOUNWIND; }
+noreturn { return NORETURN; }
+
+void { RET_TY(Type::VoidTy, VOID); }
+float { RET_TY(Type::FloatTy, FLOAT); }
+double { RET_TY(Type::DoubleTy,DOUBLE);}
+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); }
-div { RET_TOK(BinaryOpVal, Div, DIV); }
-rem { RET_TOK(BinaryOpVal, Rem, REM); }
+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); }
-setne { RET_TOK(BinaryOpVal, SetNE, SETNE); }
-seteq { RET_TOK(BinaryOpVal, SetEQ, SETEQ); }
-setlt { RET_TOK(BinaryOpVal, SetLT, SETLT); }
-setgt { RET_TOK(BinaryOpVal, SetGT, SETGT); }
-setle { RET_TOK(BinaryOpVal, SetLE, SETLE); }
-setge { RET_TOK(BinaryOpVal, SetGE, SETGE); }
-
-phi { RET_TOK(OtherOpVal, PHINode, PHI); }
+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); }
-cast { RET_TOK(OtherOpVal, Cast, CAST); }
-shl { RET_TOK(OtherOpVal, Shl, SHL); }
-shr { RET_TOK(OtherOpVal, Shr, SHR); }
-
+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); }
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); }
-{VarID} {
+
+{LocalVarName} {
UnEscapeLexed(yytext+1);
llvmAsmlval.StrVal = strdup(yytext+1); // Skip %
- return VAR_ID;
+ return LOCALVAR;
+ }
+{GlobalVarName} {
+ UnEscapeLexed(yytext+1);
+ llvmAsmlval.StrVal = strdup(yytext+1); // Skip @
+ return GLOBALVAR;
}
{Label} {
yytext[strlen(yytext)-1] = 0; // nuke colon
UnEscapeLexed(yytext);
- llvmAsmlval.StrVal = strdup(yytext);
- return LABELSTR;
+ llvmAsmlval.StrVal = strdup(yytext);
+ return LABELSTR;
+ }
+{QuoteLabel} {
+ yytext[strlen(yytext)-2] = 0; // nuke colon, end quote
+ UnEscapeLexed(yytext+1);
+ llvmAsmlval.StrVal = strdup(yytext+1);
+ return LABELSTR;
}
{StringConstant} { // Note that we cannot unescape a string constant here! The
- // string constant might contain a \00 which would not be
+ // string constant might contain a \00 which would not be
// understood by the string stuff. It is valid to make a
// [sbyte] c"Hello World\00" constant, for example.
//
- yytext[strlen(yytext)-1] = 0; // nuke end quote
- llvmAsmlval.StrVal = strdup(yytext+1); // Nuke start quote
- return STRINGCONSTANT;
+ yytext[strlen(yytext)-1] = 0; // nuke end quote
+ llvmAsmlval.StrVal = strdup(yytext+1); // Nuke start quote
+ return STRINGCONSTANT;
}
+{AtStringConstant} {
+ yytext[strlen(yytext)-1] = 0; // nuke end quote
+ llvmAsmlval.StrVal = strdup(yytext+2); // Nuke @, quote
+ return ATSTRINGCONSTANT;
+ }
+
+{PInteger} { int len = strlen(yytext);
+ uint32_t numBits = ((len * 64) / 19) + 1;
+ APInt Tmp(numBits, yytext, len, 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} { int len = strlen(yytext);
+ uint32_t numBits = (((len-1) * 64) / 19) + 2;
+ APInt Tmp(numBits, yytext, len, 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 = strlen(yytext+3) - 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;
+ }
+ }
-{PInteger} { llvmAsmlval.UInt64Val = atoull(yytext); return EUINT64VAL; }
-{NInteger} {
+{LocalVarID} {
uint64_t Val = atoull(yytext+1);
- // +1: we have bigger negative range
- if (Val > (uint64_t)INT64_MAX+1)
- ThrowException("Constant too large for signed 64 bits!");
- llvmAsmlval.SInt64Val = -Val;
- return ESINT64VAL;
+ if ((unsigned)Val != Val)
+ GenerateError("Invalid value number (too large)!");
+ llvmAsmlval.UIntVal = unsigned(Val);
+ return LOCALVAL_ID;
}
-{HexIntConstant} {
- llvmAsmlval.UInt64Val = HexIntToVal(yytext+3);
- return yytext[0] == 's' ? ESINT64VAL : EUINT64VAL;
- }
-
-{EPInteger} { llvmAsmlval.UIntVal = atoull(yytext+1); return UINTVAL; }
-{ENInteger} {
- uint64_t Val = atoull(yytext+2);
- // +1: we have bigger negative range
- if (Val > (uint64_t)INT32_MAX+1)
- ThrowException("Constant too large for signed 32 bits!");
- llvmAsmlval.SIntVal = -Val;
- return SINTVAL;
+{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 = atof(yytext); return FPVAL; }
{HexFPConstant} { llvmAsmlval.FPVal = HexToFP(yytext); return FPVAL; }
-[ \t\n] { /* Ignore whitespace */ }
+<<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]; }
%%