[regexp] back refs must not start/end in the middle of a surrogate pair

src/regexp/jsregexp.cc

 // Copyright 2012 the V8 project authors. All rights reserved.
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file.
 
 #include "src/regexp/jsregexp.h"
 
 #include "src/ast/ast.h"
 #include "src/base/platform/platform.h"
 #include "src/compilation-cache.h"
 #include "src/compiler.h"
@@ skipping 3569 matching lines @@
     return alt_gens_[i];
   }
 
  private:
   static const int kAFew = 10;
   ZoneList<AlternativeGeneration*> alt_gens_;
   AlternativeGeneration a_few_alt_gens_[kAFew];
 };
 
 
-static const uc32 kLeadSurrogateStart = 0xd800;
-static const uc32 kLeadSurrogateEnd = 0xdbff;
-static const uc32 kTrailSurrogateStart = 0xdc00;
-static const uc32 kTrailSurrogateEnd = 0xdfff;
-static const uc32 kNonBmpStart = 0x10000;
-static const uc32 kNonBmpEnd = 0x10ffff;
 static const uc32 kRangeEndMarker = 0x110000;
 
 // The '2' variant is has inclusive from and exclusive to.
 // This covers \s as defined in ECMA-262 5.1, 15.10.2.12,
 // which include WhiteSpace (7.2) or LineTerminator (7.3) values.
 static const int kSpaceRanges[] = {
     '\t',   '\r' + 1, ' ',    ' ' + 1, 0x00A0, 0x00A1, 0x1680,         0x1681,
     0x180E, 0x180F,   0x2000, 0x200B,  0x2028, 0x202A, 0x202F,         0x2030,
     0x205F, 0x2060,   0x3000, 0x3001,  0xFEFF, 0xFF00, kRangeEndMarker};
 static const int kSpaceRangeCount = arraysize(kSpaceRanges);
@@ skipping 780 matching lines @@
     trace->Flush(compiler, this);
     return;
   }
 
   LimitResult limit_result = LimitVersions(compiler, trace);
   if (limit_result == DONE) return;
   DCHECK(limit_result == CONTINUE);
 
   RecursionCheck rc(compiler);
 
+  bool surrogate_check = compiler->unicode() && !compiler->one_byte();
+  if (surrogate_check) {
+    assembler->CheckNotInSurrogatePair(trace->cp_offset(), trace->backtrack());
+  }

[erikcorry, 2016/01/26 16:25:06]

You don't need this check (and thus you don't need the surrogate_check variable
either).

[Yang, 2016/01/27 09:06:06]

Done. Thanks.

+
   DCHECK_EQ(start_reg_ + 1, end_reg_);
   if (compiler->ignore_case()) {
     assembler->CheckNotBackReferenceIgnoreCase(start_reg_, read_backward(),
                                                trace->backtrack());
   } else {
     assembler->CheckNotBackReference(start_reg_, read_backward(),
                                      trace->backtrack());
   }
   // We are going to advance backward, so we may end up at the start.
   if (read_backward()) trace->set_at_start(Trace::UNKNOWN);
+
+  if (surrogate_check) {
+    assembler->CheckNotInSurrogatePair(trace->cp_offset(), trace->backtrack());
+  }
   on_success()->Emit(compiler, trace);
 }
 
 
 // -------------------------------------------------------------------
 // Dot/dotty output
 
 
 #ifdef DEBUG
 
@@ skipping 2413 matching lines @@
 
 
 void RegExpResultsCache::Clear(FixedArray* cache) {
   for (int i = 0; i < kRegExpResultsCacheSize; i++) {
     cache->set(i, Smi::FromInt(0));
   }
 }
 
 }  // namespace internal
 }  // namespace v8