[interpreter] Support for ES6 super keyword.

[interpreter] Support for ES6 super keyword.

Adds support for ES6 super keyword and performing loads, stores, and
calls to super class members.

Implements SetHomeObject and enables ThisFunctionVariable.

BUG=v8:4280, v8:4682
LOG=N

Committed: https://crrev.com/e768bcca2494077e634dd9e1fa495b183b4344cc
Cr-Commit-Position: refs/heads/master@{#33977}

[oth, 2016-02-11 11:52:53]

oth@chromium.org changed reviewers:
+ bmeurer@chromium.org, rmcilroy@chromium.org

[oth, 2016-02-11 11:52:54]

rmcilroy@chromium.org: PTAL. Updates to test262.status in the pipe.

bmeurer@chromium.org: Please review changes in compiler.

Thanks

[rmcilroy, 2016-02-11 15:25:49]

Nice! Couple of comments and suggestions.

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
File src/interpreter/bytecode-array-builder.h (right):

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
src/interpreter/bytecode-array-builder.h:171: // The |constructor| is in a
register followed by followed by
delete extra "followed by"

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
File src/interpreter/bytecode-generator.cc (left):

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
src/interpreter/bytecode-generator.cc:2132: VisitPropertyLoad(obj, expr);
Could you merge this back please (and make the 
VisitForRegisterValue(expr->obj()); conditional on it not being a super
property.

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
File src/interpreter/bytecode-generator.cc (right):

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
src/interpreter/bytecode-generator.cc:1049: BuildNamedSuperStore(receiver,
home_object, name);
Could you adapt this function to expect the home_object in the accumulator and
the value as a register? That way you could avoid the extra LoadAccumulator
above (and probably as StoreAccumulatorInReg for the home object too).

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
src/interpreter/bytecode-generator.cc:1063: BuildKeyedSuperStore(receiver,
home_object, key);
ditto

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
src/interpreter/bytecode-generator.cc:1807: void
BytecodeGenerator::BuildKeyedSuperLoad(Register receiver,
BuildKeyedSuperPropertyLoad  (and the same for the ones below) ?

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
src/interpreter/bytecode-generator.cc:1815: }
I'd prefer just to explicitly name each of the registers for clarity (You don't
actually save any lines of code using the args array anyway AFAIKT).

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
src/interpreter/bytecode-generator.cc:2299:
VisitForAccumulatorValue(super_ref->home_object());
Could you use BuildNamedSuperLoad here?

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
src/interpreter/bytecode-generator.cc:2307: if (receiver_out.is_valid()) {
nit - move this up to be below
"builder()->StoreAccumulatorInRegister(receiver);" so that it is clear that this
visitor returns the load from the runtime call in the accumulator.

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
src/interpreter/bytecode-generator.cc:2314: RegisterAllocationScope
register_scope(this);
Same question here, could this use BuildKeyedSuperPropertyLoad?

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
src/interpreter/bytecode-generator.cc:2530: DCHECK_NOT_NULL(super);
I don't think this DCHECK is necessary (it will crash 3 lines down if it's
null).

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
src/interpreter/bytecode-generator.cc:2534: Register this_function =
constructor;
Could you move this down below the CallRuntime with a comment saying that you
reuse the constructor register since it's no longer used.

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
File src/interpreter/bytecode-generator.h (right):

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
src/interpreter/bytecode-generator.h:70: void
VisitKeyedSuperPropertyLoad(Property* property, Register receiver_out);
I think these could do with a comment to specify what is returned in the
accumulator, and that reciever_out is an optional parameter

https://codereview.chromium.org/1689573004/diff/60001/test/cctest/interpreter...
File test/cctest/interpreter/test-bytecode-generator.cc (right):

https://codereview.chromium.org/1689573004/diff/60001/test/cctest/interpreter...
test/cctest/interpreter/test-bytecode-generator.cc:5093: 
Could you add some bytecode generator tests for the super operations please?
Feel free to be a guinee-pig for Stephano's tool if that works.

[Benedikt Meurer, 2016-02-11 17:36:10]

LGTM on compiler.

[oth, 2016-02-12 14:12:54]

Ross, PTAL. There's more common code here and it tries to push all super
property loads/stores through these common paths.

Additional tests for bytecode generator to follow.

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
File src/interpreter/bytecode-array-builder.h (right):

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
src/interpreter/bytecode-array-builder.h:171: // The |constructor| is in a
register followed by followed by
On 2016/02/11 15:25:49, rmcilroy wrote:
> delete extra "followed by"

Done.

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
File src/interpreter/bytecode-generator.cc (left):

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
src/interpreter/bytecode-generator.cc:2132: VisitPropertyLoad(obj, expr);
On 2016/02/11 15:25:49, rmcilroy wrote:
> Could you merge this back please (and make the 
> VisitForRegisterValue(expr->obj()); conditional on it not being a super
> property.

Done.

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
File src/interpreter/bytecode-generator.cc (right):

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
src/interpreter/bytecode-generator.cc:1049: BuildNamedSuperStore(receiver,
home_object, name);
On 2016/02/11 15:25:49, rmcilroy wrote:
> Could you adapt this function to expect the home_object in the accumulator and
> the value as a register? That way you could avoid the extra LoadAccumulator
> above (and probably as StoreAccumulatorInReg for the home object too). 

I'd wondered about this too. Have added PrepareNamedSuperPropertyArguments and
StoreNamedSuperPropertyArguments that reduces the code here.

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
src/interpreter/bytecode-generator.cc:1063: BuildKeyedSuperStore(receiver,
home_object, key);
On 2016/02/11 15:25:49, rmcilroy wrote:
> ditto

Acknowledged.

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
src/interpreter/bytecode-generator.cc:1807: void
BytecodeGenerator::BuildKeyedSuperLoad(Register receiver,
On 2016/02/11 15:25:49, rmcilroy wrote:
> BuildKeyedSuperPropertyLoad  (and the same for the ones below) ?

Done.

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
src/interpreter/bytecode-generator.cc:1815: }
On 2016/02/11 15:25:49, rmcilroy wrote:
> I'd prefer just to explicitly name each of the registers for clarity (You
don't
> actually save any lines of code using the args array anyway AFAIKT).

Keeping this pattern as discussed and incorporated the initialization the array
suggested. WDYT?

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
src/interpreter/bytecode-generator.cc:2299:
VisitForAccumulatorValue(super_ref->home_object());
On 2016/02/11 15:25:49, rmcilroy wrote:
> Could you use BuildNamedSuperLoad here?

Done.

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
src/interpreter/bytecode-generator.cc:2307: if (receiver_out.is_valid()) {
On 2016/02/11 15:25:49, rmcilroy wrote:
> nit - move this up to be below
> "builder()->StoreAccumulatorInRegister(receiver);" so that it is clear that
this
> visitor returns the load from the runtime call in the accumulator.

Done.

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
src/interpreter/bytecode-generator.cc:2314: RegisterAllocationScope
register_scope(this);
On 2016/02/11 15:25:49, rmcilroy wrote:
> Same question here, could this use BuildKeyedSuperPropertyLoad?

Done.

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
src/interpreter/bytecode-generator.cc:2530: DCHECK_NOT_NULL(super);
On 2016/02/11 15:25:49, rmcilroy wrote:
> I don't think this DCHECK is necessary (it will crash 3 lines down if it's
> null).

Done.

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
src/interpreter/bytecode-generator.cc:2534: Register this_function =
constructor;
On 2016/02/11 15:25:49, rmcilroy wrote:
> Could you move this down below the CallRuntime with a comment saying that you
> reuse the constructor register since it's no longer used. 

Done.

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
File src/interpreter/bytecode-generator.h (right):

https://codereview.chromium.org/1689573004/diff/60001/src/interpreter/bytecod...
src/interpreter/bytecode-generator.h:70: void
VisitKeyedSuperPropertyLoad(Property* property, Register receiver_out);
On 2016/02/11 15:25:49, rmcilroy wrote:
> I think these could do with a comment to specify what is returned in the
> accumulator, and that reciever_out is an optional parameter

Done.

https://codereview.chromium.org/1689573004/diff/60001/test/cctest/interpreter...
File test/cctest/interpreter/test-bytecode-generator.cc (right):

https://codereview.chromium.org/1689573004/diff/60001/test/cctest/interpreter...
test/cctest/interpreter/test-bytecode-generator.cc:5093: 
On 2016/02/11 15:25:49, rmcilroy wrote:
> Could you add some bytecode generator tests for the super operations please?
> Feel free to be a guinee-pig for Stephano's tool if that works.

Ack, will do.

[rmcilroy, 2016-02-12 14:56:42]

Looks great, a few last comments.

https://codereview.chromium.org/1689573004/diff/120001/src/interpreter/byteco...
File src/interpreter/bytecode-generator.cc (right):

https://codereview.chromium.org/1689573004/diff/120001/src/interpreter/byteco...
src/interpreter/bytecode-generator.cc:1828:
register_allocator()->PrepareForConsecutiveAllocations(arraysize(args));
unecessary, right? (InitializeWithConsecutiveRegisters does it)

https://codereview.chromium.org/1689573004/diff/120001/src/interpreter/byteco...
src/interpreter/bytecode-generator.cc:1830:
InitializeWithConsecutiveRegisters(args);
nit - move below Register args[4]; and add whitespace before VisitForAccumulator

https://codereview.chromium.org/1689573004/diff/120001/src/interpreter/byteco...
src/interpreter/bytecode-generator.cc:1839:
.StoreAccumulatorInRegister(args[3]);
As discussed, please drop storing language mode here and 'return' args[3] as an
out parameter, or pull the preparing of the args out of the function.

https://codereview.chromium.org/1689573004/diff/120001/src/interpreter/byteco...
src/interpreter/bytecode-generator.cc:1846:
register_allocator()->PrepareForConsecutiveAllocations(arraysize(args));
ditto

https://codereview.chromium.org/1689573004/diff/120001/src/interpreter/byteco...
src/interpreter/bytecode-generator.cc:1848:
InitializeWithConsecutiveRegisters(args);
ditto

https://codereview.chromium.org/1689573004/diff/120001/src/interpreter/byteco...
File src/interpreter/bytecode-generator.h (left):

https://codereview.chromium.org/1689573004/diff/120001/src/interpreter/byteco...
src/interpreter/bytecode-generator.h:70: 
nit - add back whitespace

https://codereview.chromium.org/1689573004/diff/120001/src/interpreter/byteco...
File src/interpreter/bytecode-generator.h (right):

https://codereview.chromium.org/1689573004/diff/120001/src/interpreter/byteco...
src/interpreter/bytecode-generator.h:70: // it. The loaded value is placed in
the accumulator.
s/it./it if it's a valid register./

https://codereview.chromium.org/1689573004/diff/120001/src/interpreter/byteco...
src/interpreter/bytecode-generator.h:76: // it. The loaded value is placed in
the accumulator.
ditto

https://codereview.chromium.org/1689573004/diff/120001/src/interpreter/byteco...
src/interpreter/bytecode-generator.h:155: void
InitializeWithConsecutiveRegisters(Register (&registers)[N]);
This should probably live in the register allocator.

[oth, 2016-02-12 16:50:34]

Comment incorporated and have added class for SuperPropertyArguments as
discussed. PTAL. Thanks!

https://codereview.chromium.org/1689573004/diff/120001/src/interpreter/byteco...
File src/interpreter/bytecode-generator.cc (right):

https://codereview.chromium.org/1689573004/diff/120001/src/interpreter/byteco...
src/interpreter/bytecode-generator.cc:1828:
register_allocator()->PrepareForConsecutiveAllocations(arraysize(args));
On 2016/02/12 14:56:42, rmcilroy wrote:
> unecessary, right? (InitializeWithConsecutiveRegisters does it)

Done.

https://codereview.chromium.org/1689573004/diff/120001/src/interpreter/byteco...
src/interpreter/bytecode-generator.cc:1830:
InitializeWithConsecutiveRegisters(args);
On 2016/02/12 14:56:42, rmcilroy wrote:
> nit - move below Register args[4]; and add whitespace before
VisitForAccumulator

Done.

https://codereview.chromium.org/1689573004/diff/120001/src/interpreter/byteco...
src/interpreter/bytecode-generator.cc:1839:
.StoreAccumulatorInRegister(args[3]);
On 2016/02/12 14:56:42, rmcilroy wrote:
> As discussed, please drop storing language mode here and 'return' args[3] as
an
> out parameter, or pull the preparing of the args out of the function.

Done.

https://codereview.chromium.org/1689573004/diff/120001/src/interpreter/byteco...
src/interpreter/bytecode-generator.cc:1846:
register_allocator()->PrepareForConsecutiveAllocations(arraysize(args));
On 2016/02/12 14:56:42, rmcilroy wrote:
> ditto

Done.

https://codereview.chromium.org/1689573004/diff/120001/src/interpreter/byteco...
src/interpreter/bytecode-generator.cc:1848:
InitializeWithConsecutiveRegisters(args);
On 2016/02/12 14:56:42, rmcilroy wrote:
> ditto

Done.

https://codereview.chromium.org/1689573004/diff/120001/src/interpreter/byteco...
File src/interpreter/bytecode-generator.h (left):

https://codereview.chromium.org/1689573004/diff/120001/src/interpreter/byteco...
src/interpreter/bytecode-generator.h:70: 
On 2016/02/12 14:56:42, rmcilroy wrote:
> nit - add back whitespace

Done.

https://codereview.chromium.org/1689573004/diff/120001/src/interpreter/byteco...
File src/interpreter/bytecode-generator.h (right):

https://codereview.chromium.org/1689573004/diff/120001/src/interpreter/byteco...
src/interpreter/bytecode-generator.h:70: // it. The loaded value is placed in
the accumulator.
On 2016/02/12 14:56:42, rmcilroy wrote:
> s/it./it if it's a valid register./

Done.

https://codereview.chromium.org/1689573004/diff/120001/src/interpreter/byteco...
src/interpreter/bytecode-generator.h:76: // it. The loaded value is placed in
the accumulator.
On 2016/02/12 14:56:42, rmcilroy wrote:
> ditto

Done.

https://codereview.chromium.org/1689573004/diff/120001/src/interpreter/byteco...
src/interpreter/bytecode-generator.h:155: void
InitializeWithConsecutiveRegisters(Register (&registers)[N]);
On 2016/02/12 14:56:42, rmcilroy wrote:
> This should probably live in the register allocator.

Done.

[rmcilroy, 2016-02-12 17:08:43]

LGTM to land when the bots are happy.

https://codereview.chromium.org/1689573004/diff/140001/src/interpreter/byteco...
File src/interpreter/bytecode-generator.cc (right):

https://codereview.chromium.org/1689573004/diff/140001/src/interpreter/byteco...
src/interpreter/bytecode-generator.cc:561: size_t size() const { return
arraysize(args_); }
nit - /s/size/count/

https://codereview.chromium.org/1689573004/diff/160001/test/cctest/interprete...
File test/cctest/interpreter/test-bytecode-generator.cc (right):

https://codereview.chromium.org/1689573004/diff/160001/test/cctest/interprete...
test/cctest/interpreter/test-bytecode-generator.cc:9236: 
Please add a TODO to add class tests when you get back.

[oth, 2016-02-12 17:22:59]

Thanks, investigating test breakage that's introduced after rebase and
SuperPropertyArguments change.

https://codereview.chromium.org/1689573004/diff/140001/src/interpreter/byteco...
File src/interpreter/bytecode-generator.cc (right):

https://codereview.chromium.org/1689573004/diff/140001/src/interpreter/byteco...
src/interpreter/bytecode-generator.cc:561: size_t size() const { return
arraysize(args_); }
On 2016/02/12 17:08:43, rmcilroy wrote:
> nit - /s/size/count/

Done.

https://codereview.chromium.org/1689573004/diff/160001/test/cctest/interprete...
File test/cctest/interpreter/test-bytecode-generator.cc (right):

https://codereview.chromium.org/1689573004/diff/160001/test/cctest/interprete...
test/cctest/interpreter/test-bytecode-generator.cc:9236: 
On 2016/02/12 17:08:43, rmcilroy wrote:
> Please add a TODO to add class tests when you get back.

Done.

[oth, 2016-02-12 21:07:32]

The CQ bit was checked by oth@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-02-12 21:07:35]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1689573004/240001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1689573004/240001

[commit-bot: I haz the power, 2016-02-12 21:46:58]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-02-12 21:46:59]

Dry run: Try jobs failed on following builders:
  v8_linux_arm64_rel on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_arm64_rel/builds/1...)

[oth, 2016-02-13 07:25:27]

The CQ bit was checked by oth@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-02-13 07:25:42]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1689573004/260001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1689573004/260001

[commit-bot: I haz the power, 2016-02-13 07:44:35]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-02-13 07:44:36]

Dry run: Try jobs failed on following builders:
  v8_linux_arm_rel on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_arm_rel/builds/13713)

[oth, 2016-02-13 09:26:05]

The CQ bit was checked by oth@chromium.org

[oth, 2016-02-13 09:26:06]

The patchset sent to the CQ was uploaded after l-g-t-m from
bmeurer@chromium.org, rmcilroy@chromium.org
Link to the patchset: https://codereview.chromium.org/1689573004/#ps280001
(title: "More ARM64 failures.")

[commit-bot: I haz the power, 2016-02-13 09:26:13]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1689573004/280001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1689573004/280001

[commit-bot: I haz the power, 2016-02-13 09:30:54]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-02-13 09:30:55]

Try jobs failed on following builders:
  v8_linux_arm_rel on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_arm_rel/builds/13716)

[oth, 2016-02-13 12:06:43]

The CQ bit was checked by oth@chromium.org

[commit-bot: I haz the power, 2016-02-13 12:06:55]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1689573004/280001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1689573004/280001

[commit-bot: I haz the power, 2016-02-13 12:14:55]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-02-13 12:14:56]

Try jobs failed on following builders:
  v8_linux_arm_rel on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_arm_rel/builds/13718)

[oth, 2016-02-13 12:21:02]

The CQ bit was checked by oth@chromium.org

[oth, 2016-02-13 12:21:02]

The patchset sent to the CQ was uploaded after l-g-t-m from
bmeurer@chromium.org, rmcilroy@chromium.org
Link to the patchset: https://codereview.chromium.org/1689573004/#ps300001
(title: "No change")

[commit-bot: I haz the power, 2016-02-13 12:21:05]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1689573004/300001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1689573004/300001

[commit-bot: I haz the power, 2016-02-13 12:25:56]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-02-13 12:25:57]

Try jobs failed on following builders:
  v8_linux_arm_rel on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_arm_rel/builds/13720)

[oth, 2016-02-15 07:54:37]

The CQ bit was checked by oth@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-02-15 07:54:40]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1689573004/320001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1689573004/320001

[commit-bot: I haz the power, 2016-02-15 08:15:48]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-02-15 08:15:49]

Dry run: This issue passed the CQ dry run.

[oth, 2016-02-15 08:16:56]

The CQ bit was checked by oth@chromium.org

[oth, 2016-02-15 08:16:57]

The patchset sent to the CQ was uploaded after l-g-t-m from
bmeurer@chromium.org, rmcilroy@chromium.org
Link to the patchset: https://codereview.chromium.org/1689573004/#ps320001
(title: "Rebase")

[commit-bot: I haz the power, 2016-02-15 08:17:07]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1689573004/320001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1689573004/320001

[commit-bot: I haz the power, 2016-02-15 08:18:28]

Committed patchset #17 (id:320001)

[commit-bot: I haz the power, 2016-02-15 08:19:14]

Description was changed from

==========
[interpreter] Support for ES6 super keyword.

Adds support for ES6 super keyword and performing loads, stores, and
calls to super class members.

Implements SetHomeObject and enables ThisFunctionVariable.

BUG=v8:4280,v8:4682
LOG=N
==========

to

==========
[interpreter] Support for ES6 super keyword.

Adds support for ES6 super keyword and performing loads, stores, and
calls to super class members.

Implements SetHomeObject and enables ThisFunctionVariable.

BUG=v8:4280,v8:4682
LOG=N

Committed: https://crrev.com/e768bcca2494077e634dd9e1fa495b183b4344cc
Cr-Commit-Position: refs/heads/master@{#33977}
==========

[commit-bot: I haz the power, 2016-02-15 08:19:15]

Patchset 17 (id:??) landed as
https://crrev.com/e768bcca2494077e634dd9e1fa495b183b4344cc
Cr-Commit-Position: refs/heads/master@{#33977}

[Michael Starzinger, 2016-02-16 16:53:08]

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

[Michael Starzinger, 2016-02-16 16:53:09]

https://codereview.chromium.org/1689573004/diff/320001/src/interpreter/byteco...
File src/interpreter/bytecode-generator.cc (right):

https://codereview.chromium.org/1689573004/diff/320001/src/interpreter/byteco...
src/interpreter/bytecode-generator.cc:553: class
BytecodeGenerator::SuperPropertyArguments final {
This assumes that both, the runtime calls for loads and stores always had
exactly 4 arguments. This assumption will not hold for long. In fact I just
broke it today. Not a big fan of this abstraction.