[turbofan] Fix super property calls to act as method calls.

[turbofan] Fix super property calls to act as method calls.

This fixes [NAMED|KEYED]_SUPER_PROPERTY_CALL to perform a method call
instead of a function call. The difference is visible for sloppy mode
targets that convert primitive receivers.

R=rossberg@chromium.org
TEST=mjsunit/regress/regress-4525
BUG=v8:4525
LOG=n

Committed: https://crrev.com/26f90c95cc8b56a51e462175fd5ebec5399f9a68
Cr-Commit-Position: refs/heads/master@{#31664}

[Michael Starzinger, 2015-10-29 15:12:49]

https://codereview.chromium.org/1419173004/diff/1/test/mjsunit/regress/regres...
File test/mjsunit/regress/regress-4525.js (right):

https://codereview.chromium.org/1419173004/diff/1/test/mjsunit/regress/regres...
test/mjsunit/regress/regress-4525.js:30: assertThrows("o.bar.call(null)",
TypeError);
This call throws with the following exception. I am not 100% convinced this is
what the spec wants us to do, but it is what we are implementing (for both,
baseline and TurboFan) at the moment. I will need your help figuring out whether
this is intended or not.

regress-4525.js:15: TypeError: Cannot convert undefined or null to object
    bar() { return super.foo() }
                         ^
TypeError: Cannot convert undefined or null to object
    at bar (regress-4525.js:15:26)
    at regress-4525.js:30:20

[Michael Starzinger, 2015-10-29 15:13:29]

Description was changed from

==========
[turbofan] Fix super property calls to act as method calls.

This fixed [NAMED|KEYED]_SUPER_PROPERTY_CALL to perform a method call
instead of a function call. The difference is visible for sloppy mode
targets that convert primitive receivers.

R=rossberg@chromium.org
TEST=mjsunit/regress/regress-4525
BUG=v8:4525
LOG=n
==========

to

==========
[turbofan] Fix super property calls to act as method calls.

This fixes [NAMED|KEYED]_SUPER_PROPERTY_CALL to perform a method call
instead of a function call. The difference is visible for sloppy mode
targets that convert primitive receivers.

R=rossberg@chromium.org
TEST=mjsunit/regress/regress-4525
BUG=v8:4525
LOG=n
==========

[Michael Starzinger, 2015-10-29 16:28:57]

https://codereview.chromium.org/1419173004/diff/1/test/mjsunit/regress/regres...
File test/mjsunit/regress/regress-4525.js (right):

https://codereview.chromium.org/1419173004/diff/1/test/mjsunit/regress/regres...
test/mjsunit/regress/regress-4525.js:30: assertThrows("o.bar.call(null)",
TypeError);
On 2015/10/29 15:12:49, Michael Starzinger wrote:
> This call throws with the following exception. I am not 100% convinced this is
> what the spec wants us to do, but it is what we are implementing (for both,
> baseline and TurboFan) at the moment. I will need your help figuring out
whether
> this is intended or not.
> 
> regress-4525.js:15: TypeError: Cannot convert undefined or null to object
>     bar() { return super.foo() }
>                          ^
> TypeError: Cannot convert undefined or null to object
>     at bar (regress-4525.js:15:26)
>     at regress-4525.js:30:20

As per offline discussion, I filed a separate issue for this ...

https://code.google.com/p/v8/issues/detail?id=4526

[rossberg, 2015-10-29 16:31:28]

lgtm

https://codereview.chromium.org/1419173004/diff/20001/test/mjsunit/regress/re...
File test/mjsunit/regress/regress-4525.js (right):

https://codereview.chromium.org/1419173004/diff/20001/test/mjsunit/regress/re...
test/mjsunit/regress/regress-4525.js:30: // TODO(4526): Should convert to global
receiver!
Nit: "... instead of throwing TypeError."

[Michael Starzinger, 2015-10-29 16:41:03]

https://codereview.chromium.org/1419173004/diff/20001/test/mjsunit/regress/re...
File test/mjsunit/regress/regress-4525.js (right):

https://codereview.chromium.org/1419173004/diff/20001/test/mjsunit/regress/re...
test/mjsunit/regress/regress-4525.js:30: // TODO(4526): Should convert to global
receiver!
On 2015/10/29 16:31:28, rossberg wrote:
> Nit: "... instead of throwing TypeError."

Done.

[Michael Starzinger, 2015-10-29 17:17:40]

The CQ bit was checked by mstarzinger@chromium.org

[Michael Starzinger, 2015-10-29 17:17:40]

The patchset sent to the CQ was uploaded after l-g-t-m from
rossberg@chromium.org
Link to the patchset: https://codereview.chromium.org/1419173004/#ps40001
(title: "Addressed comment.")

[commit-bot: I haz the power, 2015-10-29 17:17:48]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1419173004/40001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1419173004/40001

[commit-bot: I haz the power, 2015-10-29 17:19:32]

Committed patchset #3 (id:40001)

[commit-bot: I haz the power, 2015-10-29 17:19:50]

Patchset 3 (id:??) landed as
https://crrev.com/26f90c95cc8b56a51e462175fd5ebec5399f9a68
Cr-Commit-Position: refs/heads/master@{#31664}