[mips] [IAS] Fix parsing of memory offset expressions with parenthesis depth >1.

Summary:
In an expression such as "(((a+b)+c)+d)", parseParenExpression() would only parse the "a+b)+c", which would result in an error later on in the parser.
This means that we can only parse one level of inner parentheses.

In order to fix this, I added a new function called parseParenExprOfDepth(), which parses a specified number of trailing parenthesis expressions
(except for the outermost parenthesis), and changed MipsAsmParser to use it in parseMemOffset instead of parseParenExpression().

Reviewers: dsanders, rafael

Reviewed By: dsanders, rafael

Subscribers: llvm-commits

Differential Revision: http://reviews.llvm.org/D9742

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@240625 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/include/llvm/MC/MCParser/MCAsmParser.h b/include/llvm/MC/MCParser/MCAsmParser.h
index 0538b9457f9ebc72d35ec9271d4cba72e8307191..ac8706d9950537e518e2988e9092654d6d5644de 100644 (file)
--- a/include/llvm/MC/MCParser/MCAsmParser.h
+++ b/include/llvm/MC/MCParser/MCAsmParser.h
@@ -197,6 +197,17 @@ public:
   /// \brief Ensure that we have a valid section set in the streamer. Otherwise,
   /// report an error and switch to .text.
   virtual void checkForValidSection() = 0;
+
+  /// \brief Parse an arbitrary expression of a specified parenthesis depth,
+  /// assuming that the initial '(' characters have already been consumed.
+  ///
+  /// \param ParenDepth - Specifies how many trailing expressions outside the
+  /// current parentheses we have to parse.
+  /// \param Res - The value of the expression. The result is undefined
+  /// on error.
+  /// \return - False on success.
+  virtual bool parseParenExprOfDepth(unsigned ParenDepth, const MCExpr *&Res,
+                                     SMLoc &EndLoc) = 0;
 };
 
 /// \brief Create an MCAsmParser instance.

diff --git a/lib/MC/MCParser/AsmParser.cpp b/lib/MC/MCParser/AsmParser.cpp
index c4f0d441f1dda1e509124fcfb491a9d81b226585..2245a466e62abbff5afa3bbde2ad6c3f0a556bbb 100644 (file)
--- a/lib/MC/MCParser/AsmParser.cpp
+++ b/lib/MC/MCParser/AsmParser.cpp
@@ -234,6 +234,8 @@ public:
   bool parseExpression(const MCExpr *&Res, SMLoc &EndLoc) override;
   bool parsePrimaryExpr(const MCExpr *&Res, SMLoc &EndLoc) override;
   bool parseParenExpression(const MCExpr *&Res, SMLoc &EndLoc) override;
+  bool parseParenExprOfDepth(unsigned ParenDepth, const MCExpr *&Res,
+                             SMLoc &EndLoc) override;
   bool parseAbsoluteExpression(int64_t &Res) override;
 
   /// \brief Parse an identifier or string (as a quoted identifier)
@@ -1066,6 +1068,27 @@ bool AsmParser::parseParenExpression(const MCExpr *&Res, SMLoc &EndLoc) {
   return parseParenExpr(Res, EndLoc) || parseBinOpRHS(1, Res, EndLoc);
 }
 
+bool AsmParser::parseParenExprOfDepth(unsigned ParenDepth, const MCExpr *&Res,
+                                      SMLoc &EndLoc) {
+  if (parseParenExpr(Res, EndLoc))
+    return true;
+
+  for (; ParenDepth > 0; --ParenDepth) {
+    if (parseBinOpRHS(1, Res, EndLoc))
+      return true;
+
+    // We don't Lex() the last RParen.
+    // This is the same behavior as parseParenExpression().
+    if (ParenDepth - 1 > 0) {
+      if (Lexer.isNot(AsmToken::RParen))
+        return TokError("expected ')' in parentheses expression");
+      EndLoc = Lexer.getTok().getEndLoc();
+      Lex();
+    }
+  }
+  return false;
+}
+
 bool AsmParser::parseAbsoluteExpression(int64_t &Res) {
   const MCExpr *Expr;
 

diff --git a/lib/Target/Mips/AsmParser/MipsAsmParser.cpp b/lib/Target/Mips/AsmParser/MipsAsmParser.cpp
index 03dbcb2c45b0ce1f4c04353f3276f732d8808a6e..e323192a0ec61282d381ffc13cf9eeaed6fe18c4 100644 (file)
--- a/lib/Target/Mips/AsmParser/MipsAsmParser.cpp
+++ b/lib/Target/Mips/AsmParser/MipsAsmParser.cpp
@@ -3105,9 +3105,12 @@ bool MipsAsmParser::parseMemOffset(const MCExpr *&Res, bool isParenExpr) {
   MCAsmParser &Parser = getParser();
   SMLoc S;
   bool Result = true;
+  unsigned NumOfLParen = 0;
 
-  while (getLexer().getKind() == AsmToken::LParen)
+  while (getLexer().getKind() == AsmToken::LParen) {
     Parser.Lex();
+    ++NumOfLParen;
+  }
 
   switch (getLexer().getKind()) {
   default:
@@ -3118,7 +3121,7 @@ bool MipsAsmParser::parseMemOffset(const MCExpr *&Res, bool isParenExpr) {
   case AsmToken::Minus:
   case AsmToken::Plus:
     if (isParenExpr)
-      Result = getParser().parseParenExpression(Res, S);
+      Result = getParser().parseParenExprOfDepth(NumOfLParen, Res, S);
     else
       Result = (getParser().parseExpression(Res));
     while (getLexer().getKind() == AsmToken::RParen)

diff --git a/test/MC/Mips/expr1.s b/test/MC/Mips/expr1.s
index 7959315a809a7b649cc53ce4196c093c3125ea6d..4af61636355febf4b6465f7ef4dc0246393bc8dd 100644 (file)
--- a/test/MC/Mips/expr1.s
+++ b/test/MC/Mips/expr1.s
@@ -16,6 +16,11 @@
 # 32R2-EL:                             #   fixup A - offset: 0, value: foo@ABS_LO, kind: fixup_Mips_LO16
 # 32R2-EL: lw   $4, %lo(foo+8)($4)     # encoding: [0x08'A',A,0x84,0x8c]
 # 32R2-EL:                             #   fixup A - offset: 0, value: foo@ABS_LO, kind: fixup_Mips_LO16
+# 32R2-EL: lw   $4, 10($4)             # encoding: [0x0a,0x00,0x84,0x8c]
+# 32R2-EL: lw   $4, 15($4)             # encoding: [0x0f,0x00,0x84,0x8c]
+# 32R2-EL: lw   $4, 21($4)             # encoding: [0x15,0x00,0x84,0x8c]
+# 32R2-EL: lw   $4, 28($4)             # encoding: [0x1c,0x00,0x84,0x8c]
+# 32R2-EL: lw   $4, 6($4)              # encoding: [0x06,0x00,0x84,0x8c]
 # 32R2-EL: .space  64
 
 # MM-32R2-EL: .text
@@ -30,6 +35,11 @@
 # MM-32R2-EL:                          #   fixup A - offset: 0, value: foo@ABS_LO, kind: fixup_MICROMIPS_LO16
 # MM-32R2-EL: lw   $4, %lo(foo+8)($4)  # encoding: [0x84'A',0xfc'A',0x08,0x00]
 # MM-32R2-EL:                          #   fixup A - offset: 0, value: foo@ABS_LO, kind: fixup_MICROMIPS_LO16
+# MM-32R2-EL: lw   $4, 10($4)          # encoding: [0x84,0xfc,0x0a,0x00]
+# MM-32R2-EL: lw   $4, 15($4)          # encoding: [0x84,0xfc,0x0f,0x00]
+# MM-32R2-EL: lw   $4, 21($4)          # encoding: [0x84,0xfc,0x15,0x00]
+# MM-32R2-EL: lw   $4, 28($4)          # encoding: [0x84,0xfc,0x1c,0x00]
+# MM-32R2-EL: lw   $4, 6($4)           # encoding: [0x84,0xfc,0x06,0x00]
 # MM-32R2-EL: .space  64
 
   .globl  foo
@@ -40,5 +50,10 @@ foo:
   lw  $4,%lo (2 * 4) + foo($4)
   lw  $4,%lo((2 * 4) + foo)($4)
   lw  $4,(((%lo ((2 * 4) + foo))))($4)
+  lw  $4, (((1+2)+3)+4)($4)
+  lw  $4, ((((1+2)+3)+4)+5)($4)
+  lw  $4, (((((1+2)+3)+4)+5)+6)($4)
+  lw  $4, ((((((1+2)+3)+4)+5)+6)+7)($4)
+  lw  $4, (%lo((1+2)+65536)+3)($4)
   .space  64
   .end  foo