Implement data property creation for assignments to super.x.

Implement data property creation for assignments to super.x.

R=rossberg@chromium.org, arv@chromium.org
BUG=v8:3330
LOG=N

Committed: https://code.google.com/p/v8/source/detail?r=24331

[Dmitry Lomov (no reviews), 2014-09-30 12:23:25]

dslomov@chromium.org changed reviewers:
+ rossberg@chromium.org
- rossberg@chromium.orh

[Dmitry Lomov (no reviews), 2014-09-30 12:23:26]

PTAL

[rossberg, 2014-09-30 12:57:40]

lgtm

[Dmitry Lomov (no reviews), 2014-09-30 13:23:25]

Committed patchset #1 (id:1) manually as 24331 (presubmit successful).

[arv (Not doing code reviews), 2014-09-30 13:52:17]

https://codereview.chromium.org/618643003/diff/1/test/mjsunit/harmony/super.js
File test/mjsunit/harmony/super.js (right):

https://codereview.chromium.org/618643003/diff/1/test/mjsunit/harmony/super.j...
test/mjsunit/harmony/super.js:126: x : "x from Base"
stick to '

https://codereview.chromium.org/618643003/diff/1/test/mjsunit/harmony/super.j...
test/mjsunit/harmony/super.js:126: x : "x from Base"
no ws before :

https://codereview.chromium.org/618643003/diff/1/test/mjsunit/harmony/super.j...
test/mjsunit/harmony/super.js:210: super.newProperty = 15;
I'm having a hard time seeing why this should throw?

https://people.mozilla.org/~jorendorff/es6-draft.html#sec-putvalue

Is it throwing because we eventually get to a 42[[DefineOwnProperty]] which
returns false?

https://codereview.chromium.org/618643003/diff/1/test/mjsunit/harmony/super.j...
test/mjsunit/harmony/super.js:211: } catch(e) { ex = e; }
ws between catch and (

[arv (Not doing code reviews), 2014-09-30 13:57:06]

LGTM

[Dmitry Lomov (no reviews), 2014-09-30 15:22:42]

On 2014/09/30 13:52:17, arv wrote:
> https://codereview.chromium.org/618643003/diff/1/test/mjsunit/harmony/super.js
> File test/mjsunit/harmony/super.js (right):
> 
>
https://codereview.chromium.org/618643003/diff/1/test/mjsunit/harmony/super.j...
> test/mjsunit/harmony/super.js:126: x : "x from Base"
> stick to '
> 
>
https://codereview.chromium.org/618643003/diff/1/test/mjsunit/harmony/super.j...
> test/mjsunit/harmony/super.js:126: x : "x from Base"
> no ws before :
> 
>
https://codereview.chromium.org/618643003/diff/1/test/mjsunit/harmony/super.j...
> test/mjsunit/harmony/super.js:210: super.newProperty = 15;
> I'm having a hard time seeing why this should throw?
> 
> https://people.mozilla.org/~jorendorff/es6-draft.html#sec-putvalue
> 
> Is it throwing because we eventually get to a 42[[DefineOwnProperty]] which
> returns false?
> 
>
https://codereview.chromium.org/618643003/diff/1/test/mjsunit/harmony/super.j...
> test/mjsunit/harmony/super.js:211: } catch(e) { ex = e; }
> ws between catch and (
Thanks a lot, style nits fixed in https://codereview.chromium.org/613253002

[Toon Verwaest, 2015-02-16 09:33:10]

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

[Toon Verwaest, 2015-02-16 09:33:11]

https://codereview.chromium.org/618643003/diff/1/src/objects.cc
File src/objects.cc (right):

https://codereview.chromium.org/618643003/diff/1/src/objects.cc#newcode2912
src/objects.cc:2912: return JSObject::SetProperty(&own_lookup, value,
strict_mode, store_mode,
This is not correct as far as I can tell. Your implementation will perform the
set based on the state of the receiver.

As far as I read the spec, it says that if no descriptor is found on the
prototype chain, you have to set "OwnDesc" to a writable, enumerable,
configurable data descriptor. If there's an existing previous "valueDesc" on the
receiver, you [[DefineOwnProperty]] to redefine the accessor. Otherwise you
CreateDataProperty(Receiver, P, V).

[Dmitry Lomov (no reviews), 2015-02-17 11:29:32]

On 2015/02/16 09:33:11, Toon Verwaest wrote:
> https://codereview.chromium.org/618643003/diff/1/src/objects.cc
> File src/objects.cc (right):
> 
> https://codereview.chromium.org/618643003/diff/1/src/objects.cc#newcode2912
> src/objects.cc:2912: return JSObject::SetProperty(&own_lookup, value,
> strict_mode, store_mode,
> This is not correct as far as I can tell. Your implementation will perform the
> set based on the state of the receiver.
> 
> As far as I read the spec, it says that if no descriptor is found on the
> prototype chain, you have to set "OwnDesc" to a writable, enumerable,
> configurable data descriptor. If there's an existing previous "valueDesc" on
the
> receiver, you [[DefineOwnProperty]] to redefine the accessor. Otherwise you
> CreateDataProperty(Receiver, P, V).

Is it, though? We only look for OWN properties in the lookup, and I though that
does exactly this.
arv, could you take a look, since I am ooo?

[Toon Verwaest, 2015-02-17 14:07:04]

You aren't guaranteed that the receiver hasn't got OWN properties, since the
receiver was never checked previously. It could have accessors or some data
property. (Or I presume anything else in case the method was applied to some
other receiver; but I'm not sure about that part.)

[Toon Verwaest, 2015-02-17 14:07:55]

Mostly, SetProperty should not know anything about SUPER, Given that SetProperty
in the spec doesn't know anything about it either. It just forces our
SetProperty to get closer to how the spec works.