Convert `obj` ToObject in Object.keys()

Convert `obj` ToObject in Object.keys()

BUG=v8:3587
LOG=Y
R=arv@chromium.org, dslomov@chromium.org

Committed: https://code.google.com/p/v8/source/detail?r=24495

[caitp (gmail), 2014-09-22 02:03:01]

Quick patch to match Harmony semantics for ObjectKeys(O)
(https://code.google.com/p/v8/issues/detail?id=3587)

[mathias, 2014-09-22 09:23:45]

Don’t forget to add yourself to the `AUTHORS` file (and sign the CLA if you
haven’t already).

[caitp (gmail), 2014-09-22 13:24:29]

On 2014/09/22 09:23:45, mathias wrote:
> Don’t forget to add yourself to the `AUTHORS` file (and sign the CLA if you
> haven’t already).

Added to V8 AUTHORS

[arv (Not doing code reviews), 2014-09-22 15:32:34]

arv@chromium.org changed reviewers:
+ arv@chromium.org, dslomov@chromium.org

[arv (Not doing code reviews), 2014-09-22 15:32:35]

LGTM

Dmitry, this needs an owners approval too.

[Dmitry Lomov (no reviews), 2014-09-23 01:57:21]

lgtm

[Dmitry Lomov (no reviews), 2014-09-29 07:07:27]

Committed patchset #2 (id:20001) manually as 24260 (presubmit successful).

[Dmitry Lomov (no reviews), 2014-09-29 07:31:01]

Had to revert this because of test262 test failures.

[caitp (gmail), 2014-09-29 14:24:13]

On 2014/09/29 07:31:01, Dmitry Lomov (chromium) wrote:
> Had to revert this because of test262 test failures.

Should see what @Domenic thinks about those --- I recall the idea was that it
wasn't important for an engine to pass 100% of test262, especially since the
requirements of different editions would be changing over time.

In any case, If it's not possible to land the change now, then I guess that's
okay

[arv (Not doing code reviews), 2014-09-29 14:41:02]

On 2014/09/29 14:24:13, caitp wrote:
> On 2014/09/29 07:31:01, Dmitry Lomov (chromium) wrote:
> > Had to revert this because of test262 test failures.
> 
> Should see what @Domenic thinks about those --- I recall the idea was that it
> wasn't important for an engine to pass 100% of test262, especially since the
> requirements of different editions would be changing over time.
> 
> In any case, If it's not possible to land the change now, then I guess that's
> okay

The solution is to disable/skip the es5 test262 tests. Our bots needs to
continue to stay green. To disable the tests update test/test262/test262.status

[caitp (gmail), 2014-09-29 16:42:56]

On 2014/09/29 14:41:02, arv wrote:
> On 2014/09/29 14:24:13, caitp wrote:
> > On 2014/09/29 07:31:01, Dmitry Lomov (chromium) wrote:
> > > Had to revert this because of test262 test failures.
> > 
> > Should see what @Domenic thinks about those --- I recall the idea was that
it
> > wasn't important for an engine to pass 100% of test262, especially since the
> > requirements of different editions would be changing over time.
> > 
> > In any case, If it's not possible to land the change now, then I guess
that's
> > okay
> 
> The solution is to disable/skip the es5 test262 tests. Our bots needs to
> continue to stay green. To disable the tests update
test/test262/test262.status

Disabled (I believe)

[arv (Not doing code reviews), 2014-09-29 18:32:17]

LGTM (test262 passes now)

caitp, fyi:


tools/run-tests.py --mode release --arch ia32 --download-data test262

Runs the test262 tests

[caitp (gmail), 2014-09-29 18:33:00]

On 2014/09/29 18:32:17, arv wrote:
> LGTM (test262 passes now)
> 
> caitp, fyi:
> 
> 
> tools/run-tests.py --mode release --arch ia32 --download-data test262
> 
> Runs the test262 tests

Yes, I have run them --- there are some failures that look unrelated to my
changes (some of the same failures which are disabled on various ARM arches)

[arv (Not doing code reviews), 2014-09-29 18:51:15]

On 2014/09/29 18:33:00, caitp wrote:

> Yes, I have run them --- there are some failures that look unrelated to my
> changes (some of the same failures which are disabled on various ARM arches)

Sorry. Your "I believe" made me assume you had not been able to run the tests.

[caitp (gmail), 2014-10-07 14:34:00]

dslomov@ PTAL?

[Dmitry Lomov (no reviews), 2014-10-09 12:23:19]

On 2014/10/07 14:34:00, caitp wrote:
> dslomov@ PTAL?

Oh sorry, will land this today.

[Dmitry Lomov (no reviews), 2014-10-09 13:19:14]

Committed patchset #3 (id:40001) manually as 24495 (presubmit successful).

[Michael Starzinger, 2014-10-09 16:57:26]

mstarzinger@chromium.org changed reviewers:
+ mstarzinger@chromium.org

[Michael Starzinger, 2014-10-09 16:57:27]

https://codereview.chromium.org/585373003/diff/40001/test/test262/test262.status
File test/test262/test262.status (right):

https://codereview.chromium.org/585373003/diff/40001/test/test262/test262.sta...
test/test262/test262.status:92: '15.2.3.14-1-1': [SKIP],
Is there a particular reason we cannot use "[PASS, FAIL_OK]" here?

[caitp (gmail), 2014-10-09 17:00:10]

On 2014/10/09 16:57:27, Michael Starzinger wrote:
>
https://codereview.chromium.org/585373003/diff/40001/test/test262/test262.status
> File test/test262/test262.status (right):
> 
>
https://codereview.chromium.org/585373003/diff/40001/test/test262/test262.sta...
> test/test262/test262.status:92: '15.2.3.14-1-1': [SKIP],
> Is there a particular reason we cannot use "[PASS, FAIL_OK]" here?

We probably can (I don't know exactly how the test harness works, so it's not
clear what the [PASS, FAIL_OK] means --- it didn't occur to me and nobody asked
for it).