Disallow subclassing Arrays.

Disallow subclassing Arrays.

R=rossberg@chromium.org,arv@chromium.org
BUG=v8:3930
LOG=Y

Committed: https://crrev.com/87f3e08e72510ee5544e82bb7ad39b2b5f001ad3
Cr-Commit-Position: refs/heads/master@{#26925}

Committed: https://crrev.com/290ee378ea44bda379cb3e1864631895c7dea774
Cr-Commit-Position: refs/heads/master@{#26931}

[Dmitry Lomov (no reviews), 2015-02-27 18:56:51]

PTAL: emergency fix for Array subclassing.

I will take a stab at a proper fix on Monday. It might actually be possible to
make a proper fix non-invasive.

[arv (Not doing code reviews), 2015-02-27 19:15:15]

LGTM

https://codereview.chromium.org/962263002/diff/40001/src/arm64/code-stubs-arm...
File src/arm64/code-stubs-arm64.cc (right):

https://codereview.chromium.org/962263002/diff/40001/src/arm64/code-stubs-arm...
src/arm64/code-stubs-arm64.cc:3075: Register original_constructor = x3;
I like this. How come you decided to do this for only arm64?

[Dmitry Lomov (no reviews), 2015-02-27 19:45:10]

https://codereview.chromium.org/962263002/diff/40001/src/arm64/code-stubs-arm...
File src/arm64/code-stubs-arm64.cc (right):

https://codereview.chromium.org/962263002/diff/40001/src/arm64/code-stubs-arm...
src/arm64/code-stubs-arm64.cc:3075: Register original_constructor = x3;
On 2015/02/27 19:15:15, arv wrote:
> I like this. How come you decided to do this for only arm64?

You mean naming registers? I am just following the style in the rest of the
port, and only arm64 does this

[Dmitry Lomov (no reviews), 2015-02-27 19:46:54]

The CQ bit was checked by dslomov@chromium.org

[commit-bot: I haz the power, 2015-02-27 19:47:58]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/962263002/40001

[arv (Not doing code reviews), 2015-02-27 19:48:15]

On 2015/02/27 19:45:10, Dmitry Lomov (chromium) wrote:
>
https://codereview.chromium.org/962263002/diff/40001/src/arm64/code-stubs-arm...
> File src/arm64/code-stubs-arm64.cc (right):
> 
>
https://codereview.chromium.org/962263002/diff/40001/src/arm64/code-stubs-arm...
> src/arm64/code-stubs-arm64.cc:3075: Register original_constructor = x3;
> On 2015/02/27 19:15:15, arv wrote:
> > I like this. How come you decided to do this for only arm64?
> 
> You mean naming registers? I am just following the style in the rest of the
> port, and only arm64 does this

Yup and I find that it makes the code more readable.

[arv (Not doing code reviews), 2015-02-27 19:48:29]

On 2015/02/27 19:48:15, arv wrote:
> On 2015/02/27 19:45:10, Dmitry Lomov (chromium) wrote:
> >
>
https://codereview.chromium.org/962263002/diff/40001/src/arm64/code-stubs-arm...
> > File src/arm64/code-stubs-arm64.cc (right):
> > 
> >
>
https://codereview.chromium.org/962263002/diff/40001/src/arm64/code-stubs-arm...
> > src/arm64/code-stubs-arm64.cc:3075: Register original_constructor = x3;
> > On 2015/02/27 19:15:15, arv wrote:
> > > I like this. How come you decided to do this for only arm64?
> > 
> > You mean naming registers? I am just following the style in the rest of the
> > port, and only arm64 does this
> 
> Yup and I find that it makes the code more readable.

But no need to change this...

[Dmitry Lomov (no reviews), 2015-02-27 19:52:43]

On 2015/02/27 19:48:29, arv wrote:
> On 2015/02/27 19:48:15, arv wrote:
> > On 2015/02/27 19:45:10, Dmitry Lomov (chromium) wrote:
> > >
> >
>
https://codereview.chromium.org/962263002/diff/40001/src/arm64/code-stubs-arm...
> > > File src/arm64/code-stubs-arm64.cc (right):
> > > 
> > >
> >
>
https://codereview.chromium.org/962263002/diff/40001/src/arm64/code-stubs-arm...
> > > src/arm64/code-stubs-arm64.cc:3075: Register original_constructor = x3;
> > > On 2015/02/27 19:15:15, arv wrote:
> > > > I like this. How come you decided to do this for only arm64?
> > > 
> > > You mean naming registers? I am just following the style in the rest of
the
> > > port, and only arm64 does this
> > 
> > Yup and I find that it makes the code more readable.
> 
> But no need to change this...

Agreed it will be nice.

[commit-bot: I haz the power, 2015-02-27 20:18:02]

Committed patchset #3 (id:40001)

[commit-bot: I haz the power, 2015-02-27 20:18:16]

Patchset 3 (id:??) landed as
https://crrev.com/87f3e08e72510ee5544e82bb7ad39b2b5f001ad3
Cr-Commit-Position: refs/heads/master@{#26925}

[Dmitry Lomov (no reviews), 2015-02-27 20:43:19]

Oops looks like this commit missed a regression test - adding in
https://codereview.chromium.org/961353002

[Dmitry Lomov (no reviews), 2015-02-27 22:26:01]

A revert of this CL (patchset #3 id:40001) has been created in
https://codereview.chromium.org/965053002/ by dslomov@chromium.org.

The reason for reverting is: Fails on nosnap build.

[Dmitry Lomov (no reviews), 2015-02-27 23:36:53]

Please take another look.

Issues with nosnap (with --debug-code, really) fixed. The problem was that I
didn't set up the pointer to feedback vector properly in generated default
constructor. 
All other construct stubs currently ignore feedback vector if
original_constructor != constructor, but ArrayConstructStub didn't (it checks
under --debug-code that it is either an undefined or an allocation site)

[arv (Not doing code reviews), 2015-02-27 23:42:27]

LGTM

[Dmitry Lomov (no reviews), 2015-02-27 23:44:11]

The CQ bit was checked by dslomov@chromium.org

[commit-bot: I haz the power, 2015-02-27 23:45:19]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/962263002/30006

[commit-bot: I haz the power, 2015-02-28 00:28:26]

Committed patchset #4 (id:30006)

[commit-bot: I haz the power, 2015-02-28 00:28:44]

Patchset 4 (id:??) landed as
https://crrev.com/290ee378ea44bda379cb3e1864631895c7dea774
Cr-Commit-Position: refs/heads/master@{#26931}