Introduce {FAST,SLOW}_STRING_WRAPPER_ELEMENTS

Introduce {FAST,SLOW}_STRING_WRAPPER_ELEMENTS

String wrappers (new String("foo")) are special objects: their string
characters are accessed like elements, and they also have an elements
backing store. This used to require a bunch of explicit checks like:

if (obj->IsJSValue() && JSValue::cast(obj)->value()->IsString()) {
  /* Handle string characters */
}
// Handle regular elements (for string wrappers and other objects)
obj->GetElementsAccessor()->Whatever(...);

This CL introduces new ElementsKinds for string wrapper objects (one for
fast elements, one for dictionary elements), which allow folding the
special-casing into new StringWrapperElementsAccessors.

No observable change in behavior is intended.

Committed: https://crrev.com/f4872f7477e34ae090e1cb862f22b3b89fba7329
Cr-Commit-Position: refs/heads/master@{#33616}

[Jakob Kummerow, 2016-01-26 14:21:23]

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

[Jakob Kummerow, 2016-01-26 14:21:24]

PTAL. The meat of the change is in elements.cc; the benefit is mostly in
lookup.cc.

I'm gonna need OWNERS' reviews for the (trivial) changes to src/heap/ and
src/interpreter/.

https://codereview.chromium.org/1612323003/diff/40001/src/elements-kind.h
File src/elements-kind.h (right):

https://codereview.chromium.org/1612323003/diff/40001/src/elements-kind.h#new...
src/elements-kind.h:54: NO_ELEMENTS,
Just adding this while I'm touching all the switch statements anyway. Not used
yet.

https://codereview.chromium.org/1612323003/diff/40001/src/elements.cc
File src/elements.cc (left):

https://codereview.chromium.org/1612323003/diff/40001/src/elements.cc#oldcode563
src/elements.cc:563: static inline void SetImpl(FixedArrayBase* backing_store,
uint32_t entry,
Having the compiler complain about not finding a SetImpl implementation for some
subclass is better than failing at runtime when trying to execute it.

https://codereview.chromium.org/1612323003/diff/40001/src/elements.cc
File src/elements.cc (right):

https://codereview.chromium.org/1612323003/diff/40001/src/elements.cc#newcode...
src/elements.cc:1030: static Handle<Object> GetImpl(Handle<JSObject> holder,
uint32_t entry) {
This does exactly the same as the superclass implementation, but the compiler
won't let me override only one of the two overloaded versions (see right below).

https://codereview.chromium.org/1612323003/diff/40001/src/elements.h
File src/elements.h (left):

https://codereview.chromium.org/1612323003/diff/40001/src/elements.h#oldcode88
src/elements.h:88: virtual void CopyElements(
Drive-by cleanup: drop unused functions (one of these three is moved to the
private section, the others aren't needed at all).

https://codereview.chromium.org/1612323003/diff/40001/src/factory.h
File src/factory.h (right):

https://codereview.chromium.org/1612323003/diff/40001/src/factory.h#newcode704
src/factory.h:704: Handle<JSArray> NewJSArray(ElementsKind elements_kind,
This function creates JSArray objects in inconsistent state (length ==
undefined)! Must not be called from the outside -> made it private.

https://codereview.chromium.org/1612323003/diff/40001/src/key-accumulator.h
File src/key-accumulator.h (left):

https://codereview.chromium.org/1612323003/diff/40001/src/key-accumulator.h#o...
src/key-accumulator.h:39: bool AddKey(Object* key, AddKeyConversion convert =
DO_NOT_CONVERT);
Forcing callers to think about what behavior they want is better than silently
doing the wrong default thing.

https://codereview.chromium.org/1612323003/diff/40001/src/lookup.cc
File src/lookup.cc (right):

https://codereview.chromium.org/1612323003/diff/40001/src/lookup.cc#newcode403
src/lookup.cc:403: if (current == holder) return true;
Unrelated to the rest of this CL: avoid instantiating the PrototypeIterator
unless we really need it.

https://codereview.chromium.org/1612323003/diff/40001/src/objects.cc
File src/objects.cc (right):

https://codereview.chromium.org/1612323003/diff/40001/src/objects.cc#newcode1...
src/objects.cc:16492: if (IsJSValue()) {
Intentionally not ElementsAccessor-izing this because the entire function is
deprecated (only being kept around for Object.observe).

https://codereview.chromium.org/1612323003/diff/40001/src/objects.h
File src/objects.h (right):

https://codereview.chromium.org/1612323003/diff/40001/src/objects.h#newcode2637
src/objects.h:2637: static inline Handle<Object> get(FixedArray* array, int
index,
Taking a raw pointer allows some callers to skip creation of a short-lived
handle.

[Hannes Payer (out of office), 2016-01-26 14:39:42]

On 2016/01/26 14:21:24, Jakob wrote:
> PTAL. The meat of the change is in elements.cc; the benefit is mostly in
> lookup.cc.
> 
> I'm gonna need OWNERS' reviews for the (trivial) changes to src/heap/ and
> src/interpreter/.
> 
> https://codereview.chromium.org/1612323003/diff/40001/src/elements-kind.h
> File src/elements-kind.h (right):
> 
>
https://codereview.chromium.org/1612323003/diff/40001/src/elements-kind.h#new...
> src/elements-kind.h:54: NO_ELEMENTS,
> Just adding this while I'm touching all the switch statements anyway. Not used
> yet.
> 
> https://codereview.chromium.org/1612323003/diff/40001/src/elements.cc
> File src/elements.cc (left):
> 
>
https://codereview.chromium.org/1612323003/diff/40001/src/elements.cc#oldcode563
> src/elements.cc:563: static inline void SetImpl(FixedArrayBase* backing_store,
> uint32_t entry,
> Having the compiler complain about not finding a SetImpl implementation for
some
> subclass is better than failing at runtime when trying to execute it.
> 
> https://codereview.chromium.org/1612323003/diff/40001/src/elements.cc
> File src/elements.cc (right):
> 
>
https://codereview.chromium.org/1612323003/diff/40001/src/elements.cc#newcode...
> src/elements.cc:1030: static Handle<Object> GetImpl(Handle<JSObject> holder,
> uint32_t entry) {
> This does exactly the same as the superclass implementation, but the compiler
> won't let me override only one of the two overloaded versions (see right
below).
> 
> https://codereview.chromium.org/1612323003/diff/40001/src/elements.h
> File src/elements.h (left):
> 
> https://codereview.chromium.org/1612323003/diff/40001/src/elements.h#oldcode88
> src/elements.h:88: virtual void CopyElements(
> Drive-by cleanup: drop unused functions (one of these three is moved to the
> private section, the others aren't needed at all).
> 
> https://codereview.chromium.org/1612323003/diff/40001/src/factory.h
> File src/factory.h (right):
> 
> https://codereview.chromium.org/1612323003/diff/40001/src/factory.h#newcode704
> src/factory.h:704: Handle<JSArray> NewJSArray(ElementsKind elements_kind,
> This function creates JSArray objects in inconsistent state (length ==
> undefined)! Must not be called from the outside -> made it private.
> 
> https://codereview.chromium.org/1612323003/diff/40001/src/key-accumulator.h
> File src/key-accumulator.h (left):
> 
>
https://codereview.chromium.org/1612323003/diff/40001/src/key-accumulator.h#o...
> src/key-accumulator.h:39: bool AddKey(Object* key, AddKeyConversion convert =
> DO_NOT_CONVERT);
> Forcing callers to think about what behavior they want is better than silently
> doing the wrong default thing.
> 
> https://codereview.chromium.org/1612323003/diff/40001/src/lookup.cc
> File src/lookup.cc (right):
> 
> https://codereview.chromium.org/1612323003/diff/40001/src/lookup.cc#newcode403
> src/lookup.cc:403: if (current == holder) return true;
> Unrelated to the rest of this CL: avoid instantiating the PrototypeIterator
> unless we really need it.
> 
> https://codereview.chromium.org/1612323003/diff/40001/src/objects.cc
> File src/objects.cc (right):
> 
>
https://codereview.chromium.org/1612323003/diff/40001/src/objects.cc#newcode1...
> src/objects.cc:16492: if (IsJSValue()) {
> Intentionally not ElementsAccessor-izing this because the entire function is
> deprecated (only being kept around for Object.observe).
> 
> https://codereview.chromium.org/1612323003/diff/40001/src/objects.h
> File src/objects.h (right):
> 
>
https://codereview.chromium.org/1612323003/diff/40001/src/objects.h#newcode2637
> src/objects.h:2637: static inline Handle<Object> get(FixedArray* array, int
> index,
> Taking a raw pointer allows some callers to skip creation of a short-lived
> handle.
Heap lgtm

[Yang, 2016-01-26 15:45:11]

On 2016/01/26 14:39:42, Hannes Payer wrote:
> On 2016/01/26 14:21:24, Jakob wrote:
> > PTAL. The meat of the change is in elements.cc; the benefit is mostly in
> > lookup.cc.
> > 
> > I'm gonna need OWNERS' reviews for the (trivial) changes to src/heap/ and
> > src/interpreter/.
> > 
> > https://codereview.chromium.org/1612323003/diff/40001/src/elements-kind.h
> > File src/elements-kind.h (right):
> > 
> >
>
https://codereview.chromium.org/1612323003/diff/40001/src/elements-kind.h#new...
> > src/elements-kind.h:54: NO_ELEMENTS,
> > Just adding this while I'm touching all the switch statements anyway. Not
used
> > yet.
> > 
> > https://codereview.chromium.org/1612323003/diff/40001/src/elements.cc
> > File src/elements.cc (left):
> > 
> >
>
https://codereview.chromium.org/1612323003/diff/40001/src/elements.cc#oldcode563
> > src/elements.cc:563: static inline void SetImpl(FixedArrayBase*
backing_store,
> > uint32_t entry,
> > Having the compiler complain about not finding a SetImpl implementation for
> some
> > subclass is better than failing at runtime when trying to execute it.
> > 
> > https://codereview.chromium.org/1612323003/diff/40001/src/elements.cc
> > File src/elements.cc (right):
> > 
> >
>
https://codereview.chromium.org/1612323003/diff/40001/src/elements.cc#newcode...
> > src/elements.cc:1030: static Handle<Object> GetImpl(Handle<JSObject> holder,
> > uint32_t entry) {
> > This does exactly the same as the superclass implementation, but the
compiler
> > won't let me override only one of the two overloaded versions (see right
> below).
> > 
> > https://codereview.chromium.org/1612323003/diff/40001/src/elements.h
> > File src/elements.h (left):
> > 
> >
https://codereview.chromium.org/1612323003/diff/40001/src/elements.h#oldcode88
> > src/elements.h:88: virtual void CopyElements(
> > Drive-by cleanup: drop unused functions (one of these three is moved to the
> > private section, the others aren't needed at all).
> > 
> > https://codereview.chromium.org/1612323003/diff/40001/src/factory.h
> > File src/factory.h (right):
> > 
> >
https://codereview.chromium.org/1612323003/diff/40001/src/factory.h#newcode704
> > src/factory.h:704: Handle<JSArray> NewJSArray(ElementsKind elements_kind,
> > This function creates JSArray objects in inconsistent state (length ==
> > undefined)! Must not be called from the outside -> made it private.
> > 
> > https://codereview.chromium.org/1612323003/diff/40001/src/key-accumulator.h
> > File src/key-accumulator.h (left):
> > 
> >
>
https://codereview.chromium.org/1612323003/diff/40001/src/key-accumulator.h#o...
> > src/key-accumulator.h:39: bool AddKey(Object* key, AddKeyConversion convert
=
> > DO_NOT_CONVERT);
> > Forcing callers to think about what behavior they want is better than
silently
> > doing the wrong default thing.
> > 
> > https://codereview.chromium.org/1612323003/diff/40001/src/lookup.cc
> > File src/lookup.cc (right):
> > 
> >
https://codereview.chromium.org/1612323003/diff/40001/src/lookup.cc#newcode403
> > src/lookup.cc:403: if (current == holder) return true;
> > Unrelated to the rest of this CL: avoid instantiating the PrototypeIterator
> > unless we really need it.
> > 
> > https://codereview.chromium.org/1612323003/diff/40001/src/objects.cc
> > File src/objects.cc (right):
> > 
> >
>
https://codereview.chromium.org/1612323003/diff/40001/src/objects.cc#newcode1...
> > src/objects.cc:16492: if (IsJSValue()) {
> > Intentionally not ElementsAccessor-izing this because the entire function is
> > deprecated (only being kept around for Object.observe).
> > 
> > https://codereview.chromium.org/1612323003/diff/40001/src/objects.h
> > File src/objects.h (right):
> > 
> >
>
https://codereview.chromium.org/1612323003/diff/40001/src/objects.h#newcode2637
> > src/objects.h:2637: static inline Handle<Object> get(FixedArray* array, int
> > index,
> > Taking a raw pointer allows some callers to skip creation of a short-lived
> > handle.
> Heap lgtm

debug lgtm

[rmcilroy, 2016-01-26 15:53:27]

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

[rmcilroy, 2016-01-26 15:53:28]

interpreter LGTM.

https://codereview.chromium.org/1612323003/diff/40001/src/interpreter/bytecod...
File src/interpreter/bytecode-array-iterator.cc (right):

https://codereview.chromium.org/1612323003/diff/40001/src/interpreter/bytecod...
src/interpreter/bytecode-array-iterator.cc:110: Handle<FixedArray> constants =
handle(bytecode_array()->constant_pool());
Just skip the handle creation here and call get on
bytecode_array()->constant_pool() directly ?

[adamk, 2016-01-26 22:56:51]

Description was changed from

==========
Introduce {FAST,SLOW}_STRING_WRAPPER_ELEMENTS

WIP
==========

to

==========
Introduce {FAST,SLOW}_STRING_WRAPPER_ELEMENTS

WIP
==========

[Toon Verwaest, 2016-01-27 15:02:22]

I'm not such a fan of the changes to Fixed(Double)Array::get; but I won't block
it for that reason either ...
lgtm.

https://codereview.chromium.org/1612323003/diff/40001/src/builtins.cc
File src/builtins.cc (right):

https://codereview.chromium.org/1612323003/diff/40001/src/builtins.cc#newcode839
src/builtins.cc:839: case SLOW_SLOPPY_ARGUMENTS_ELEMENTS:
I guess these cases cannot happen either since arguments objects are not arrays.

https://codereview.chromium.org/1612323003/diff/40001/src/builtins.cc#newcode...
src/builtins.cc:1000: int limit = length < range ? length : range;
int limit = Min(length, range)

https://codereview.chromium.org/1612323003/diff/40001/src/key-accumulator.h
File src/key-accumulator.h (left):

https://codereview.chromium.org/1612323003/diff/40001/src/key-accumulator.h#o...
src/key-accumulator.h:39: bool AddKey(Object* key, AddKeyConversion convert =
DO_NOT_CONVERT);
On 2016/01/26 14:21:24, Jakob wrote:
> Forcing callers to think about what behavior they want is better than silently
> doing the wrong default thing.

Acknowledged.

https://codereview.chromium.org/1612323003/diff/40001/src/lookup.cc
File src/lookup.cc (right):

https://codereview.chromium.org/1612323003/diff/40001/src/lookup.cc#newcode404
src/lookup.cc:404: // JSProxy do not occur as hidden prototypes.
What about adding if (!holder->map()->is_hidden_prototype()) return false;

Perhaps both in a small separate CL?

[Jakob Kummerow, 2016-01-29 14:31:27]

Thanks for the reviews. Addressed comments.

https://codereview.chromium.org/1612323003/diff/40001/src/builtins.cc
File src/builtins.cc (right):

https://codereview.chromium.org/1612323003/diff/40001/src/builtins.cc#newcode839
src/builtins.cc:839: case SLOW_SLOPPY_ARGUMENTS_ELEMENTS:
On 2016/01/27 15:02:21, Toon Verwaest wrote:
> I guess these cases cannot happen either since arguments objects are not
arrays.

Done.

https://codereview.chromium.org/1612323003/diff/40001/src/builtins.cc#newcode...
src/builtins.cc:1000: int limit = length < range ? length : range;
On 2016/01/27 15:02:21, Toon Verwaest wrote:
> int limit = Min(length, range)

Done.

https://codereview.chromium.org/1612323003/diff/40001/src/interpreter/bytecod...
File src/interpreter/bytecode-array-iterator.cc (right):

https://codereview.chromium.org/1612323003/diff/40001/src/interpreter/bytecod...
src/interpreter/bytecode-array-iterator.cc:110: Handle<FixedArray> constants =
handle(bytecode_array()->constant_pool());
On 2016/01/26 15:53:27, rmcilroy wrote:
> Just skip the handle creation here and call get on
> bytecode_array()->constant_pool() directly ?

Most excellent point, dunno why I didn't see that before. Done.

https://codereview.chromium.org/1612323003/diff/40001/src/lookup.cc
File src/lookup.cc (right):

https://codereview.chromium.org/1612323003/diff/40001/src/lookup.cc#newcode404
src/lookup.cc:404: // JSProxy do not occur as hidden prototypes.
On 2016/01/27 15:02:22, Toon Verwaest wrote:
> What about adding if (!holder->map()->is_hidden_prototype()) return false;
> 
> Perhaps both in a small separate CL?

Done.

[Jakob Kummerow, 2016-01-29 14:43:55]

The CQ bit was checked by jkummerow@chromium.org

[Jakob Kummerow, 2016-01-29 14:43:55]

The patchset sent to the CQ was uploaded after l-g-t-m from hpayer@chromium.org,
verwaest@chromium.org, rmcilroy@chromium.org, yangguo@chromium.org
Link to the patchset: https://codereview.chromium.org/1612323003/#ps80001
(title: "rebased to ToT")

[commit-bot: I haz the power, 2016-01-29 14:44:05]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1612323003/80001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1612323003/80001

[commit-bot: I haz the power, 2016-01-29 14:49:08]

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

[commit-bot: I haz the power, 2016-01-29 14:49:08]

Try jobs failed on following builders:
  v8_win_rel_ng on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_win_rel_ng/builds/2105)

[Jakob Kummerow, 2016-01-29 17:51:36]

Patchset #6 (id:100001) has been deleted

[Jakob Kummerow, 2016-01-29 17:51:57]

The CQ bit was checked by jkummerow@chromium.org

[Jakob Kummerow, 2016-01-29 17:51:57]

The patchset sent to the CQ was uploaded after l-g-t-m from hpayer@chromium.org,
verwaest@chromium.org, rmcilroy@chromium.org, yangguo@chromium.org
Link to the patchset: https://codereview.chromium.org/1612323003/#ps120001
(title: "make MSVS and test262 happy")

[commit-bot: I haz the power, 2016-01-29 17:52:05]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1612323003/120001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1612323003/120001

[commit-bot: I haz the power, 2016-01-29 18:07:35]

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

[commit-bot: I haz the power, 2016-01-29 18:07:36]

Try jobs failed on following builders:
  v8_win_rel_ng on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_win_rel_ng/builds/2112)
  v8_win_rel_ng_triggered on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_win_rel_ng_triggered/bui...)

[adamk, 2016-01-29 18:16:20]

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

[adamk, 2016-01-29 18:16:20]

DBC: Please give this a more detailed CL description before landing.

[Jakob Kummerow, 2016-01-29 18:17:44]

On 2016/01/29 18:16:20, adamk wrote:
> DBC: Please give this a more detailed CL description before landing.

Oops, somehow that didn't get updated. Fixing.

[Jakob Kummerow, 2016-01-29 18:18:20]

Description was changed from

==========
Introduce {FAST,SLOW}_STRING_WRAPPER_ELEMENTS

WIP
==========

to

==========
Introduce {FAST,SLOW}_STRING_WRAPPER_ELEMENTS

String wrappers (new String("foo")) are special objects: their string
characters are accessed like elements, and they also have an elements
backing store. This used to require a bunch of explicit checks like:

if (obj->IsJSValue() && JSValue::cast(obj)->value()->IsString()) {
  /* Handle string characters */
}
// Handle regular elements (for string wrappers and other objects)
obj->GetElementsAccessor()->Whatever(...);

This CL introduces new ElementsKinds for string wrapper objects (one for
fast elements, one for dictionary elements), which allow folding the
special-casing into new StringWrapperElementsAccessors.

No observable change in behavior is intended.
==========

[Jakob Kummerow, 2016-01-29 18:31:22]

The CQ bit was checked by jkummerow@chromium.org

[Jakob Kummerow, 2016-01-29 18:31:22]

The patchset sent to the CQ was uploaded after l-g-t-m from hpayer@chromium.org,
verwaest@chromium.org, rmcilroy@chromium.org, yangguo@chromium.org
Link to the patchset: https://codereview.chromium.org/1612323003/#ps140001
(title: "one more DCHECK fix")

[commit-bot: I haz the power, 2016-01-29 18:31:24]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1612323003/140001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1612323003/140001

[commit-bot: I haz the power, 2016-01-29 18:57:31]

Description was changed from

==========
Introduce {FAST,SLOW}_STRING_WRAPPER_ELEMENTS

String wrappers (new String("foo")) are special objects: their string
characters are accessed like elements, and they also have an elements
backing store. This used to require a bunch of explicit checks like:

if (obj->IsJSValue() && JSValue::cast(obj)->value()->IsString()) {
  /* Handle string characters */
}
// Handle regular elements (for string wrappers and other objects)
obj->GetElementsAccessor()->Whatever(...);

This CL introduces new ElementsKinds for string wrapper objects (one for
fast elements, one for dictionary elements), which allow folding the
special-casing into new StringWrapperElementsAccessors.

No observable change in behavior is intended.
==========

to

==========
Introduce {FAST,SLOW}_STRING_WRAPPER_ELEMENTS

String wrappers (new String("foo")) are special objects: their string
characters are accessed like elements, and they also have an elements
backing store. This used to require a bunch of explicit checks like:

if (obj->IsJSValue() && JSValue::cast(obj)->value()->IsString()) {
  /* Handle string characters */
}
// Handle regular elements (for string wrappers and other objects)
obj->GetElementsAccessor()->Whatever(...);

This CL introduces new ElementsKinds for string wrapper objects (one for
fast elements, one for dictionary elements), which allow folding the
special-casing into new StringWrapperElementsAccessors.

No observable change in behavior is intended.
==========

[commit-bot: I haz the power, 2016-01-29 18:57:32]

Committed patchset #7 (id:140001)

[commit-bot: I haz the power, 2016-01-29 18:58:10]

Description was changed from

==========
Introduce {FAST,SLOW}_STRING_WRAPPER_ELEMENTS

String wrappers (new String("foo")) are special objects: their string
characters are accessed like elements, and they also have an elements
backing store. This used to require a bunch of explicit checks like:

if (obj->IsJSValue() && JSValue::cast(obj)->value()->IsString()) {
  /* Handle string characters */
}
// Handle regular elements (for string wrappers and other objects)
obj->GetElementsAccessor()->Whatever(...);

This CL introduces new ElementsKinds for string wrapper objects (one for
fast elements, one for dictionary elements), which allow folding the
special-casing into new StringWrapperElementsAccessors.

No observable change in behavior is intended.
==========

to

==========
Introduce {FAST,SLOW}_STRING_WRAPPER_ELEMENTS

String wrappers (new String("foo")) are special objects: their string
characters are accessed like elements, and they also have an elements
backing store. This used to require a bunch of explicit checks like:

if (obj->IsJSValue() && JSValue::cast(obj)->value()->IsString()) {
  /* Handle string characters */
}
// Handle regular elements (for string wrappers and other objects)
obj->GetElementsAccessor()->Whatever(...);

This CL introduces new ElementsKinds for string wrapper objects (one for
fast elements, one for dictionary elements), which allow folding the
special-casing into new StringWrapperElementsAccessors.

No observable change in behavior is intended.

Committed: https://crrev.com/f4872f7477e34ae090e1cb862f22b3b89fba7329
Cr-Commit-Position: refs/heads/master@{#33616}
==========

[commit-bot: I haz the power, 2016-01-29 18:58:12]

Patchset 7 (id:??) landed as
https://crrev.com/f4872f7477e34ae090e1cb862f22b3b89fba7329
Cr-Commit-Position: refs/heads/master@{#33616}