Refactor and cleanup Assembly Parsing / Lexing

[niravd]

Summary

Add explicit Comment Token in Assembly Lexing for future support for
outputting explicit comments from inline assembly. As part of this,
CPPHash Directives are now explicitly distinguished from Hash line
comments in Lexer.

Line comments are recorded as EndOfStatement tokens, not Comment tokens
to simplify compatibility with current TargetParsers. This slightly
complicates comment output.

This remove all lexing tasks out of the parser, does minor cleanup
to remove extraneous newlines Asm Output, and some improvements white
space handling.

[niravd]

Modifications to keep line comments in correct place with TargetParser

[niravd]

Update to fix CPP Hash Directive parsing and add testcase

[rnk]

Overall seems like a good cleanup of a janky parser. =/

include/llvm/MC/MCParser/AsmLexer.h

32–34

These should both follow LLVM style, i.e. leading upper case for variables & fields.

lib/MC/MCParser/AsmLexer.cpp

53

Removing this attempt to optimize seems fine.

500

Follow the LLVM naming conventions please.

lib/MC/MCParser/AsmParser.cpp

630

Really it skips them, it doesn't parse them.

[niravd]

Revert enum as requested in D20020, fix LLVM style issues

[niravd]

Everything appears to be resolved. Can I get a LGTM?

[rnk]

lgtm

[loladiro]

Please note https://llvm.org/bugs/show_bug.cgi?id=28921, which has a testcase bisected to this revision.

include/llvm/MC/MCParser/AsmLexer.h

 class MCAsmInfo;
 
 /// AsmLexer - Lexer class for assembly files.
 class AsmLexer : public MCAsmLexer {
   const MCAsmInfo &MAI;
 
   const char *CurPtr;
   StringRef CurBuf;
-  bool isAtStartOfLine;
+  bool IsAtStartOfLine;
+  bool IsAtStartOfStatement;
 
   void operator=(const AsmLexer&) = delete;
   AsmLexer(const AsmLexer&) = delete;
 
 protected:
   /// LexToken - Read the next token and return its code.
   AsmToken LexToken() override;
 
 public:
   AsmLexer(const MCAsmInfo &MAI);
   ~AsmLexer() override;
 
   void setBuffer(StringRef Buf, const char *ptr = nullptr);
 
   StringRef LexUntilEndOfStatement() override;
-  StringRef LexUntilEndOfLine();
 
   size_t peekTokens(MutableArrayRef<AsmToken> Buf,
                     bool ShouldSkipSpace = true) override;
 
-  bool isAtStartOfComment(const char *Ptr);
-  bool isAtStatementSeparator(const char *Ptr);
-
   const MCAsmInfo &getMAI() const { return MAI; }
 
 private:
+  bool isAtStartOfComment(const char *Ptr);
+  bool isAtStatementSeparator(const char *Ptr);
   int getNextChar();
   AsmToken ReturnError(const char *Loc, const std::string &Msg);
 
   AsmToken LexIdentifier();
   AsmToken LexSlash();
   AsmToken LexLineComment();
   AsmToken LexDigit();
   AsmToken LexSingleQuote();
   AsmToken LexQuote();
   AsmToken LexFloatLiteral();
   AsmToken LexHexFloatLiteral(bool NoIntDigits);
+
+  StringRef LexUntilEndOfLine();
 };
 
 } // end namespace llvm
 
 #endif

include/llvm/MC/MCParser/MCAsmLexer.h

enum TokenKind {
 
     // Integer values.
     Integer,
     BigNum, // larger than 64 bits
 
     // Real values.
     Real,
 
+    // Comments
+    Comment,
+    HashDirective,
     // No-value.
     EndOfStatement,
     Colon,
     Space,
     Plus, Minus, Tilde,
-    Slash,    // '/'
+    Slash,     // '/'
     BackSlash, // '\'
     LParen, RParen, LBrac, RBrac, LCurly, RCurly,
     Star, Dot, Comma, Dollar, Equal, EqualEqual,
 
     Pipe, PipePipe, Caret,
     Amp, AmpAmp, Exclaim, ExclaimEqual, Percent, Hash,
     Less, LessEqual, LessLess, LessGreater,
     Greater, GreaterEqual, GreaterGreater, At
public:
 
   /// Consume the next token from the input stream and return it.
   ///
   /// The lexer will continuosly return the end-of-file token once the end of
   /// the main input file has been reached.
   const AsmToken &Lex() {
     assert(!CurTok.empty());
     CurTok.erase(CurTok.begin());
+    // Always place in front as LexToken may generate multiple tokens via UnLex.
     if (CurTok.empty())
-      CurTok.emplace_back(LexToken());
+      CurTok.insert(CurTok.begin(), LexToken());
     return CurTok.front();
   }
 
   void UnLex(AsmToken const &Token) {
     CurTok.insert(CurTok.begin(), Token);
   }
 
   virtual StringRef LexUntilEndOfStatement() = 0;

lib/MC/MCParser/AsmLexer.cpp

 #include <cctype>
 #include <cerrno>
 #include <cstdio>
 #include <cstdlib>
 using namespace llvm;
 
 AsmLexer::AsmLexer(const MCAsmInfo &MAI) : MAI(MAI) {
   CurPtr = nullptr;
-  isAtStartOfLine = true;
+  IsAtStartOfLine = true;
+  IsAtStartOfStatement = true;
   AllowAtInIdentifier = !StringRef(MAI.getCommentString()).startswith("@");
 }
 
 AsmLexer::~AsmLexer() {
 }
 
 void AsmLexer::setBuffer(StringRef Buf, const char *ptr) {
   CurBuf = Buf;

 /// location.  This is defined to always return AsmToken::Error.
 AsmToken AsmLexer::ReturnError(const char *Loc, const std::string &Msg) {
   SetError(SMLoc::getFromPointer(Loc), Msg);
 
   return AsmToken(AsmToken::Error, StringRef(Loc, CurPtr - Loc));
 }
 
 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]*)?
 ///
 /// The leading integral digit sequence and dot should have already been
 /// consumed, some or all of the fractional digit sequence *can* have been
 /// consumed.
 AsmToken AsmLexer::LexFloatLiteral() {
case '/':
     ++CurPtr;
     return LexLineComment();
   default:
     return AsmToken(AsmToken::Slash, StringRef(CurPtr - 1, 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));
+  // Whis is a whole line comment. leave newline
+  if (IsAtStartOfStatement)
+    return AsmToken(AsmToken::EndOfStatement,
+                    StringRef(TokStart, CurPtr - TokStart));
+  IsAtStartOfLine = true;
+  IsAtStartOfStatement = true;
+
+  return AsmToken(AsmToken::EndOfStatement,
+                  StringRef(TokStart, CurPtr - 1 - TokStart));
 }
 
 static void SkipIgnoredIntegerSuffix(const char *&CurPtr) {
   // Skip ULL, UL, U, L and LL suffices.
   if (CurPtr[0] == 'U')
     ++CurPtr;
   if (CurPtr[0] == 'L')
     ++CurPtr;
AsmToken AsmLexer::LexQuote() {
   return AsmToken(AsmToken::String, StringRef(TokStart, CurPtr - TokStart));
 }
 
 StringRef AsmLexer::LexUntilEndOfStatement() {
   TokStart = CurPtr;
 
   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);
 }
 
 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);
 }
 
 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();
   SMLoc SavedErrLoc = getErrLoc();
 
   SkipSpace = ShouldSkipSpace;
 
   size_t ReadCount;
   for (ReadCount = 0; ReadCount < Buf.size(); ++ReadCount) {
     AsmToken Token = LexToken();
 
     Buf[ReadCount] = Token;
 
     if (Token.is(AsmToken::Eof))
       break;
   }
 
   SetError(SavedErrLoc, SavedErr);
 
   SkipSpace = SavedSkipSpace;
-  isAtStartOfLine = SavedAtStartOfLine;
+  IsAtStartOfLine = SavedAtStartOfLine;
+  IsAtStartOfStatement = SavedAtStartOfStatement;
   CurPtr = SavedCurPtr;
   TokStart = SavedTokStart;
 
   return ReadCount;
 }
 
 bool AsmLexer::isAtStartOfComment(const char *Ptr) {
   const char *CommentString = MAI.getCommentString();
return strncmp(Ptr, MAI.getSeparatorString(),
                  strlen(MAI.getSeparatorString())) == 0;
 }
 
 AsmToken AsmLexer::LexToken() {
   TokStart = CurPtr;
   // 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_$.@]*
     if (isalpha(CurChar) || CurChar == '_' || CurChar == '.')
       return LexIdentifier();
 
     // 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')
+    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));
   case '-': return AsmToken(AsmToken::Minus, StringRef(TokStart, 1));
   case '~': return AsmToken(AsmToken::Tilde, StringRef(TokStart, 1));
   case '(': return AsmToken(AsmToken::LParen, StringRef(TokStart, 1));
   case ')': return AsmToken(AsmToken::RParen, StringRef(TokStart, 1));
   case '[': return AsmToken(AsmToken::LBrac, StringRef(TokStart, 1));

lib/MC/MCParser/AsmParser.cpp

public:
   void checkForValidSection() override;
   /// }
 
 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,
                         ArrayRef<MCAsmMacroParameter> Parameters);
   bool expandMacro(raw_svector_ostream &OS, StringRef Body,
                    ArrayRef<MCAsmMacroParameter> Parameters,
                    ArrayRef<MCAsmMacroArgument> A, bool EnableAtPseudoVariable,
                    SMLoc L);
Lexer.setBuffer(SrcMgr.getMemoryBuffer(CurBuffer)->getBuffer(),
                   Loc.getPointer());
 }
 
 const AsmToken &AsmParser::Lex() {
   if (Lexer.getTok().is(AsmToken::Error))
     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
     // include stack.
     SMLoc ParentIncludeLoc = SrcMgr.getParentIncludeLoc(CurBuffer);
     if (ParentIncludeLoc != SMLoc()) {
       jumpToLoc(ParentIncludeLoc);
-      tok = &Lexer.Lex();
+      return Lex();
     }
   }
 
 
   return *tok;
 }
 
 bool AsmParser::Run(bool NoInitialTextSection, bool NoFinalize) {
if (MAI.hasSubsectionsViaSymbols()) {
         // Variable symbols may not be marked as defined, so check those
         // explicitly. If we know it's a variable, we have a definition for
         // the purposes of this check.
         if (Sym->isTemporary() && !Sym->isVariable() && !Sym->isDefined())
           // FIXME: We would really like to refer back to where the symbol was
           // 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 '" +
       }
     }
 
     // Temporary symbols like the ones for directional jumps don't go in the
     // symbol table. They also need to be diagnosed in all (final) cases.
     for (std::tuple<SMLoc, CppHashInfoTy, MCSymbol *> &LocSym : DirLabels) {
       if (std::get<2>(LocSym)->isUndefined()) {
         // Reset the state of any "# line file" directives we've seen to the
void AsmParser::eatToEndOfStatement() {
   if (Lexer.is(AsmToken::EndOfStatement))
     Lexer.Lex();
 }
 
 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);
 }
 
 StringRef AsmParser::parseStringToComma() {
   const char *Start = getTok().getLoc().getPointer();
 
   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);
 }
 
 /// \brief Parse a paren expression and return it.
 /// NOTE: This assumes the leading '(' has already been consumed.
 ///
if (parseIdentifier(Identifier)) {
         return Error(FirstTokenLoc, "invalid token in expression");
       }
     }
     // Parse symbol variant
     std::pair<StringRef, StringRef> Split;
     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))
             return Error(AtLoc, "expected symbol variant after '@'");
 
           Split = std::make_pair(Identifier, VName);
         }
       } else {
         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);
     }
 
     EndLoc = SMLoc::getFromPointer(Identifier.end());
 
     // This is a symbol reference.
     StringRef SymbolName = Identifier;
     MCSymbolRefExpr::VariantKind Variant = MCSymbolRefExpr::VK_None;

 }
 
 /// ParseStatement:
 ///   ::= EndOfStatement
 ///   ::= Label* Directive ...Operands... EndOfStatement
 ///   ::= 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)) {
+    // 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();
     if (LocalLabelVal < 0) {
       if (!TheCondState.Ignore)
         return TokError("unexpected token at start of statement");
       IDVal = "";
     } else {
if (IDVal[0] == '.' && IDVal != ".") {
     case DK_ABORT:
       return parseDirectiveAbort();
     case DK_INCLUDE:
       return parseDirectiveInclude();
     case DK_INCBIN:
       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:
       return parseDirectiveIrp(IDLoc);
     case DK_IRPC:
       return parseDirectiveIrpc(IDLoc);
     case DK_ENDR:
       return parseDirectiveEndr(IDLoc);
if (Lexer.is(AsmToken::EndOfStatement))
     Lex(); // Eat EndOfStatement following the brace
 
   // Erase the block start/end brace from the output asm string
   AsmStrRewrites.emplace_back(AOK_Skip, StartLoc, Lexer.getLoc().getPointer() -
                                                   StartLoc.getPointer());
   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);
 
   // Save the SMLoc, Filename and LineNumber for later use by diagnostics.
   CppHashInfo.Loc = L;
   CppHashInfo.Filename = Filename;
   CppHashInfo.LineNumber = LineNumber;
   CppHashInfo.Buf = CurBuffer;
-
-  // Ignore any trailing characters, they're just comment.
-  eatToEndOfLine();
   return false;
 }
 
 /// \brief will use the last parsed cpp hash line filename comment
 /// for the Filename and LineNo if any in the diagnostic.
 void AsmParser::DiagHandler(const SMDiagnostic &Diag, void *Context) {
   const AsmParser *Parser = static_cast<const AsmParser *>(Context);
   raw_ostream &OS = errs();
for (unsigned Parameter = 0; !NParameters || Parameter < NParameters;
     unsigned PI = Parameter;
     if (!FA.Name.empty()) {
       unsigned FAI = 0;
       for (FAI = 0; FAI < NParameters; ++FAI)
         if (M->Parameters[FAI].Name == FA.Name)
           break;
 
       if (FAI >= NParameters) {
         assert(M && "expected macro to be defined");
         Error(IDLoc,
               "parameter named '" + FA.Name + "' does not exist for macro '" +
               M->Name + "'");
         return true;
       }
       PI = FAI;
     }
 
bool AsmParser::parseDirectiveReloc(SMLoc DirectiveLoc) {
   if (!Offset->evaluateAsAbsolute(OffsetValue))
     return Error(OffsetLoc, "expression is not a constant value");
 
   if (OffsetValue < 0)
     return Error(OffsetLoc, "expression is negative");
 
   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;
 
     MCValue Value;
     if (!Expr->evaluateAsRelocatable(Value, nullptr, nullptr))
       return Error(ExprLoc, "expression must be relocatable");
   }
static bool isSymbolUsedInExpression(const MCSymbol *Sym, const MCExpr *Value) {
   }
 
   llvm_unreachable("Unknown expr kind!");
 }
 
 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");
     Parser.eatToEndOfStatement();
     return true;
   }
 
   // Note: we don't count b as used in "a = b". This is to allow
   // 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.
   Parser.Lex();
 
   // Validate that the LHS is allowed to be a variable (either it has not been
   // used as a symbol, or it is an absolute symbol).
   Sym = Parser.getContext().lookupSymbol(Name);

lib/Target/ARM/AsmParser/ARMAsmParser.cpp

bool ARMAsmParser::parseDirectiveObjectArch(SMLoc L) {
   if (getLexer().isNot(AsmToken::Identifier)) {
     Error(getLexer().getLoc(), "unexpected token");
     Parser.eatToEndOfStatement();
     return false;
   }
 
   StringRef Arch = Parser.getTok().getString();
   SMLoc ArchLoc = Parser.getTok().getLoc();
-  getLexer().Lex();
+  Lex();
 
   unsigned ID = ARM::parseArch(Arch);
 
   if (ID == ARM::AK_INVALID) {
     Error(ArchLoc, "unknown architecture '" + Arch + "'");
     Parser.eatToEndOfStatement();
     return false;
   }
bool ARMAsmParser::parseDirectiveArchExtension(SMLoc L) {
   if (getLexer().isNot(AsmToken::Identifier)) {
     Error(getLexer().getLoc(), "unexpected token");
     Parser.eatToEndOfStatement();
     return false;
   }
 
   StringRef Name = Parser.getTok().getString();
   SMLoc ExtLoc = Parser.getTok().getLoc();
-  getLexer().Lex();
+  Lex();
 
   bool EnableFeature = true;
   if (Name.startswith_lower("no")) {
     EnableFeature = false;
     Name = Name.substr(2);
   }
   unsigned FeatureKind = ARM::parseArchExt(Name);
   if (FeatureKind == ARM::AEK_INVALID)

lib/Target/Hexagon/AsmParser/HexagonAsmParser.cpp

if (CheckOk) {
     }
   }
 
   return false; // No error
 }
 
 bool HexagonAsmParser::matchBundleOptions() {
   MCAsmParser &Parser = getParser();
-  MCAsmLexer &Lexer = getLexer();
   while (true) {
     if (!Parser.getTok().is(AsmToken::Colon))
       return false;
-    Lexer.Lex();
+    Lex();
     StringRef Option = Parser.getTok().getString();
     if (Option.compare_lower("endloop0") == 0)
       HexagonMCInstrInfo::setInnerLoop(MCB);
     else if (Option.compare_lower("endloop1") == 0)
       HexagonMCInstrInfo::setOuterLoop(MCB);
     else if (Option.compare_lower("mem_noshuf") == 0)
       HexagonMCInstrInfo::setMemReorderDisabled(MCB);
     else if (Option.compare_lower("mem_shuf") == 0)
       HexagonMCInstrInfo::setMemStoreReorderEnabled(MCB);
     else
       return true;
-    Lexer.Lex();
+    Lex();
   }
 }
 
 // For instruction aliases, immediates are generated rather than
 // MCConstantExpr.  Convert them for uniform MCExpr.
 // Also check for signed/unsigned mismatches and warn
 void HexagonAsmParser::canonicalizeImmediates(MCInst &MCI) {
   MCInst NewInst;
return previousEqual(Operands, Index, "loop0") ||
          previousEqual(Operands, Index, "sp3loop0");
 }
 }
 
 bool HexagonAsmParser::splitIdentifier(OperandVector &Operands) {
   AsmToken const &Token = getParser().getTok();
   StringRef String = Token.getString();
   SMLoc Loc = Token.getLoc();
-  getLexer().Lex();
+  Lex();
   do {
     std::pair<StringRef, StringRef> HeadTail = String.split('.');
     if (!HeadTail.first.empty())
       Operands.push_back(HexagonOperand::CreateToken(HeadTail.first, Loc));
     if (!HeadTail.second.empty())
       Operands.push_back(HexagonOperand::CreateToken(
           String.substr(HeadTail.first.size(), 1), Loc));
     String = HeadTail.second;

 
 bool HexagonAsmParser::parseExpression(MCExpr const *& Expr) {
   llvm::SmallVector<AsmToken, 4> Tokens;
   MCAsmLexer &Lexer = getLexer();
   bool Done = false;
   static char const * Comma = ",";
   do {
     Tokens.emplace_back (Lexer.getTok());
-    Lexer.Lex();
+    Lex();
     switch (Tokens.back().getKind())
     {
     case AsmToken::TokenKind::Hash:
       if (Tokens.size () > 1)
         if ((Tokens.end () - 2)->getKind() == AsmToken::TokenKind::Plus) {
           Tokens.insert(Tokens.end() - 2,
                         AsmToken(AsmToken::TokenKind::Comma, Comma));
           Done = true;

 /// Parse an instruction.
 bool HexagonAsmParser::parseInstruction(OperandVector &Operands) {
   MCAsmParser &Parser = getParser();
   MCAsmLexer &Lexer = getLexer();
   while (true) {
     AsmToken const &Token = Parser.getTok();
     switch (Token.getKind()) {
     case AsmToken::EndOfStatement: {
-      Lexer.Lex();
+      Lex();
       return false;
     }
     case AsmToken::LCurly: {
       if (!Operands.empty())
         return true;
       Operands.push_back(
           HexagonOperand::CreateToken(Token.getString(), Token.getLoc()));
-      Lexer.Lex();
+      Lex();
       return false;
     }
     case AsmToken::RCurly: {
       if (Operands.empty()) {
         Operands.push_back(
             HexagonOperand::CreateToken(Token.getString(), Token.getLoc()));
-        Lexer.Lex();
+        Lex();
       }
       return false;
     }
     case AsmToken::Comma: {
-      Lexer.Lex();
+      Lex();
       continue;
     }
     case AsmToken::EqualEqual:
     case AsmToken::ExclaimEqual:
     case AsmToken::GreaterEqual:
     case AsmToken::GreaterGreater:
     case AsmToken::LessEqual:
     case AsmToken::LessLess: {
       Operands.push_back(HexagonOperand::CreateToken(
           Token.getString().substr(0, 1), Token.getLoc()));
       Operands.push_back(HexagonOperand::CreateToken(
           Token.getString().substr(1, 1), Token.getLoc()));
-      Lexer.Lex();
+      Lex();
       continue;
     }
     case AsmToken::Hash: {
       bool MustNotExtend = false;
       bool ImplicitExpression = implicitExpressionLocation(Operands);
       SMLoc ExprLoc = Lexer.getLoc();
       if (!ImplicitExpression)
         Operands.push_back(
           HexagonOperand::CreateToken(Token.getString(), Token.getLoc()));
-      Lexer.Lex();
+      Lex();
       bool MustExtend = false;
       bool HiOnly = false;
       bool LoOnly = false;
       if (Lexer.is(AsmToken::Hash)) {
-        Lexer.Lex();
+        Lex();
         MustExtend = true;
       } else if (ImplicitExpression)
         MustNotExtend = true;
       AsmToken const &Token = Parser.getTok();
       if (Token.is(AsmToken::Identifier)) {
         StringRef String = Token.getString();
         if (String.lower() == "hi") {
           HiOnly = true;
         } else if (String.lower() == "lo") {
           LoOnly = true;
         }
         if (HiOnly || LoOnly) {
           AsmToken LParen = Lexer.peekTok();
           if (!LParen.is(AsmToken::LParen)) {
             HiOnly = false;
             LoOnly = false;
           } else {
-            Lexer.Lex();
+            Lex();
           }
         }
       }
       MCExpr const *Expr = nullptr;
       if (parseExpression(Expr))
         return true;
       int64_t Value;
       MCContext &Context = Parser.getContext();

lib/Target/Lanai/AsmParser/LanaiAsmParser.cpp

if (Lexer.is(AsmToken::EndOfStatement) && Name.startswith("bt") &&
       Operands.size() == 3) {
     Operands.erase(Operands.begin(), Operands.begin() + 2);
     Operands.insert(Operands.begin(), LanaiOperand::CreateToken("bt", NameLoc));
   }
 
   // Parse until end of statement, consuming commas between operands
   while (Lexer.isNot(AsmToken::EndOfStatement) && Lexer.is(AsmToken::Comma)) {
     // Consume comma token
-    Lexer.Lex();
+    Lex();
 
     // Parse next operand
     if (parseOperand(&Operands, Mnemonic) != MatchOperand_Success)
       return true;
   }
 
   if (IsMemoryAssignmentError(Operands)) {
     Error(Parser.getTok().getLoc(),

lib/Target/PowerPC/AsmParser/PPCAsmParser.cpp

bool PPCAsmParser::ParseInstruction(ParseInstructionInfo &Info, StringRef Name,
 
   // Parse the first operand
   if (ParseOperand(Operands))
     return true;
 
   while (getLexer().isNot(AsmToken::EndOfStatement) &&
          getLexer().is(AsmToken::Comma)) {
     // Consume the comma token
-    getLexer().Lex();
+    Lex();
 
     // Parse the next operand
     if (ParseOperand(Operands))
       return true;
   }
 
   // We'll now deal with an unfortunate special case: the syntax for the dcbt
   // and dcbtst instructions differs for server vs. embedded cores.

test/MC/AsmParser/hash-directive.s

+# RUN: not llvm-mc -triple i386-unknown-unknown %s 2>&1 | FileCheck %s
+error
+# CHECK: hash-directive.s:[[@LINE-1]]:1: error
+# 3 "FILE1" 1 #<- This is a CPP Hash w/ comment
+error
+# CHECK: FILE1:3:1: error
+# 0 "" 2 #<- This is too
+error
+# CHECK: hash-directive.s:[[@LINE-1]]:1: error
+ # 1 "FILE2" 2 #<- This is a comment
+error
+# CHECK: hash-directive.s:[[@LINE-1]]:1: error
+nop; # 6 "FILE3" 2 #<- This is a still comment
+error
+# CHECK: hash-directive.s:[[@LINE-1]]:1: error
+nop;# 6 "FILE4" 2
+  nop;
+error
+# CHECK: FILE4:7:1: error
+# 0 "" 2
+/*comment*/# 6 "FILE5" 2 #<- This is a comment
+error
+# CHECK: hash-directive.s:[[@LINE-1]]:1: error