Implement sealing, freezing, and related functions for proxies.

Implement sealing, freezing, and related functions for proxies.

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

Committed: http://code.google.com/p/v8/source/detail?r=8673

[Mads Ager (chromium), 2011-07-17 10:43:47]

LGTM

http://codereview.chromium.org/7391001/diff/3001/src/factory.h
File src/factory.h (right):

http://codereview.chromium.org/7391001/diff/3001/src/factory.h#newcode256
src/factory.h:256: void BecomeJSObject(Handle<JSReceiver> object);
A comment explaining what this operation does would be nice. Also, I guess the
object has to be a proxy or function proxy? If that is the case I think we
should restrict the type from JSReceiver to JSProxy for the argument.

http://codereview.chromium.org/7391001/diff/3001/src/heap.cc
File src/heap.cc (right):

http://codereview.chromium.org/7391001/diff/3001/src/heap.cc#newcode3282
src/heap.cc:3282: MaybeObject* Heap::AllocateGlobalObject(JSFunction*
constructor) {
Accidental edit. :-)

http://codereview.chromium.org/7391001/diff/3001/src/heap.h
File src/heap.h (right):

http://codereview.chromium.org/7391001/diff/3001/src/heap.h#newcode447
src/heap.h:447: MUST_USE_RESULT MaybeObject* ReinitializeJSReceiverAsJSObject(
JSReceiver -> JSProxy

We don't need this operation for anything else and the current implementation
restrictions on the size of the input means that it will not work for much else.
Let's reflect that in the type of the input?

http://codereview.chromium.org/7391001/diff/3001/test/mjsunit/harmony/proxies.js
File test/mjsunit/harmony/proxies.js (right):

http://codereview.chromium.org/7391001/diff/3001/test/mjsunit/harmony/proxies...
test/mjsunit/harmony/proxies.js:589: TestFix(["a", "b", "c", "d", "zz"], {
I would add a newline between each TestFix invocation here to make it easier to
read.

[rossberg, 2011-07-18 10:53:53]

Also removed reservation of in-object properties, and made sure that I
initialize the padding slot.

http://codereview.chromium.org/7391001/diff/3001/src/factory.h
File src/factory.h (right):

http://codereview.chromium.org/7391001/diff/3001/src/factory.h#newcode256
src/factory.h:256: void BecomeJSObject(Handle<JSReceiver> object);
On 2011/07/17 10:43:47, Mads Ager wrote:
> A comment explaining what this operation does would be nice. Also, I guess the
> object has to be a proxy or function proxy? If that is the case I think we
> should restrict the type from JSReceiver to JSProxy for the argument.

Done.

http://codereview.chromium.org/7391001/diff/3001/src/heap.cc
File src/heap.cc (right):

http://codereview.chromium.org/7391001/diff/3001/src/heap.cc#newcode3282
src/heap.cc:3282: MaybeObject* Heap::AllocateGlobalObject(JSFunction*
constructor) {
On 2011/07/17 10:43:47, Mads Ager wrote:
> Accidental edit. :-)

Done.

http://codereview.chromium.org/7391001/diff/3001/src/heap.h
File src/heap.h (right):

http://codereview.chromium.org/7391001/diff/3001/src/heap.h#newcode447
src/heap.h:447: MUST_USE_RESULT MaybeObject* ReinitializeJSReceiverAsJSObject(
On 2011/07/17 10:43:47, Mads Ager wrote:
> JSReceiver -> JSProxy
> 
> We don't need this operation for anything else and the current implementation
> restrictions on the size of the input means that it will not work for much
else.
> Let's reflect that in the type of the input?

Done.

http://codereview.chromium.org/7391001/diff/3001/test/mjsunit/harmony/proxies.js
File test/mjsunit/harmony/proxies.js (right):

http://codereview.chromium.org/7391001/diff/3001/test/mjsunit/harmony/proxies...
test/mjsunit/harmony/proxies.js:589: TestFix(["a", "b", "c", "d", "zz"], {
On 2011/07/17 10:43:47, Mads Ager wrote:
> I would add a newline between each TestFix invocation here to make it easier
to
> read.

Done.

[Mads Ager (chromium), 2011-07-18 11:02:37]

Even better. LGTM!

[Vyacheslav Egorov (Chromium), 2011-07-19 04:07:56]

DBC.

Sorry did not send it earlier had poor connection.

http://codereview.chromium.org/7391001/diff/9002/src/objects.cc
File src/objects.cc (right):

http://codereview.chromium.org/7391001/diff/9002/src/objects.cc#newcode2310
src/objects.cc:2310: ASSERT(IsJSObject());
self->IsJSObject() as we have GC above

http://codereview.chromium.org/7391001/diff/9002/src/runtime.cc
File src/runtime.cc (right):

http://codereview.chromium.org/7391001/diff/9002/src/runtime.cc#newcode622
src/runtime.cc:622: return proxy;
You are potentially returning invalid pointer here cause Fix can cause a GC
(though it does not matter cause return value is ignored atm).

Better return undefined I think

[rossberg, 2011-07-20 09:11:09]

Thanks Slava. Addressed in http://codereview.chromium.org/7460009

http://codereview.chromium.org/7391001/diff/9002/src/objects.cc
File src/objects.cc (right):

http://codereview.chromium.org/7391001/diff/9002/src/objects.cc#newcode2310
src/objects.cc:2310: ASSERT(IsJSObject());
On 2011/07/19 04:07:57, Vyacheslav Egorov wrote:
> self->IsJSObject() as we have GC above

Done.

http://codereview.chromium.org/7391001/diff/9002/src/runtime.cc
File src/runtime.cc (right):

http://codereview.chromium.org/7391001/diff/9002/src/runtime.cc#newcode622
src/runtime.cc:622: return proxy;
On 2011/07/19 04:07:57, Vyacheslav Egorov wrote:
> You are potentially returning invalid pointer here cause Fix can cause a GC
> (though it does not matter cause return value is ignored atm).
> 
> Better return undefined I think

Done.