Fix asserts and GC unsafeness in stub generation, bug=1689.

test/mozilla/mozilla.status

 # Copyright 2011 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 51 matching lines @@
 js1_5/Array/regress-465980-02: SKIP
 
 # These tests are simply wrong (i.e., they do not test what they intend
 # to test).
 # In particular, these two compare numbers to NaN with != in the current
 # version of the Mozilla tests. This is *fixed* in a later version.
 # The tests should be re-enabled when switching to a new version.
 ecma_3/Date/15.9.3.2-1: SKIP
 js1_2/function/Number: SKIP
 
-# Causes assert to be triggered: http://code.google.com/p/v8/issues/detail?id=1689
-js1_5/GC/regress-311497: SKIP
-
 ##################### SLOW TESTS #####################
 
 # This takes a long time to run (~100 seconds). It should only be run
 # by the really patient.
 js1_5/GC/regress-324278: SLOW
 
 # This takes a long time to run because our indexOf operation is
 # pretty slow - it causes a lot of GCs; see issue
 # #926379. We could consider marking this SKIP because it takes a
 # while to run to completion.
@@ skipping 775 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