Refix issue 1472. The previous fix worked for the example in the bug

src/jsregexp.cc

Index: src/jsregexp.cc
===================================================================
--- src/jsregexp.cc	(revision 8315)
+++ src/jsregexp.cc	(working copy)
@@ -810,6 +810,11 @@
   inline bool ignore_case() { return ignore_case_; }
   inline bool ascii() { return ascii_; }
 
+  int current_expansion_factor() { return current_expansion_factor_; }
+  void set_current_expansion_factor(int value) {
+    current_expansion_factor_ = value;
+  }
+
   static const int kNoRegister = -1;
 
  private:
@@ -821,6 +826,7 @@
   bool ignore_case_;
   bool ascii_;
   bool reg_exp_too_big_;
+  int current_expansion_factor_;
 };
 
 
@@ -848,7 +854,8 @@
       recursion_depth_(0),
       ignore_case_(ignore_case),
       ascii_(ascii),
-      reg_exp_too_big_(false) {
+      reg_exp_too_big_(false),
+      current_expansion_factor_(1) {
   accept_ = new EndNode(EndNode::ACCEPT);
   ASSERT(next_register_ - 1 <= RegExpMacroAssembler::kMaxRegister);
 }
@@ -3730,6 +3737,43 @@
 }
 
 
+// Scoped object to keep track of how much we unroll quantifier loops in the
+// regexp graph generator.
+class RegExpExpansionLimiter {
+ public:
+  static const int kMaxExpansionFactor = 6;
+  RegExpExpansionLimiter(RegExpCompiler* compiler, int factor)
+      : compiler_(compiler),
+        saved_expansion_factor_(compiler->current_expansion_factor()),
+        ok_to_expand_(true) {

[Lasse Reichstein, 2011/06/17 07:18:49]

How about initializing this as:
  ok_to_expand_(saved_expansion_factor_ <= kMaxExpansionFactor)
and not doing anything later if it's already false.

[Erik Corry, 2011/06/17 07:31:14]

Done.

+    ASSERT(factor > 0);
+    if (factor > kMaxExpansionFactor) {

[Lasse Reichstein, 2011/06/17 07:18:49]

Add comment saying that this prevents overflow in later computation, because
kMaxExpansionFactor^2 < kMaxInt.

Well, actually it doesn't if you have a very deep nesting of {6} quantifiers,
because you don't cap the stored expansion_factor. Initializing ok_to_expand as
above, and bailing out if it's already false will prevent this.

[Erik Corry, 2011/06/17 07:31:14]

Done.

+      ok_to_expand_ = false;
+      compiler->set_current_expansion_factor(kMaxExpansionFactor + 1);
+    } else {
+      int new_factor = saved_expansion_factor_ * factor;
+      if (new_factor > kMaxExpansionFactor) {
+        ok_to_expand_ = false;
+      }

[Lasse Reichstein, 2011/06/17 07:18:49]

Remove the if and just always assing:
 ok_to_expand = (new_factor <= kMaxExpansionFactor);

[Erik Corry, 2011/06/17 07:31:14]

Done.

+      compiler->set_current_expansion_factor(new_factor);
+    }
+  }
+
+  ~RegExpExpansionLimiter() {
+    compiler_->set_current_expansion_factor(saved_expansion_factor_);
+  }
+
+  bool ok_to_expand() { return ok_to_expand_; }
+
+ private:
+  RegExpCompiler* compiler_;
+  int saved_expansion_factor_;
+  bool ok_to_expand_;
+
+  DISALLOW_IMPLICIT_CONSTRUCTORS(RegExpExpansionLimiter);
+};
+
+
 RegExpNode* RegExpQuantifier::ToNode(int min,
                                      int max,
                                      bool is_greedy,
@@ -3766,45 +3810,49 @@
   bool needs_capture_clearing = !capture_registers.is_empty();
   if (body_can_be_empty) {
     body_start_reg = compiler->AllocateRegister();
-  } else if (FLAG_regexp_optimization &&
-             !body->ContainsExpandedQuantifier() &&
-             !needs_capture_clearing) {
+  } else if (FLAG_regexp_optimization && !needs_capture_clearing) {
     // Only unroll if there are no captures and the body can't be
     // empty.
-    if (min > 0 && min <= kMaxUnrolledMinMatches) {
-      int new_max = (max == kInfinity) ? max : max - min;
-      // Recurse once to get the loop or optional matches after the fixed ones.
-      RegExpNode* answer = ToNode(
-          0, new_max, is_greedy, body, compiler, on_success, true);
-      // Unroll the forced matches from 0 to min.  This can cause chains of
-      // TextNodes (which the parser does not generate).  These should be
-      // combined if it turns out they hinder good code generation.
-      for (int i = 0; i < min; i++) {
-        answer = body->ToNode(compiler, answer);
+    {
+      RegExpExpansionLimiter limiter(
+          compiler, min + ((max - min) ? 1 : 0));

[Lasse Reichstein, 2011/06/17 07:18:49]

I just noticed the using of (max - min) as a boolean. Please add a  " != 0".

[Erik Corry, 2011/06/17 07:31:14]

max - min replaced with max != min

+      if (min > 0 && min <= kMaxUnrolledMinMatches && limiter.ok_to_expand()) {
+        int new_max = (max == kInfinity) ? max : max - min;
+        // Recurse once to get the loop or optional matches after the fixed
+        // ones.
+        RegExpNode* answer = ToNode(
+            0, new_max, is_greedy, body, compiler, on_success, true);
+        // Unroll the forced matches from 0 to min.  This can cause chains of
+        // TextNodes (which the parser does not generate).  These should be
+        // combined if it turns out they hinder good code generation.
+        for (int i = 0; i < min; i++) {
+          answer = body->ToNode(compiler, answer);
+        }
+        return answer;
       }
-      if (min > 1) body->set_contains_expanded_quantifier(true);
-      return answer;
     }
-    if (max <= kMaxUnrolledMaxMatches) {
-      ASSERT(min == 0);
-      // Unroll the optional matches up to max.
-      RegExpNode* answer = on_success;
-      for (int i = 0; i < max; i++) {
-        ChoiceNode* alternation = new ChoiceNode(2);
-        if (is_greedy) {
-          alternation->AddAlternative(GuardedAlternative(body->ToNode(compiler,
-                                                                      answer)));
-          alternation->AddAlternative(GuardedAlternative(on_success));
-        } else {
-          alternation->AddAlternative(GuardedAlternative(on_success));
-          alternation->AddAlternative(GuardedAlternative(body->ToNode(compiler,
-                                                                      answer)));
+    if (max <= kMaxUnrolledMaxMatches && min == 0) {
+      ASSERT(max > 0);  // Due to the 'if' above.
+      RegExpExpansionLimiter limiter(compiler, max);
+      if (limiter.ok_to_expand()) {
+        // Unroll the optional matches up to max.
+        RegExpNode* answer = on_success;
+        for (int i = 0; i < max; i++) {
+          ChoiceNode* alternation = new ChoiceNode(2);
+          if (is_greedy) {
+            alternation->AddAlternative(
+                GuardedAlternative(body->ToNode(compiler, answer)));
+            alternation->AddAlternative(GuardedAlternative(on_success));
+          } else {
+            alternation->AddAlternative(GuardedAlternative(on_success));
+            alternation->AddAlternative(
+                GuardedAlternative(body->ToNode(compiler, answer)));
+          }
+          answer = alternation;
+          if (not_at_start) alternation->set_not_at_start();
         }
-        answer = alternation;
-        if (not_at_start) alternation->set_not_at_start();
+        return answer;
       }
-      if (max > 1) body->set_contains_expanded_quantifier(true);
-      return answer;
     }
   }
   bool has_min = min > 0;
@@ -4130,12 +4178,6 @@
 }
 
 
-static void AddUncanonicals(Isolate* isolate,
-                            ZoneList<CharacterRange>* ranges,
-                            int bottom,
-                            int top);
-
-
 void CharacterRange::AddCaseEquivalents(ZoneList<CharacterRange>* ranges,
                                         bool is_ascii) {
   Isolate* isolate = Isolate::Current();
@@ -4296,101 +4338,6 @@
 }
 
 
-static void AddUncanonicals(Isolate* isolate,
-                            ZoneList<CharacterRange>* ranges,
-                            int bottom,
-                            int top) {
-  unibrow::uchar chars[unibrow::Ecma262UnCanonicalize::kMaxWidth];
-  // Zones with no case mappings.  There is a DEBUG-mode loop to assert that
-  // this table is correct.
-  // 0x0600 - 0x0fff
-  // 0x1100 - 0x1cff
-  // 0x2000 - 0x20ff
-  // 0x2200 - 0x23ff
-  // 0x2500 - 0x2bff
-  // 0x2e00 - 0xa5ff
-  // 0xa800 - 0xfaff
-  // 0xfc00 - 0xfeff
-  const int boundary_count = 18;
-  int boundaries[] = {
-      0x600, 0x1000, 0x1100, 0x1d00, 0x2000, 0x2100, 0x2200, 0x2400, 0x2500,
-      0x2c00, 0x2e00, 0xa600, 0xa800, 0xfb00, 0xfc00, 0xff00};
-
-  // Special ASCII rule from spec can save us some work here.
-  if (bottom == 0x80 && top == 0xffff) return;
-
-  if (top <= boundaries[0]) {
-    CharacterRange range(bottom, top);
-    range.AddCaseEquivalents(ranges, false);
-    return;
-  }
-
-  // Split up very large ranges.  This helps remove ranges where there are no
-  // case mappings.
-  for (int i = 0; i < boundary_count; i++) {
-    if (bottom < boundaries[i] && top >= boundaries[i]) {
-      AddUncanonicals(isolate, ranges, bottom, boundaries[i] - 1);
-      AddUncanonicals(isolate, ranges, boundaries[i], top);
-      return;
-    }
-  }
-
-  // If we are completely in a zone with no case mappings then we are done.
-  for (int i = 0; i < boundary_count; i += 2) {
-    if (bottom >= boundaries[i] && top < boundaries[i + 1]) {
-#ifdef DEBUG
-      for (int j = bottom; j <= top; j++) {
-        unsigned current_char = j;
-        int length = isolate->jsregexp_uncanonicalize()->get(current_char,
-                                                             '\0', chars);
-        for (int k = 0; k < length; k++) {
-          ASSERT(chars[k] == current_char);
-        }
-      }
-#endif
-      return;
-    }
-  }
-
-  // Step through the range finding equivalent characters.
-  ZoneList<unibrow::uchar> *characters = new ZoneList<unibrow::uchar>(100);
-  for (int i = bottom; i <= top; i++) {
-    int length = isolate->jsregexp_uncanonicalize()->get(i, '\0', chars);
-    for (int j = 0; j < length; j++) {
-      uc32 chr = chars[j];
-      if (chr != i && (chr < bottom || chr > top)) {
-        characters->Add(chr);
-      }
-    }
-  }
-
-  // Step through the equivalent characters finding simple ranges and
-  // adding ranges to the character class.
-  if (characters->length() > 0) {
-    int new_from = characters->at(0);
-    int new_to = new_from;
-    for (int i = 1; i < characters->length(); i++) {
-      int chr = characters->at(i);
-      if (chr == new_to + 1) {
-        new_to++;
-      } else {
-        if (new_to == new_from) {
-          ranges->Add(CharacterRange::Singleton(new_from));
-        } else {
-          ranges->Add(CharacterRange(new_from, new_to));
-        }
-        new_from = new_to = chr;
-      }
-    }
-    if (new_to == new_from) {
-      ranges->Add(CharacterRange::Singleton(new_from));
-    } else {
-      ranges->Add(CharacterRange(new_from, new_to));
-    }
-  }
-}
-
-
 ZoneList<CharacterRange>* CharacterSet::ranges() {
   if (ranges_ == NULL) {
     ranges_ = new ZoneList<CharacterRange>(2);