From c956219ef42bc46761dcca209e07d7aee684fc9a Mon Sep 17 00:00:00 2001
From: "erik.corry@gmail.com"
 <erik.corry@gmail.com@ce2b1a6d-e550-0410-aec6-3dcde31c8c00>
Date: Mon, 26 Jan 2009 13:04:49 +0000
Subject: [PATCH] * Remember to check for end of string even where we   know
 the character class must match. Thanks to Mads and Christian for finding this
 bug Review URL: http://codereview.chromium.org/18750

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@1150 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
---
 src/jsregexp.cc                   | 9 +++++----
 src/regexp-macro-assembler-ia32.h | 7 +++----
 src/regexp-macro-assembler.h      | 6 ++++++
 test/mjsunit/regexp.js            | 4 ++++
 4 files changed, 18 insertions(+), 8 deletions(-)

diff --git a/src/jsregexp.cc b/src/jsregexp.cc
index 14cab72bcb..f1daf81c87 100644
--- a/src/jsregexp.cc
+++ b/src/jsregexp.cc
@@ -1835,6 +1835,9 @@ static void EmitCharClass(RegExpMacroAssembler* macro_assembler,
       // ASCII optimizations for us.
       macro_assembler->GoTo(on_failure);
     }
+    if (check_offset) {
+      macro_assembler->CheckPosition(cp_offset, on_failure);
+    }
     return;
   }
 
@@ -1842,10 +1845,8 @@ static void EmitCharClass(RegExpMacroAssembler* macro_assembler,
       !cc->is_negated() &&
       ranges->at(0).IsEverything(max_char)) {
     // This is a common case hit by non-anchored expressions.
-    // TODO(erikcorry): We should have a macro assembler instruction that just
-    // checks for end of string without loading the character.
     if (check_offset) {
-      macro_assembler->LoadCurrentCharacter(cp_offset, on_failure);
+      macro_assembler->CheckPosition(cp_offset, on_failure);
     }
     return;
   }
@@ -2477,7 +2478,7 @@ bool AssertionNode::Emit(RegExpCompiler* compiler, Trace* trace) {
   switch (type_) {
     case AT_END: {
       Label ok;
-      assembler->LoadCurrentCharacter(trace->cp_offset(), &ok);
+      assembler->CheckPosition(trace->cp_offset(), &ok);
       assembler->GoTo(trace->backtrack());
       assembler->Bind(&ok);
       break;
diff --git a/src/regexp-macro-assembler-ia32.h b/src/regexp-macro-assembler-ia32.h
index dd74c655f7..116b03db48 100644
--- a/src/regexp-macro-assembler-ia32.h
+++ b/src/regexp-macro-assembler-ia32.h
@@ -71,6 +71,9 @@ class RegExpMacroAssemblerIA32: public RegExpMacroAssembler {
                                               uc16 minus,
                                               uc16 mask,
                                               Label* on_not_equal);
+  // Checks whether the given offset from the current position is before
+  // the end of the string.
+  virtual void CheckPosition(int cp_offset, Label* on_outside_input);
   virtual bool CheckSpecialCharacterClass(uc16 type,
                                           int cp_offset,
                                           bool check_offset,
@@ -171,10 +174,6 @@ class RegExpMacroAssemblerIA32: public RegExpMacroAssembler {
   // This function must not trigger a garbage collection.
   static Address GrowStack(Address stack_top);
 
-  // Checks whether the given offset from the current position is before
-  // the end of the string.
-  void CheckPosition(int cp_offset, Label* on_outside_input);
-
   // The ebp-relative location of a regexp register.
   Operand register_location(int register_index);
 
diff --git a/src/regexp-macro-assembler.h b/src/regexp-macro-assembler.h
index 74133d587e..0d27fac4bd 100644
--- a/src/regexp-macro-assembler.h
+++ b/src/regexp-macro-assembler.h
@@ -111,6 +111,12 @@ class RegExpMacroAssembler {
   virtual void CheckNotRegistersEqual(int reg1,
                                       int reg2,
                                       Label* on_not_equal) = 0;
+
+  // Checks whether the given offset from the current position is before
+  // the end of the string.  May overwrite the current character.
+  virtual void CheckPosition(int cp_offset, Label* on_outside_input) {
+    LoadCurrentCharacter(cp_offset, on_outside_input, true);
+  }
   // Check whether a standard/default character class matches the current
   // character. Returns false if the type of special character class does
   // not have custom support.
diff --git a/test/mjsunit/regexp.js b/test/mjsunit/regexp.js
index 716b54be04..705754bef1 100644
--- a/test/mjsunit/regexp.js
+++ b/test/mjsunit/regexp.js
@@ -332,3 +332,7 @@ assertFalse(/()x\1(y([0-7]%%%x|[0-6]%%%y)|dkjasldkas)/.test('xy%%%y'), 'qt5');
 assertFalse(/()x\1y([0-7]%%%x|[0-6]%%%y)/.test('xy7%%%y'), 'qt6');
 assertFalse(/xy([0-7]%%%x|[0-6]%%%y)/.test('xy7%%%y'), 'qt7');
 assertFalse(/x([0-7]%%%x|[0-6]%%%y)/.test('x7%%%y'), 'qt8');
+
+
+// Don't hang on this one.
+/[^\xfe-\xff]*/.test("");