Merge up to 8597 to experimental/gc from the bleeding edge.

test/mozilla/mozilla.status

 # Copyright 2009 the V8 project authors. All rights reserved.
 # Redistribution and use in source and binary forms, with or without
 # modification, are permitted provided that the following conditions are
 # met:
 #
 #     * Redistributions of source code must retain the above copyright
 #       notice, this list of conditions and the following disclaimer.
 #     * Redistributions in binary form must reproduce the above
 #       copyright notice, this list of conditions and the following
 #       disclaimer in the documentation and/or other materials provided
@@ skipping 188 matching lines @@
 
 # Test that assumes specific runtime for a regexp, flaky in debug mode.
 ecma_3/RegExp/regress-85721: PASS || FAIL if $mode == debug
 
 
 ##################### INCOMPATIBLE TESTS #####################
 
 # This section is for tests that fail in both V8 and JSC.  Thus they
 # have been determined to be incompatible between Mozilla and V8/JSC.
 
+# Fail because it calls builtins as functions and do not expect the
+# builtin to have undefined as the receiver.
+ecma/String/15.5.4.6-2: FAIL_OK
+
 # Fail because of toLowerCase and toUpperCase conversion.
 ecma/String/15.5.4.11-2: FAIL_OK
 ecma/String/15.5.4.11-5: FAIL_OK
 ecma/String/15.5.4.12-1: FAIL_OK
 ecma/String/15.5.4.12-4: FAIL_OK
 
 # This test uses an older version of the unicode standard that fails
 # us because we correctly convert the armenian small ligature ech-yiwn
 # to the two upper-case characters ECH and YIWN, whereas the older
 # unicode version converts it to itself.
@@ skipping 99 matching lines @@
 
 
 # Stack overflows should be InternalError: too much recursion?
 js1_5/Regress/regress-234389: FAIL_OK
 
 
 # This may very well be a bogus test. I'm not sure yet.
 js1_5/Regress/regress-320119: FAIL_OK
 
 
-# We do not recognize a multiline comment as newline character.
-# We are compatible with JSC.
-ecma_3/LexicalConventions/7.4-01: FAIL_OK
-
-
 # No support for toSource().
 js1_5/Regress/regress-248444: FAIL_OK
 js1_5/Regress/regress-313967-01: FAIL_OK
 js1_5/Regress/regress-313967-02: FAIL_OK
 
 # This fails because we don't have stack space for Function.prototype.apply
 # with very large numbers of arguments.  The test uses 2^24 arguments.
 js1_5/Array/regress-350256-03: FAIL_OK
 
 
@@ skipping 82 matching lines @@
 js1_2/operator/equality: FAIL_OK
 js1_2/version120/boolean-001: FAIL_OK
 js1_2/String/concat: FAIL_OK
 js1_2/function/Function_object: FAIL_OK
 js1_2/function/tostring-1: FAIL_OK
 js1_2/version120/regress-99663: FAIL_OK
 js1_2/regexp/RegExp_lastIndex: FAIL_OK
 js1_2/regexp/string_split: FAIL_OK
 
 
+# RegExps are not callable.
+js1_2/regexp/simple_form: FAIL_OK
+js1_2/regexp/regress-6359: FAIL_OK
+js1_2/regexp/regress-9141: FAIL_OK
+js1_5/Regress/regress-224956: FAIL_OK
+js1_5/Regress/regress-325925: FAIL_OK
+ecma_2/RegExp/regress-001: FAIL_OK
+
 # We do not check for bad surrogate pairs when quoting strings.
 js1_5/Regress/regress-315974: FAIL_OK
 
 
 # Use unsupported "watch".
 js1_5/Regress/regress-213482: FAIL_OK
 js1_5/Regress/regress-240577: FAIL_OK
 js1_5/Regress/regress-355344: FAIL_OK
 js1_5/Object/regress-362872-01: FAIL_OK
 js1_5/Object/regress-362872-02: FAIL_OK
 js1_5/Regress/regress-361467: FAIL_OK
 js1_5/Regress/regress-385393-06: FAIL_OK
 js1_5/Regress/regress-506567: FAIL_OK
 
 
 # Use special Mozilla getter/setter syntax
 js1_5/Regress/regress-354924: FAIL_OK
 js1_5/Regress/regress-355341: FAIL_OK
 js1_5/GC/regress-316885-01: FAIL_OK
 js1_5/GetSet/getset-002: FAIL_OK
 js1_5/GetSet/regress-353264: FAIL_OK
 js1_5/Regress/regress-361617: FAIL_OK
 js1_5/Regress/regress-362583: FAIL_OK
 js1_5/extensions/regress-356378: FAIL_OK
 js1_5/extensions/regress-452178: FAIL_OK
 
 
-# 'native' *is* a keyword in V8.
-js1_5/Regress/regress-240317: FAIL_OK
-
-
 # Requires Mozilla-specific strict mode or options() function.
 ecma_3/Object/8.6.1-01: FAIL_OK
 js1_5/Exceptions/regress-315147: FAIL_OK
 js1_5/Regress/regress-106244: FAIL_OK
 js1_5/Regress/regress-317533: FAIL_OK
 js1_5/Regress/regress-323314-1: FAIL_OK
 js1_5/Regress/regress-352197: FAIL_OK
 
 
 # Equivalent to assert(false).
@@ skipping 45 matching lines @@
 
 
 # The length of Error functions is 1 not 3.
 js1_5/Exceptions/regress-123002: FAIL_OK
 
 
 # Reserved keywords as function names, etc is not supported.
 js1_5/LexicalConventions/regress-343675: FAIL_OK
 
 
+# Tests if future reserved keywords of ECMA-262, edition 3 emit warnings. We
+# implement the edition 5 behaviour and fail on use of edition 5 future reserved
+# keywords as identifiers.
+js1_5/Regress/regress-240317: FAIL_OK
+
+
 # Unsupported list comprehensions: [ ... for ... ] and for each.
 js1_5/Regress/regress-352009: FAIL_OK
 js1_5/Regress/regress-349648: FAIL_OK
 
 
 # Expects top level arguments (passed on command line?) to be
 # the empty string?
 js1_5/Regress/regress-336100: FAIL_OK
 
 
@@ skipping 39 matching lines @@
 # This section is for tests that fail in V8 and pass in JSC.
 # Tests that fail in both V8 and JSC belong in the FAIL_OK
 # category.
 
 # This fails because we don't handle Function.prototype.apply with very large
 # numbers of arguments (depending on max stack size).  350256-02 needs more than
 # 4Mbytes of stack space.
 js1_5/Array/regress-350256-02: FAIL
 
 
-# This fails because 'delete arguments[i]' does not disconnect the
-# argument from the arguments array.  See issue #900066.
-ecma_3/Function/regress-137181: FAIL
-
-
-# 'export' and 'import' are not keywords in V8.
-ecma_2/Exceptions/lexical-010: FAIL
-ecma_2/Exceptions/lexical-022: FAIL
-
-
-# Requires Mozilla-specific strict mode.
-ecma_2/Exceptions/lexical-011: FAIL
-ecma_2/Exceptions/lexical-014: FAIL
-ecma_2/Exceptions/lexical-016: FAIL
-ecma_2/Exceptions/lexical-021: FAIL
-ecma_2/LexicalConventions/keywords-001: FAIL
-js1_5/Regress/regress-306633: FAIL
-
-
 # This test seems designed to fail (it produces a 700Mbyte string).
 # We fail on out of memory.  The important thing is not to crash.
 js1_5/Regress/regress-303213: FAIL || TIMEOUT if $mode == debug
 
 # This test fails since we now throw in String.prototype.match when apply
 # is given null or undefined as this argument (and so does firefox nightly).
 js1_5/Regress/regress-295052: FAIL
 
 # Bug 1202592: New ecma_3/String/15.5.4.11 is failing.
 ecma_3/String/15.5.4.11: FAIL
@@ skipping 49 matching lines @@
 js1_5/extensions/regress-350312-03: FAIL_OK
 js1_5/extensions/regress-350531: FAIL_OK
 js1_5/extensions/regress-351102-01: FAIL_OK
 js1_5/extensions/regress-351102-02: FAIL_OK
 js1_5/extensions/regress-351102-06: FAIL_OK
 js1_5/extensions/regress-351973: FAIL_OK
 js1_5/extensions/regress-352060: FAIL_OK
 js1_5/extensions/regress-352094: FAIL_OK
 js1_5/extensions/regress-352261: FAIL_OK
 js1_5/extensions/regress-352281: FAIL_OK
-js1_5/extensions/regress-352372: FAIL_OK
 js1_5/extensions/regress-352455: FAIL_OK
 js1_5/extensions/regress-352604: FAIL_OK
 js1_5/extensions/regress-353214: FAIL_OK
 js1_5/extensions/regress-355339: FAIL_OK
 js1_5/extensions/regress-355497: FAIL_OK
 js1_5/extensions/regress-355622: FAIL_OK
 js1_5/extensions/regress-355736: FAIL_OK
 js1_5/extensions/regress-356085: FAIL_OK
 js1_5/extensions/regress-356106: FAIL_OK
 js1_5/extensions/regress-358594-01: FAIL_OK
@@ skipping 173 matching lines @@
 ecma/Date/15.9.5.12-2: SKIP
 ecma/Date/15.9.5.8: SKIP
 ecma/Date/15.9.5.9: SKIP
 ecma/Date/15.9.5.11-2: SKIP
 ecma/Expressions/11.7.2: SKIP
 ecma/Expressions/11.10-2: SKIP
 ecma/Expressions/11.7.3: SKIP
 ecma/Expressions/11.10-3: SKIP
 ecma/Expressions/11.7.1: SKIP
 ecma_3/RegExp/regress-209067: SKIP