Implement the new semantics for 'super(...)'

Implement the new semantics for 'super(...)'

Per the latest ES6 draft, super(...) translates into a call
to function's prototype.

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

Committed: https://code.google.com/p/v8/source/detail?r=24683

[Dmitry Lomov (no reviews), 2014-10-15 10:19:45]

FYI, before platform ports. 
Take a look if you have time.

[Dmitry Lomov (no reviews), 2014-10-15 11:59:10]

PTAL, platform ports finished.

[arv (Not doing code reviews), 2014-10-15 14:47:18]

LGTM

https://codereview.chromium.org/661433002/diff/20001/src/ia32/full-codegen-ia...
File src/ia32/full-codegen-ia32.cc (right):

https://codereview.chromium.org/661433002/diff/20001/src/ia32/full-codegen-ia...
src/ia32/full-codegen-ia32.cc:1: / Copyright 2012 the V8 project authors. All
rights reserved.
oops

https://codereview.chromium.org/661433002/diff/20001/test/mjsunit/harmony/sup...
File test/mjsunit/harmony/super.js (right):

https://codereview.chromium.org/661433002/diff/20001/test/mjsunit/harmony/sup...
test/mjsunit/harmony/super.js:1254: super();
Just to be clear that I understand this. This one calls?

ImplicitSubclassOfObject.__proto__.call(ImplicitSubclassOfObject);

which happens to be

Function.prototype.call(ImplicitSubclassOfObject);

https://codereview.chromium.org/661433002/diff/20001/test/mjsunit/harmony/sup...
test/mjsunit/harmony/super.js:1269: // we throw TypeError.
Completely agree. I was reading up on the changes to the spec in preparation for
this CL and it should obviously by TypeError.

[arv (Not doing code reviews), 2014-10-15 14:49:54]

https://codereview.chromium.org/661433002/diff/20001/test/mjsunit/harmony/sup...
File test/mjsunit/harmony/super.js (right):

https://codereview.chromium.org/661433002/diff/20001/test/mjsunit/harmony/sup...
test/mjsunit/harmony/super.js:1272: }());
maybe add a test that mutates __proto__ and then uses the new proto?

var calls = 0;
function G() {
  calls++;
}
function F() {}
  super();
}
F.__proto__ = G;
new F();
assertEquals(1, calls);
F.__proto__ = function() {};
new F();
assertEquals(1, calls);

[Dmitry Lomov (no reviews), 2014-10-15 17:22:57]

Erik, thanks for review! CL updated.

Toon/Igor, could you take a look at the codegen as well?

https://codereview.chromium.org/661433002/diff/20001/src/ia32/full-codegen-ia...
File src/ia32/full-codegen-ia32.cc (right):

https://codereview.chromium.org/661433002/diff/20001/src/ia32/full-codegen-ia...
src/ia32/full-codegen-ia32.cc:1: / Copyright 2012 the V8 project authors. All
rights reserved.
On 2014/10/15 14:47:18, arv wrote:
> oops

Done.

https://codereview.chromium.org/661433002/diff/20001/test/mjsunit/harmony/sup...
File test/mjsunit/harmony/super.js (right):

https://codereview.chromium.org/661433002/diff/20001/test/mjsunit/harmony/sup...
test/mjsunit/harmony/super.js:1254: super();
On 2014/10/15 14:47:18, arv wrote:
> Just to be clear that I understand this. This one calls?
> 
> ImplicitSubclassOfObject.__proto__.call(ImplicitSubclassOfObject);
> 
> which happens to be
> 
> Function.prototype.call(ImplicitSubclassOfObject);

You are right, this has to be called ImplicitSubclassOfFunction.
Renamed.

https://codereview.chromium.org/661433002/diff/20001/test/mjsunit/harmony/sup...
test/mjsunit/harmony/super.js:1272: }());
On 2014/10/15 14:49:54, arv wrote:
> maybe add a test that mutates __proto__ and then uses the new proto?
> 
> var calls = 0;
> function G() {
>   calls++;
> }
> function F() {}
>   super();
> }
> F.__proto__ = G;
> new F();
> assertEquals(1, calls);
> F.__proto__ = function() {};
> new F();
> assertEquals(1, calls);
> 

Done.

[Igor Sheludko, 2014-10-17 08:40:34]

lgtm with a nit:

https://codereview.chromium.org/661433002/diff/40001/src/compiler/ast-graph-b...
File src/compiler/ast-graph-builder.cc (right):

https://codereview.chromium.org/661433002/diff/40001/src/compiler/ast-graph-b...
src/compiler/ast-graph-builder.cc:1266: UNREACHABLE();
UNIMPLEMENTED(); (UNREACHABLE is expanded to no-op in release mode)

[Dmitry Lomov (no reviews), 2014-10-17 09:15:47]

Comments addressed, landing

https://codereview.chromium.org/661433002/diff/40001/src/compiler/ast-graph-b...
File src/compiler/ast-graph-builder.cc (right):

https://codereview.chromium.org/661433002/diff/40001/src/compiler/ast-graph-b...
src/compiler/ast-graph-builder.cc:1266: UNREACHABLE();
On 2014/10/17 08:40:34, Igor Sheludko wrote:
> UNIMPLEMENTED(); (UNREACHABLE is expanded to no-op in release mode)

Done.

[Dmitry Lomov (no reviews), 2014-10-17 09:16:16]

Committed patchset #4 (id:60001) manually as 24683 (presubmit successful).