2 files changed, 114 insertions, 111 deletions

diff --git a/llvm/lib/MC/MCParser/AsmLexer.cpp b/llvm/lib/MC/MCParser/AsmLexer.cpp
index f653304186b..d56071aea4d 100644
--- a/llvm/lib/MC/MCParser/AsmLexer.cpp
+++ b/llvm/lib/MC/MCParser/AsmLexer.cpp
@@ -23,7 +23,8 @@ using namespace llvm;
 
 AsmLexer::AsmLexer(const MCAsmInfo &MAI) : MAI(MAI) {
   CurPtr = nullptr;
-  isAtStartOfLine = true;
+  IsAtStartOfLine = true;
+  IsAtStartOfStatement = true;
   AllowAtInIdentifier = !StringRef(MAI.getCommentString()).startswith("@");
 }
 
@@ -50,20 +51,9 @@ AsmToken AsmLexer::ReturnError(const char *Loc, const std::string &Msg) {
 }
 
 int AsmLexer::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.end())
-      return 0;  // Just whitespace.
-
-    // Otherwise, return end of file.
-    --CurPtr;  // Another call to lex will return EOF again.
+  if (CurPtr == CurBuf.end())
     return EOF;
-  }
+  return (unsigned char)*CurPtr++;
 }
 
 /// LexFloatLiteral: [0-9]*[.][0-9]*([eE][+-]?[0-9]*)?
@@ -169,43 +159,52 @@ AsmToken AsmLexer::LexIdentifier() {
 AsmToken AsmLexer::LexSlash() {
   switch (*CurPtr) {
   case '*':
+    IsAtStartOfStatement = false;
     break; // C style comment.
   case '/':
     ++CurPtr;
     return LexLineComment();
   default:
-    return AsmToken(AsmToken::Slash, StringRef(CurPtr - 1, 1));
+    IsAtStartOfStatement = false;
+    return AsmToken(AsmToken::Slash, StringRef(TokStart, 1));
   }
 
   // C Style comment.
   ++CurPtr;  // skip the star.
-  while (1) {
-    int CurChar = getNextChar();
-    switch (CurChar) {
-    case EOF:
-      return ReturnError(TokStart, "unterminated comment");
+  while (CurPtr != CurBuf.end()) {
+    switch (*CurPtr++) {
     case '*':
       // End of the comment?
-      if (CurPtr[0] != '/') break;
-
+      if (*CurPtr != '/')
+        break;
       ++CurPtr;   // End the */.
-      return LexToken();
+      return AsmToken(AsmToken::Comment,
+                      StringRef(TokStart, CurPtr - TokStart));
     }
   }
+  return ReturnError(TokStart, "unterminated comment");
 }
 
 /// LexLineComment: Comment: #[^\n]*
 ///                        : //[^\n]*
 AsmToken AsmLexer::LexLineComment() {
-  // FIXME: This is broken if we happen to a comment at the end of a file, which
-  // was .included, and which doesn't end with a newline.
+  // Mark This as an end of statement with a body of the
+  // comment. While it would be nicer to leave this two tokens,
+  // backwards compatability with TargetParsers makes keeping this in this form
+  // better.
   int CurChar = getNextChar();
   while (CurChar != '\n' && CurChar != '\r' && CurChar != EOF)
     CurChar = getNextChar();
 
-  if (CurChar == EOF)
-    return AsmToken(AsmToken::Eof, StringRef(TokStart, 0));
-  return AsmToken(AsmToken::EndOfStatement, StringRef(TokStart, 0));
+  IsAtStartOfLine = true;
+  // Whis is a whole line comment. leave newline
+  if (IsAtStartOfStatement)
+    return AsmToken(AsmToken::EndOfStatement,
+                    StringRef(TokStart, CurPtr - TokStart));
+  IsAtStartOfStatement = true;
+
+  return AsmToken(AsmToken::EndOfStatement,
+                  StringRef(TokStart, CurPtr - 1 - TokStart));
 }
 
 static void SkipIgnoredIntegerSuffix(const char *&CurPtr) {
@@ -423,8 +422,7 @@ StringRef AsmLexer::LexUntilEndOfStatement() {
 
   while (!isAtStartOfComment(CurPtr) &&     // Start of line comment.
          !isAtStatementSeparator(CurPtr) && // End of statement marker.
-         *CurPtr != '\n' && *CurPtr != '\r' &&
-         (*CurPtr != 0 || CurPtr != CurBuf.end())) {
+         *CurPtr != '\n' && *CurPtr != '\r' && CurPtr != CurBuf.end()) {
     ++CurPtr;
   }
   return StringRef(TokStart, CurPtr-TokStart);
@@ -433,8 +431,7 @@ StringRef AsmLexer::LexUntilEndOfStatement() {
 StringRef AsmLexer::LexUntilEndOfLine() {
   TokStart = CurPtr;
 
-  while (*CurPtr != '\n' && *CurPtr != '\r' &&
-         (*CurPtr != 0 || CurPtr != CurBuf.end())) {
+  while (*CurPtr != '\n' && *CurPtr != '\r' && CurPtr != CurBuf.end()) {
     ++CurPtr;
   }
   return StringRef(TokStart, CurPtr-TokStart);
@@ -444,7 +441,8 @@ size_t AsmLexer::peekTokens(MutableArrayRef<AsmToken> Buf,
                             bool ShouldSkipSpace) {
   const char *SavedTokStart = TokStart;
   const char *SavedCurPtr = CurPtr;
-  bool SavedAtStartOfLine = isAtStartOfLine;
+  bool SavedAtStartOfLine = IsAtStartOfLine;
+  bool SavedAtStartOfStatement = IsAtStartOfStatement;
   bool SavedSkipSpace = SkipSpace;
 
   std::string SavedErr = getErr();
@@ -465,7 +463,8 @@ size_t AsmLexer::peekTokens(MutableArrayRef<AsmToken> Buf,
   SetError(SavedErrLoc, SavedErr);
 
   SkipSpace = SavedSkipSpace;
-  isAtStartOfLine = SavedAtStartOfLine;
+  IsAtStartOfLine = SavedAtStartOfLine;
+  IsAtStartOfStatement = SavedAtStartOfStatement;
   CurPtr = SavedCurPtr;
   TokStart = SavedTokStart;
 
@@ -495,29 +494,45 @@ AsmToken AsmLexer::LexToken() {
   // This always consumes at least one character.
   int CurChar = getNextChar();
 
-  if (isAtStartOfComment(TokStart)) {
-    // If this comment starts with a '#', then return the Hash token and let
-    // the assembler parser see if it can be parsed as a cpp line filename
-    // comment. We do this only if we are at the start of a line.
-    if (CurChar == '#' && isAtStartOfLine)
-      return AsmToken(AsmToken::Hash, StringRef(TokStart, 1));
-    isAtStartOfLine = true;
+  if (CurChar == '#' && IsAtStartOfStatement) {
+    // If this starts with a '#', this may be a cpp
+    // hash directive and otherwise a line comment.
+    AsmToken TokenBuf[2];
+    MutableArrayRef<AsmToken> Buf(TokenBuf, 2);
+    size_t num = peekTokens(Buf, true);
+    // There cannot be a space preceeding this
+    if (IsAtStartOfLine && num == 2 && TokenBuf[0].is(AsmToken::Integer) &&
+        TokenBuf[1].is(AsmToken::String)) {
+      CurPtr = TokStart; // reset curPtr;
+      StringRef s = LexUntilEndOfLine();
+      UnLex(TokenBuf[1]);
+      UnLex(TokenBuf[0]);
+      return AsmToken(AsmToken::HashDirective, s);
+    }
     return LexLineComment();
   }
+
+  if (isAtStartOfComment(TokStart))
+    return LexLineComment();
+
   if (isAtStatementSeparator(TokStart)) {
     CurPtr += strlen(MAI.getSeparatorString()) - 1;
+    IsAtStartOfLine = true;
+    IsAtStartOfStatement = true;
     return AsmToken(AsmToken::EndOfStatement,
                     StringRef(TokStart, strlen(MAI.getSeparatorString())));
   }
 
   // If we're missing a newline at EOF, make sure we still get an
   // EndOfStatement token before the Eof token.
-  if (CurChar == EOF && !isAtStartOfLine) {
-    isAtStartOfLine = true;
+  if (CurChar == EOF && !IsAtStartOfStatement) {
+    IsAtStartOfLine = true;
+    IsAtStartOfStatement = true;
     return AsmToken(AsmToken::EndOfStatement, StringRef(TokStart, 1));
   }
-
-  isAtStartOfLine = false;
+  IsAtStartOfLine = false;
+  bool OldIsAtStartOfStatement = IsAtStartOfStatement;
+  IsAtStartOfStatement = false;
   switch (CurChar) {
   default:
     // Handle identifier: [a-zA-Z_.][a-zA-Z0-9_$.@]*
@@ -526,24 +541,24 @@ AsmToken AsmLexer::LexToken() {
 
     // Unknown character, emit an error.
     return ReturnError(TokStart, "invalid character in input");
-  case EOF: return AsmToken(AsmToken::Eof, StringRef(TokStart, 0));
+  case EOF:
+    IsAtStartOfLine = true;
+    IsAtStartOfStatement = true;
+    return AsmToken(AsmToken::Eof, StringRef(TokStart, 0));
   case 0:
   case ' ':
   case '\t':
-    if (SkipSpace) {
-      // Ignore whitespace.
-      return LexToken();
-    } else {
-      int len = 1;
-      while (*CurPtr==' ' || *CurPtr=='\t') {
-        CurPtr++;
-        len++;
-      }
-      return AsmToken(AsmToken::Space, StringRef(TokStart, len));
-    }
-  case '\n': // FALL THROUGH.
+    IsAtStartOfStatement = OldIsAtStartOfStatement;
+    while (*CurPtr == ' ' || *CurPtr == '\t')
+      CurPtr++;
+    if (SkipSpace)
+      return LexToken(); // Ignore whitespace.
+    else
+      return AsmToken(AsmToken::Space, StringRef(TokStart, CurPtr - TokStart));
+  case '\n':
   case '\r':
-    isAtStartOfLine = true;
+    IsAtStartOfLine = true;
+    IsAtStartOfStatement = true;
     return AsmToken(AsmToken::EndOfStatement, StringRef(TokStart, 1));
   case ':': return AsmToken(AsmToken::Colon, StringRef(TokStart, 1));
   case '+': return AsmToken(AsmToken::Plus, StringRef(TokStart, 1));
@@ -586,7 +601,9 @@ AsmToken AsmLexer::LexToken() {
     }
     return AsmToken(AsmToken::Exclaim, StringRef(TokStart, 1));
   case '%': return AsmToken(AsmToken::Percent, StringRef(TokStart, 1));
-  case '/': return LexSlash();
+  case '/':
+    IsAtStartOfStatement = OldIsAtStartOfStatement;
+    return LexSlash();
   case '#': return AsmToken(AsmToken::Hash, StringRef(TokStart, 1));
   case '\'': return LexSingleQuote();
   case '"': return LexQuote();
diff --git a/llvm/lib/MC/MCParser/AsmParser.cpp b/llvm/lib/MC/MCParser/AsmParser.cpp
index 5b8c4ee5df1..9a331ec4458 100644
--- a/llvm/lib/MC/MCParser/AsmParser.cpp
+++ b/llvm/lib/MC/MCParser/AsmParser.cpp
@@ -257,7 +257,6 @@ private:
   bool parseStatement(ParseStatementInfo &Info,
                       MCAsmParserSemaCallback *SI);
   bool parseCurlyBlockScope(SmallVectorImpl<AsmRewrite>& AsmStrRewrites);
-  void eatToEndOfLine();
   bool parseCppHashLineFilenameComment(SMLoc L);
 
   void checkForBadMacro(SMLoc DirectiveLoc, StringRef Name, StringRef Body,
@@ -628,6 +627,10 @@ const AsmToken &AsmParser::Lex() {
     Error(Lexer.getErrLoc(), Lexer.getErr());
 
   const AsmToken *tok = &Lexer.Lex();
+  // Drop comments here.
+  while (tok->is(AsmToken::Comment)) {
+    tok = &Lexer.Lex();
+  }
 
   if (tok->is(AsmToken::Eof)) {
     // If this is the end of an included file, pop the parent file off the
@@ -635,7 +638,7 @@ const AsmToken &AsmParser::Lex() {
     SMLoc ParentIncludeLoc = SrcMgr.getParentIncludeLoc(CurBuffer);
     if (ParentIncludeLoc != SMLoc()) {
       jumpToLoc(ParentIncludeLoc);
-      tok = &Lexer.Lex();
+      return Lex();
     }
   }
 
@@ -720,8 +723,8 @@ bool AsmParser::Run(bool NoInitialTextSection, bool NoFinalize) {
           // first referenced for a source location. We need to add something
           // to track that. Currently, we just point to the end of the file.
           HadError |=
-              Error(getLexer().getLoc(), "assembler local symbol '" +
-                                             Sym->getName() + "' not defined");
+              Error(getTok().getLoc(), "assembler local symbol '" +
+                                           Sym->getName() + "' not defined");
       }
     }
 
@@ -766,7 +769,7 @@ StringRef AsmParser::parseStringToEndOfStatement() {
   const char *Start = getTok().getLoc().getPointer();
 
   while (Lexer.isNot(AsmToken::EndOfStatement) && Lexer.isNot(AsmToken::Eof))
-    Lex();
+    Lexer.Lex();
 
   const char *End = getTok().getLoc().getPointer();
   return StringRef(Start, End - Start);
@@ -777,7 +780,7 @@ StringRef AsmParser::parseStringToComma() {
 
   while (Lexer.isNot(AsmToken::EndOfStatement) &&
          Lexer.isNot(AsmToken::Comma) && Lexer.isNot(AsmToken::Eof))
-    Lex();
+    Lexer.Lex();
 
   const char *End = getTok().getLoc().getPointer();
   return StringRef(Start, End - Start);
@@ -859,7 +862,7 @@ bool AsmParser::parsePrimaryExpr(const MCExpr *&Res, SMLoc &EndLoc) {
     if (!MAI.useParensForSymbolVariant()) {
       if (FirstTokenKind == AsmToken::String) {
         if (Lexer.is(AsmToken::At)) {
-          Lexer.Lex(); // eat @
+          Lex(); // eat @
           SMLoc AtLoc = getLexer().getLoc();
           StringRef VName;
           if (parseIdentifier(VName))
@@ -871,14 +874,14 @@ bool AsmParser::parsePrimaryExpr(const MCExpr *&Res, SMLoc &EndLoc) {
         Split = Identifier.split('@');
       }
     } else if (Lexer.is(AsmToken::LParen)) {
-      Lexer.Lex(); // eat (
+      Lex(); // eat '('.
       StringRef VName;
       parseIdentifier(VName);
       if (Lexer.isNot(AsmToken::RParen)) {
           return Error(Lexer.getTok().getLoc(),
                        "unexpected token in variant, expected ')'");
       }
-      Lexer.Lex(); // eat )
+      Lex(); // eat ')'.
       Split = std::make_pair(Identifier, VName);
     }
 
@@ -1343,21 +1346,24 @@ bool AsmParser::parseBinOpRHS(unsigned Precedence, const MCExpr *&Res,
 ///   ::= Label* Identifier OperandList* EndOfStatement
 bool AsmParser::parseStatement(ParseStatementInfo &Info,
                                MCAsmParserSemaCallback *SI) {
+  // Eat initial spaces and comments
+  while (Lexer.is(AsmToken::Space))
+    Lex();
   if (Lexer.is(AsmToken::EndOfStatement)) {
-    Out.AddBlankLine();
+    // if this is a line comment we can drop it safely
+    if (getTok().getString().front() == '\r' ||
+        getTok().getString().front() == '\n')
+      Out.AddBlankLine();
     Lex();
     return false;
   }
-
-  // Statements always start with an identifier or are a full line comment.
+  // Statements always start with an identifier.
   AsmToken ID = getTok();
   SMLoc IDLoc = ID.getLoc();
   StringRef IDVal;
   int64_t LocalLabelVal = -1;
-  // A full line comment is a '#' as the first token.
-  if (Lexer.is(AsmToken::Hash))
+  if (Lexer.is(AsmToken::HashDirective))
     return parseCppHashLineFilenameComment(IDLoc);
-
   // Allow an integer followed by a ':' as a directional local label.
   if (Lexer.is(AsmToken::Integer)) {
     LocalLabelVal = getTok().getIntVal();
@@ -1648,7 +1654,8 @@ bool AsmParser::parseStatement(ParseStatementInfo &Info,
       return parseDirectiveIncbin();
     case DK_CODE16:
     case DK_CODE16GCC:
-      return TokError(Twine(IDVal) + " not supported yet");
+      return TokError(Twine(IDVal) +
+                      " not currently supported for this target");
     case DK_REPT:
       return parseDirectiveRept(IDLoc, IDVal);
     case DK_IRP:
@@ -1868,37 +1875,20 @@ AsmParser::parseCurlyBlockScope(SmallVectorImpl<AsmRewrite> &AsmStrRewrites) {
   return true;
 }
 
-/// eatToEndOfLine uses the Lexer to eat the characters to the end of the line
-/// since they may not be able to be tokenized to get to the end of line token.
-void AsmParser::eatToEndOfLine() {
-  if (!Lexer.is(AsmToken::EndOfStatement))
-    Lexer.LexUntilEndOfLine();
-  // Eat EOL.
-  Lex();
-}
-
 /// parseCppHashLineFilenameComment as this:
 ///   ::= # number "filename"
-/// or just as a full line comment if it doesn't have a number and a string.
 bool AsmParser::parseCppHashLineFilenameComment(SMLoc L) {
   Lex(); // Eat the hash token.
-
-  if (getLexer().isNot(AsmToken::Integer)) {
-    // Consume the line since in cases it is not a well-formed line directive,
-    // as if were simply a full line comment.
-    eatToEndOfLine();
-    return false;
-  }
-
+  // Lexer only ever emits HashDirective if it fully formed if it's
+  // done the checking already so this is an internal error.
+  assert(getTok().is(AsmToken::Integer) &&
+         "Lexing Cpp line comment: Expected Integer");
   int64_t LineNumber = getTok().getIntVal();
   Lex();
-
-  if (getLexer().isNot(AsmToken::String)) {
-    eatToEndOfLine();
-    return false;
-  }
-
+  assert(getTok().is(AsmToken::String) &&
+         "Lexing Cpp line comment: Expected String");
   StringRef Filename = getTok().getString();
+  Lex();
   // Get rid of the enclosing quotes.
   Filename = Filename.substr(1, Filename.size() - 2);
 
@@ -1907,9 +1897,6 @@ bool AsmParser::parseCppHashLineFilenameComment(SMLoc L) {
   CppHashInfo.Filename = Filename;
   CppHashInfo.LineNumber = LineNumber;
   CppHashInfo.Buf = CurBuffer;
-
-  // Ignore any trailing characters, they're just comment.
-  eatToEndOfLine();
   return false;
 }
 
@@ -2268,7 +2255,7 @@ bool AsmParser::parseMacroArguments(const MCAsmMacro *M,
           break;
 
       if (FAI >= NParameters) {
-    assert(M && "expected macro to be defined");
+        assert(M && "expected macro to be defined");
         Error(IDLoc,
               "parameter named '" + FA.Name + "' does not exist for macro '" +
               M->Name + "'");
@@ -2426,7 +2413,7 @@ bool AsmParser::parseIdentifier(StringRef &Res) {
     // Construct the joined identifier and consume the token.
     Res =
         StringRef(PrefixLoc.getPointer(), getTok().getIdentifier().size() + 1);
-    Lexer.Lex(); // Lexer's Lex guarantees consecutive token
+    Lex(); // Parser Lex to maintain invariants.
     return false;
   }
 
@@ -2568,16 +2555,16 @@ bool AsmParser::parseDirectiveReloc(SMLoc DirectiveLoc) {
 
   if (Lexer.isNot(AsmToken::Comma))
     return TokError("expected comma");
-  Lexer.Lex();
+  Lex();
 
   if (Lexer.isNot(AsmToken::Identifier))
     return TokError("expected relocation name");
   SMLoc NameLoc = Lexer.getTok().getLoc();
   StringRef Name = Lexer.getTok().getIdentifier();
-  Lexer.Lex();
+  Lex();
 
   if (Lexer.is(AsmToken::Comma)) {
-    Lexer.Lex();
+    Lex();
     SMLoc ExprLoc = Lexer.getLoc();
     if (parseExpression(Expr))
       return true;
@@ -5250,10 +5237,9 @@ static bool isSymbolUsedInExpression(const MCSymbol *Sym, const MCExpr *Value) {
 bool parseAssignmentExpression(StringRef Name, bool allow_redef,
                                MCAsmParser &Parser, MCSymbol *&Sym,
                                const MCExpr *&Value) {
-  MCAsmLexer &Lexer = Parser.getLexer();
 
   // FIXME: Use better location, we should use proper tokens.
-  SMLoc EqualLoc = Lexer.getLoc();
+  SMLoc EqualLoc = Parser.getTok().getLoc();
 
   if (Parser.parseExpression(Value)) {
     Parser.TokError("missing expression");
@@ -5265,7 +5251,7 @@ bool parseAssignmentExpression(StringRef Name, bool allow_redef,
   // a = b
   // b = c
 
-  if (Lexer.isNot(AsmToken::EndOfStatement))
+  if (Parser.getTok().isNot(AsmToken::EndOfStatement))
     return Parser.TokError("unexpected token in assignment");
 
   // Eat the end of statement marker.