Review URL: http://codereview.chromium.org/8271007

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9655 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/7904008

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9650 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Just use script being native and FLAG_allow_natives_syntax directly.

Review URL: http://codereview.chromium.org/8314018

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9643 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

BUG=v8:1767
TEST=mjsunit/numops-fuzz doesn't timeout in debug mode

Review URL: http://codereview.chromium.org/8302023

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9642 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

According to the ES5 spec all ">" and "<=" expressions should be be
evaluated left-to-right. This obsoletes old hacks for reversing the
order to be ES3 compliant.

R=lrn@chromium.org
BUG=v8:1752

Review URL: http://codereview.chromium.org/8275035

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9641 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Problematic code is e.g.:

   function typeofViaVariable() {
     var foo = typeof({})
     return foo === "undefined";
   }
Review URL: http://codereview.chromium.org/8283036

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9640 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/8291004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9639 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

It causes an assertion with deoptimizing from inlined code.
Review URL: http://codereview.chromium.org/8277034

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9636 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

TBR=whesse@chromium.org
Review URL: http://codereview.chromium.org/8286023

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9635 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

TBR=whesse@chromium.org
Review URL: http://codereview.chromium.org/8286022

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9633 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/8276030

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9626 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

We currently report how much memory is available without changing page, i.e., the actual amount of available memory is not correctly stated for new space when doing --trace-gc.
Review URL: http://codereview.chromium.org/8275026

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9624 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/8274032

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9622 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

This has already been fixed, we not actually use the constraints.
Review URL: http://codereview.chromium.org/8276029

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9620 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

1. Record AST id for CallFunctionStub.

2. Correctly extract cached target from CallFunctionStub inline cache.

3. Fix a bug when inling call as a function in effect or value context:
   Handle abnormal exits correcty.

4. Fix a bug when inlining call as a function in test context: drop function
   correctly from true and false block.

5. Avoid inlining mutually recursive functions by checking the stack of function
   states before inlining. This was not a bug, but is just a more general
   check to avoid recursive inlining.
Review URL: http://codereview.chromium.org/8258012

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9619 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Omit smi check inside write barriers if the value is known to be a heap object.

Refine inferred types of some instructions.
Review URL: http://codereview.chromium.org/8256016

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9618 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Currently, preparser-process crashes when given flags with arguments (e.g. --gc_interval 10). It can be fixed by ignoring everything in the command line except the "throws" command and its immediate arguments. This assumes that no flags appear between "throws" and its arguments.

TEST=make ia32.release.check TESTFLAGS="preparser --special-command=\"@ --gc_interval 10\""

Review URL: http://codereview.chromium.org/8268004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9617 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

BUG=
TEST=

Review URL: http://codereview.chromium.org/8283025
Patch from Paul Lind <plind44@gmail.com>.

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9616 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/8281001

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9615 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=danno@chromium.org

Review URL: http://codereview.chromium.org/8278001

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9614 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Fix & tweak some proxy-related error messages.

R=kmillikin@chromium.org
BUG=v8:1543
TEST=

Review URL: http://codereview.chromium.org/8229008

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9613 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Previously, comparisons against null/undefined and comparisons of the result of
typeof against a constant string were done syntactically. Now we do this via
pattern matching on the hydrogen graph, which opens up more opportunities for
better code generation, e.g. the following comparisons are now recognized to be
special:

   var bar = typeof foo;
   var baz = "undefined";
   if (bar == baz) ...

   var blah = undefined;
   if (hurz == blah) ...

If we did this handling of special cases even later at lithium generation time,
even more cases could be recognized, but this would involve bigger changes and
this CL handles most common cases.
Review URL: http://codereview.chromium.org/8242002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9612 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

We run with --nolimit-inlining when --stress-opt is on.
Review URL: http://codereview.chromium.org/8267007

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9610 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=erik.corry@gmail.com

Review URL: http://codereview.chromium.org/8261008

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9609 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/8261007

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9608 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/8267004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9607 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/8256012

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9606 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

BUG=
TEST=

Review URL: http://codereview.chromium.org/8241003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9605 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/8262010

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9602 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

This ports the following list of commits to mips. It is provided in this form
so that the mips port of the new-gc can be easily tested on your end.

This must be downloaded or landed after http://codereview.chromium.org/8106002
This is based on r9585. With these two mips commits, all tests are passing.

9319    Fix a harmless assert and a genuine bug in the GC-safety of stub generation ...
9329    ARM: Pregenerate some stubs that we call from other stubs.
9335    Initialize pre-allocated fields of JSObject with undefined.
9344    Put back the asserts in RememberedSetHelper, but correct this time
9370    Reorganize object type enum, such that proxies are no longer in the middle...
9392    Basic support for tracking smi-only arrays on ia32.
9402    Notify collector about lazily deoptimized code objects.
9411    Porting r9392 to arm (smi-only arrays).
9418    Small refactor to KeyedStoreIC::GenerateGeneric to make it slightly faster.
9447    Tighten up assertions checking GC-safety of stub calls.
9449    Record function call targets, use them for inlining.
9459    Make sure we don't flush the pregenerated stubs, since they need to b
9461    Fix the build on ARM
9466    Move the is_pregenerated flag so it does not overlap other flags....
9468    Fix the no-VFP3 build on ARM.
9475    Pass correct anchor_slot for EMBEDDED_OBJECT pointers from
9490    Adjust assertions in UpdateSlot to match UpdatePointer in PointersUpdatingVisitor.
9511    Clean list of external references from internal objects like the hole value.

9514    Simplify compares in KeyedStoreIC::GenerateGeneric.
9531    Porting r9456 to arm (Optimize KeyedStoreGeneric for Smi arrays).
9541    Fix load of potentially eval-shadowed let bindings.
9542    Fast allocation of block contexts.
9553    Activate smi-only optimizations for large array literals.

9575    Move declaration of SerializedScopeInfo from variables.h to objects.h
9577    Track elements_kind transitions in KeyedStoreICs.
9583    Fixing a bug in arm as pointed out in issue 1759.
9584    Refactor and fix polymorphic KeyedStoreIC creation.

BUG=
TEST=

Review URL: http://codereview.chromium.org/8112008
Patch from Paul Lind <pling44@gmail.com>.

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9601 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Simplified based on Michael's change Refactor how embedded pointers are visited. (9597)

Ported r9328 (bdc13b7)

BUG=
TEST=

Review URL: http://codereview.chromium.org/8106002
Patch from Paul Lind <pling44@gmail.com>.

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9600 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

TEST=Compiling Chromium with -Duse_system_v8 no longer prints a GYP warning about "invalid type 'settings'".

Review URL: http://codereview.chromium.org/8256008

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9599 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/8240007

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9598 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

This refactoring (almost) gets rid of the requirement to get the target
object address for an object pointer embedded in code objects. This is
not possible on MIPS as pointers are encoded using two instructions. All
usages of RelocInfo::target_object_address() are (almost) obsoleted by
this change. The serializer still uses it, so MIPS will not yet work
with snapshots turned on.

R=danno@chromium.org,vegorov@chromium.org

Review URL: http://codereview.chromium.org/8245007

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9597 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

According to the ES5 spec the implementation of DefineOwnProperty() has
to special case handling of arrays. This is a preliminary implementation
correctly handling definition of array index properties, defining length
properties is not completely covered yet.

R=rossberg@chromium.org
TEST=test262

Review URL: http://codereview.chromium.org/8221002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9596 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/8229009

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9590 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Shorten --harmony-block-scoping to --harmony-scoping.

R=keuchel@chromium.org
BUG=
TEST=

Review URL: http://codereview.chromium.org/8226017

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9589 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=fschneider@chromium.org
BUG=v8:1763

Review URL: http://codereview.chromium.org/8240004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9588 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

This is a temporary fix which avoids compaction when incremental marking
is restarted during an old-space-step. That could turn the page that
holds the chosen free list node into an evacuation candidate. It could
also cause several other inconsistencies if it happens during scavenge.

R=vegorov@chromium.org

Review URL: http://codereview.chromium.org/8228010

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9585 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/8233011

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@9584 ce2b1a6d-e550-0410-aec6-3dcde31c8c00