Add ExternalStringResourceBase::IsCompressible

Add Add ExternalStringResourceBase::IsCompressible

This CL introduces ExternalStringResourceBase::IsCompressible.

This CL is a preparation for CompressibleString, which can
be compressed for memory reduction in Blink. We've found that
JavaScript strings account for a relatively large part of Blink
memory usage, and we are now trying to replace JavaScript String/
AtomicString with CompressibleString.

When a string is compressed, the original char data is deleted
and V8 pointer cache becomes invalid. This CL introduces
isCompressible  property and if an external string's isCompressble
return true, this is stored short_external_*_map instead of
external_*_map so that V8 always requires the char pointer whenever
V8 needs the string data.

BUG=chromium:574317
LOG=n

Committed: https://crrev.com/150887a13c303665103aed8aa97c604c462e725c
Cr-Commit-Position: refs/heads/master@{#33224}

[hajimehoshi, 2015-12-17 11:05:11]

Description was changed from

==========
WIP: Introduce CompressibleString

BUG=
==========

to

==========
Add Add ExternalStringResourceBase::isCompressible

This CL introduces ExternalStringResourceBase::isCompressible.

This CL is a preparation for CompressibleString, which can
be compressed for memory reduction in Blink. We've found that
JavaScript strings account for a relatively large part of Blink
memory usage, and we are now trying to replace JavaScript String/
AtomicString with CompressibleString.

When a string is compressed, the original char data is deleted
and V8 pointer cache becomes invalid. This CL introduces
isCompressible  property and if an external string's isCompressble
return true, this is stored short_external_*_map instead of
external_*_map so that V8 always requires the char pointer whenever
V8 needs the string data.

BUG=n/a
TEST=n/a
==========

[hajimehoshi, 2015-12-17 11:05:11]

hajimehoshi@chromium.org changed reviewers:
+ yangguo@chromium.org

[hajimehoshi, 2015-12-17 11:05:53]

PTAL

[Yang, 2015-12-17 11:41:53]

looking good. some comments.

https://codereview.chromium.org/1490193002/diff/40001/include/v8.h
File include/v8.h (right):

https://codereview.chromium.org/1490193002/diff/40001/include/v8.h#newcode2195
include/v8.h:2195: virtual bool isCompressible() const { return false; }
Naming convention would name this "IsCompressible".

https://codereview.chromium.org/1490193002/diff/40001/src/factory.cc
File src/factory.cc (right):

https://codereview.chromium.org/1490193002/diff/40001/src/factory.cc#newcode675
src/factory.cc:675: map = short_external_one_byte_string_map();
Please add a TODO to rename this to "uncached_external_one_byte_string_map()"
and similar for other short external string maps.
Or do the renaming in this CL.

https://codereview.chromium.org/1490193002/diff/40001/src/objects.cc
File src/objects.cc (right):

https://codereview.chromium.org/1490193002/diff/40001/src/objects.cc#newcode1690
src/objects.cc:1690: if (size < ExternalString::kSize ||
resource->isCompressible()) {
Is this necessary? For source strings, we always create external strings from
scratch, and do not turn V8 strings into external strings, right? Can we instead
assert that resource is not compressible?

[hajimehoshi, 2016-01-05 08:07:44]

Thank you!

https://codereview.chromium.org/1490193002/diff/40001/include/v8.h
File include/v8.h (right):

https://codereview.chromium.org/1490193002/diff/40001/include/v8.h#newcode2195
include/v8.h:2195: virtual bool isCompressible() const { return false; }
On 2015/12/17 11:41:53, Yang wrote:
> Naming convention would name this "IsCompressible".

Done.

https://codereview.chromium.org/1490193002/diff/40001/src/factory.cc
File src/factory.cc (right):

https://codereview.chromium.org/1490193002/diff/40001/src/factory.cc#newcode675
src/factory.cc:675: map = short_external_one_byte_string_map();
On 2015/12/17 11:41:53, Yang wrote:
> Please add a TODO to rename this to "uncached_external_one_byte_string_map()"
> and similar for other short external string maps.
> Or do the renaming in this CL.

Done.

https://codereview.chromium.org/1490193002/diff/40001/src/objects.cc
File src/objects.cc (right):

https://codereview.chromium.org/1490193002/diff/40001/src/objects.cc#newcode1690
src/objects.cc:1690: if (size < ExternalString::kSize ||
resource->isCompressible()) {
On 2015/12/17 11:41:53, Yang wrote:
> Is this necessary? For source strings, we always create external strings from
> scratch, and do not turn V8 strings into external strings, right? Can we
instead
> assert that resource is not compressible?

I meant compressible string should use short_*_map to avoid caching in the same
way we did in factory.cc. Am I missing something?

[hajimehoshi, 2016-01-05 08:15:39]

Description was changed from

==========
Add Add ExternalStringResourceBase::isCompressible

This CL introduces ExternalStringResourceBase::isCompressible.

This CL is a preparation for CompressibleString, which can
be compressed for memory reduction in Blink. We've found that
JavaScript strings account for a relatively large part of Blink
memory usage, and we are now trying to replace JavaScript String/
AtomicString with CompressibleString.

When a string is compressed, the original char data is deleted
and V8 pointer cache becomes invalid. This CL introduces
isCompressible  property and if an external string's isCompressble
return true, this is stored short_external_*_map instead of
external_*_map so that V8 always requires the char pointer whenever
V8 needs the string data.

BUG=n/a
TEST=n/a
==========

to

==========
Add Add ExternalStringResourceBase::IsCompressible

This CL introduces ExternalStringResourceBase::IsCompressible.

This CL is a preparation for CompressibleString, which can
be compressed for memory reduction in Blink. We've found that
JavaScript strings account for a relatively large part of Blink
memory usage, and we are now trying to replace JavaScript String/
AtomicString with CompressibleString.

When a string is compressed, the original char data is deleted
and V8 pointer cache becomes invalid. This CL introduces
isCompressible  property and if an external string's isCompressble
return true, this is stored short_external_*_map instead of
external_*_map so that V8 always requires the char pointer whenever
V8 needs the string data.

BUG=n/a
TEST=n/a
==========

[Yang, 2016-01-05 08:16:43]

On 2016/01/05 08:07:44, hajimehoshi wrote:
> Thank you!
> 
> https://codereview.chromium.org/1490193002/diff/40001/include/v8.h
> File include/v8.h (right):
> 
> https://codereview.chromium.org/1490193002/diff/40001/include/v8.h#newcode2195
> include/v8.h:2195: virtual bool isCompressible() const { return false; }
> On 2015/12/17 11:41:53, Yang wrote:
> > Naming convention would name this "IsCompressible".
> 
> Done.
> 
> https://codereview.chromium.org/1490193002/diff/40001/src/factory.cc
> File src/factory.cc (right):
> 
>
https://codereview.chromium.org/1490193002/diff/40001/src/factory.cc#newcode675
> src/factory.cc:675: map = short_external_one_byte_string_map();
> On 2015/12/17 11:41:53, Yang wrote:
> > Please add a TODO to rename this to
"uncached_external_one_byte_string_map()"
> > and similar for other short external string maps.
> > Or do the renaming in this CL.
> 
> Done.
> 
> https://codereview.chromium.org/1490193002/diff/40001/src/objects.cc
> File src/objects.cc (right):
> 
>
https://codereview.chromium.org/1490193002/diff/40001/src/objects.cc#newcode1690
> src/objects.cc:1690: if (size < ExternalString::kSize ||
> resource->isCompressible()) {
> On 2015/12/17 11:41:53, Yang wrote:
> > Is this necessary? For source strings, we always create external strings
from
> > scratch, and do not turn V8 strings into external strings, right? Can we
> instead
> > assert that resource is not compressible?
> 
> I meant compressible string should use short_*_map to avoid caching in the
same
> way we did in factory.cc. Am I missing something?

I was saying that the purpose you are intending to use compressible external
strings for, would never run into this code path. Source strings are external
strings from the start (created by factory), and not externalized later on. So
it seems to me that we could just assert that this does not happen here.

[hajimehoshi, 2016-01-05 08:25:37]

Description was changed from

==========
Add Add ExternalStringResourceBase::IsCompressible

This CL introduces ExternalStringResourceBase::IsCompressible.

This CL is a preparation for CompressibleString, which can
be compressed for memory reduction in Blink. We've found that
JavaScript strings account for a relatively large part of Blink
memory usage, and we are now trying to replace JavaScript String/
AtomicString with CompressibleString.

When a string is compressed, the original char data is deleted
and V8 pointer cache becomes invalid. This CL introduces
isCompressible  property and if an external string's isCompressble
return true, this is stored short_external_*_map instead of
external_*_map so that V8 always requires the char pointer whenever
V8 needs the string data.

BUG=n/a
TEST=n/a
==========

to

==========
Add Add ExternalStringResourceBase::IsCompressible

This CL introduces ExternalStringResourceBase::IsCompressible.

This CL is a preparation for CompressibleString, which can
be compressed for memory reduction in Blink. We've found that
JavaScript strings account for a relatively large part of Blink
memory usage, and we are now trying to replace JavaScript String/
AtomicString with CompressibleString.

When a string is compressed, the original char data is deleted
and V8 pointer cache becomes invalid. This CL introduces
isCompressible  property and if an external string's isCompressble
return true, this is stored short_external_*_map instead of
external_*_map so that V8 always requires the char pointer whenever
V8 needs the string data.

BUG=n/a (crbug/574317)
TEST=n/a
==========

[hajimehoshi, 2016-01-05 08:50:07]

On 2016/01/05 08:16:43, Yang wrote:
> On 2016/01/05 08:07:44, hajimehoshi wrote:
> > Thank you!
> > 
> > https://codereview.chromium.org/1490193002/diff/40001/include/v8.h
> > File include/v8.h (right):
> > 
> >
https://codereview.chromium.org/1490193002/diff/40001/include/v8.h#newcode2195
> > include/v8.h:2195: virtual bool isCompressible() const { return false; }
> > On 2015/12/17 11:41:53, Yang wrote:
> > > Naming convention would name this "IsCompressible".
> > 
> > Done.
> > 
> > https://codereview.chromium.org/1490193002/diff/40001/src/factory.cc
> > File src/factory.cc (right):
> > 
> >
>
https://codereview.chromium.org/1490193002/diff/40001/src/factory.cc#newcode675
> > src/factory.cc:675: map = short_external_one_byte_string_map();
> > On 2015/12/17 11:41:53, Yang wrote:
> > > Please add a TODO to rename this to
> "uncached_external_one_byte_string_map()"
> > > and similar for other short external string maps.
> > > Or do the renaming in this CL.
> > 
> > Done.
> > 
> > https://codereview.chromium.org/1490193002/diff/40001/src/objects.cc
> > File src/objects.cc (right):
> > 
> >
>
https://codereview.chromium.org/1490193002/diff/40001/src/objects.cc#newcode1690
> > src/objects.cc:1690: if (size < ExternalString::kSize ||
> > resource->isCompressible()) {
> > On 2015/12/17 11:41:53, Yang wrote:
> > > Is this necessary? For source strings, we always create external strings
> from
> > > scratch, and do not turn V8 strings into external strings, right? Can we
> > instead
> > > assert that resource is not compressible?
> > 
> > I meant compressible string should use short_*_map to avoid caching in the
> same
> > way we did in factory.cc. Am I missing something?
> 
> I was saying that the purpose you are intending to use compressible external
> strings for, would never run into this code path. Source strings are external
> strings from the start (created by factory), and not externalized later on. So
> it seems to me that we could just assert that this does not happen here.

I understood. Added DCHECKs. Thank you!

[hajimehoshi, 2016-01-06 10:47:10]

On 2016/01/05 08:50:07, hajimehoshi wrote:
> On 2016/01/05 08:16:43, Yang wrote:
> > On 2016/01/05 08:07:44, hajimehoshi wrote:
> > > Thank you!
> > > 
> > > https://codereview.chromium.org/1490193002/diff/40001/include/v8.h
> > > File include/v8.h (right):
> > > 
> > >
> https://codereview.chromium.org/1490193002/diff/40001/include/v8.h#newcode2195
> > > include/v8.h:2195: virtual bool isCompressible() const { return false; }
> > > On 2015/12/17 11:41:53, Yang wrote:
> > > > Naming convention would name this "IsCompressible".
> > > 
> > > Done.
> > > 
> > > https://codereview.chromium.org/1490193002/diff/40001/src/factory.cc
> > > File src/factory.cc (right):
> > > 
> > >
> >
>
https://codereview.chromium.org/1490193002/diff/40001/src/factory.cc#newcode675
> > > src/factory.cc:675: map = short_external_one_byte_string_map();
> > > On 2015/12/17 11:41:53, Yang wrote:
> > > > Please add a TODO to rename this to
> > "uncached_external_one_byte_string_map()"
> > > > and similar for other short external string maps.
> > > > Or do the renaming in this CL.
> > > 
> > > Done.
> > > 
> > > https://codereview.chromium.org/1490193002/diff/40001/src/objects.cc
> > > File src/objects.cc (right):
> > > 
> > >
> >
>
https://codereview.chromium.org/1490193002/diff/40001/src/objects.cc#newcode1690
> > > src/objects.cc:1690: if (size < ExternalString::kSize ||
> > > resource->isCompressible()) {
> > > On 2015/12/17 11:41:53, Yang wrote:
> > > > Is this necessary? For source strings, we always create external strings
> > from
> > > > scratch, and do not turn V8 strings into external strings, right? Can we
> > > instead
> > > > assert that resource is not compressible?
> > > 
> > > I meant compressible string should use short_*_map to avoid caching in the
> > same
> > > way we did in factory.cc. Am I missing something?
> > 
> > I was saying that the purpose you are intending to use compressible external
> > strings for, would never run into this code path. Source strings are
external
> > strings from the start (created by factory), and not externalized later on.
So
> > it seems to me that we could just assert that this does not happen here.
> 
> I understood. Added DCHECKs. Thank you!

ping

[Yang, 2016-01-06 11:02:37]

On 2016/01/06 10:47:10, hajimehoshi wrote:
> On 2016/01/05 08:50:07, hajimehoshi wrote:
> > On 2016/01/05 08:16:43, Yang wrote:
> > > On 2016/01/05 08:07:44, hajimehoshi wrote:
> > > > Thank you!
> > > > 
> > > > https://codereview.chromium.org/1490193002/diff/40001/include/v8.h
> > > > File include/v8.h (right):
> > > > 
> > > >
> >
https://codereview.chromium.org/1490193002/diff/40001/include/v8.h#newcode2195
> > > > include/v8.h:2195: virtual bool isCompressible() const { return false; }
> > > > On 2015/12/17 11:41:53, Yang wrote:
> > > > > Naming convention would name this "IsCompressible".
> > > > 
> > > > Done.
> > > > 
> > > > https://codereview.chromium.org/1490193002/diff/40001/src/factory.cc
> > > > File src/factory.cc (right):
> > > > 
> > > >
> > >
> >
>
https://codereview.chromium.org/1490193002/diff/40001/src/factory.cc#newcode675
> > > > src/factory.cc:675: map = short_external_one_byte_string_map();
> > > > On 2015/12/17 11:41:53, Yang wrote:
> > > > > Please add a TODO to rename this to
> > > "uncached_external_one_byte_string_map()"
> > > > > and similar for other short external string maps.
> > > > > Or do the renaming in this CL.
> > > > 
> > > > Done.
> > > > 
> > > > https://codereview.chromium.org/1490193002/diff/40001/src/objects.cc
> > > > File src/objects.cc (right):
> > > > 
> > > >
> > >
> >
>
https://codereview.chromium.org/1490193002/diff/40001/src/objects.cc#newcode1690
> > > > src/objects.cc:1690: if (size < ExternalString::kSize ||
> > > > resource->isCompressible()) {
> > > > On 2015/12/17 11:41:53, Yang wrote:
> > > > > Is this necessary? For source strings, we always create external
strings
> > > from
> > > > > scratch, and do not turn V8 strings into external strings, right? Can
we
> > > > instead
> > > > > assert that resource is not compressible?
> > > > 
> > > > I meant compressible string should use short_*_map to avoid caching in
the
> > > same
> > > > way we did in factory.cc. Am I missing something?
> > > 
> > > I was saying that the purpose you are intending to use compressible
external
> > > strings for, would never run into this code path. Source strings are
> external
> > > strings from the start (created by factory), and not externalized later
on.
> So
> > > it seems to me that we could just assert that this does not happen here.
> > 
> > I understood. Added DCHECKs. Thank you!
> 
> ping

I wont be back from vacation until tomorrow.

[Yang, 2016-01-07 05:46:20]

On 2016/01/06 11:02:37, Yang wrote:
> On 2016/01/06 10:47:10, hajimehoshi wrote:
> > On 2016/01/05 08:50:07, hajimehoshi wrote:
> > > On 2016/01/05 08:16:43, Yang wrote:
> > > > On 2016/01/05 08:07:44, hajimehoshi wrote:
> > > > > Thank you!
> > > > > 
> > > > > https://codereview.chromium.org/1490193002/diff/40001/include/v8.h
> > > > > File include/v8.h (right):
> > > > > 
> > > > >
> > >
> https://codereview.chromium.org/1490193002/diff/40001/include/v8.h#newcode2195
> > > > > include/v8.h:2195: virtual bool isCompressible() const { return false;
}
> > > > > On 2015/12/17 11:41:53, Yang wrote:
> > > > > > Naming convention would name this "IsCompressible".
> > > > > 
> > > > > Done.
> > > > > 
> > > > > https://codereview.chromium.org/1490193002/diff/40001/src/factory.cc
> > > > > File src/factory.cc (right):
> > > > > 
> > > > >
> > > >
> > >
> >
>
https://codereview.chromium.org/1490193002/diff/40001/src/factory.cc#newcode675
> > > > > src/factory.cc:675: map = short_external_one_byte_string_map();
> > > > > On 2015/12/17 11:41:53, Yang wrote:
> > > > > > Please add a TODO to rename this to
> > > > "uncached_external_one_byte_string_map()"
> > > > > > and similar for other short external string maps.
> > > > > > Or do the renaming in this CL.
> > > > > 
> > > > > Done.
> > > > > 
> > > > > https://codereview.chromium.org/1490193002/diff/40001/src/objects.cc
> > > > > File src/objects.cc (right):
> > > > > 
> > > > >
> > > >
> > >
> >
>
https://codereview.chromium.org/1490193002/diff/40001/src/objects.cc#newcode1690
> > > > > src/objects.cc:1690: if (size < ExternalString::kSize ||
> > > > > resource->isCompressible()) {
> > > > > On 2015/12/17 11:41:53, Yang wrote:
> > > > > > Is this necessary? For source strings, we always create external
> strings
> > > > from
> > > > > > scratch, and do not turn V8 strings into external strings, right?
Can
> we
> > > > > instead
> > > > > > assert that resource is not compressible?
> > > > > 
> > > > > I meant compressible string should use short_*_map to avoid caching in
> the
> > > > same
> > > > > way we did in factory.cc. Am I missing something?
> > > > 
> > > > I was saying that the purpose you are intending to use compressible
> external
> > > > strings for, would never run into this code path. Source strings are
> > external
> > > > strings from the start (created by factory), and not externalized later
> on.
> > So
> > > > it seems to me that we could just assert that this does not happen here.
> > > 
> > > I understood. Added DCHECKs. Thank you!
> > 
> > ping
> 
> I wont be back from vacation until tomorrow.

lgtm.

[hajimehoshi, 2016-01-07 05:50:59]

The CQ bit was checked by hajimehoshi@chromium.org

[commit-bot: I haz the power, 2016-01-07 05:51:01]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1490193002/80001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1490193002/80001

[commit-bot: I haz the power, 2016-01-07 05:53:28]

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

[commit-bot: I haz the power, 2016-01-07 05:53:29]

Try jobs failed on following builders:
  v8_presubmit on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_presubmit/builds/9504)

[hajimehoshi, 2016-01-12 05:40:44]

Description was changed from

==========
Add Add ExternalStringResourceBase::IsCompressible

This CL introduces ExternalStringResourceBase::IsCompressible.

This CL is a preparation for CompressibleString, which can
be compressed for memory reduction in Blink. We've found that
JavaScript strings account for a relatively large part of Blink
memory usage, and we are now trying to replace JavaScript String/
AtomicString with CompressibleString.

When a string is compressed, the original char data is deleted
and V8 pointer cache becomes invalid. This CL introduces
isCompressible  property and if an external string's isCompressble
return true, this is stored short_external_*_map instead of
external_*_map so that V8 always requires the char pointer whenever
V8 needs the string data.

BUG=n/a (crbug/574317)
TEST=n/a
==========

to

==========
Add Add ExternalStringResourceBase::IsCompressible

This CL introduces ExternalStringResourceBase::IsCompressible.

This CL is a preparation for CompressibleString, which can
be compressed for memory reduction in Blink. We've found that
JavaScript strings account for a relatively large part of Blink
memory usage, and we are now trying to replace JavaScript String/
AtomicString with CompressibleString.

When a string is compressed, the original char data is deleted
and V8 pointer cache becomes invalid. This CL introduces
isCompressible  property and if an external string's isCompressble
return true, this is stored short_external_*_map instead of
external_*_map so that V8 always requires the char pointer whenever
V8 needs the string data.

BUG=chromium:574317
==========

[hajimehoshi, 2016-01-12 05:41:29]

Description was changed from

==========
Add Add ExternalStringResourceBase::IsCompressible

This CL introduces ExternalStringResourceBase::IsCompressible.

This CL is a preparation for CompressibleString, which can
be compressed for memory reduction in Blink. We've found that
JavaScript strings account for a relatively large part of Blink
memory usage, and we are now trying to replace JavaScript String/
AtomicString with CompressibleString.

When a string is compressed, the original char data is deleted
and V8 pointer cache becomes invalid. This CL introduces
isCompressible  property and if an external string's isCompressble
return true, this is stored short_external_*_map instead of
external_*_map so that V8 always requires the char pointer whenever
V8 needs the string data.

BUG=chromium:574317
==========

to

==========
Add Add ExternalStringResourceBase::IsCompressible

This CL introduces ExternalStringResourceBase::IsCompressible.

This CL is a preparation for CompressibleString, which can
be compressed for memory reduction in Blink. We've found that
JavaScript strings account for a relatively large part of Blink
memory usage, and we are now trying to replace JavaScript String/
AtomicString with CompressibleString.

When a string is compressed, the original char data is deleted
and V8 pointer cache becomes invalid. This CL introduces
isCompressible  property and if an external string's isCompressble
return true, this is stored short_external_*_map instead of
external_*_map so that V8 always requires the char pointer whenever
V8 needs the string data.

BUG=chromium:574317
LOG=n
==========

[hajimehoshi, 2016-01-12 05:45:17]

The CQ bit was checked by hajimehoshi@chromium.org

[commit-bot: I haz the power, 2016-01-12 05:45:22]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1490193002/80001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1490193002/80001

[commit-bot: I haz the power, 2016-01-12 06:28:13]

Description was changed from

==========
Add Add ExternalStringResourceBase::IsCompressible

This CL introduces ExternalStringResourceBase::IsCompressible.

This CL is a preparation for CompressibleString, which can
be compressed for memory reduction in Blink. We've found that
JavaScript strings account for a relatively large part of Blink
memory usage, and we are now trying to replace JavaScript String/
AtomicString with CompressibleString.

When a string is compressed, the original char data is deleted
and V8 pointer cache becomes invalid. This CL introduces
isCompressible  property and if an external string's isCompressble
return true, this is stored short_external_*_map instead of
external_*_map so that V8 always requires the char pointer whenever
V8 needs the string data.

BUG=chromium:574317
LOG=n
==========

to

==========
Add Add ExternalStringResourceBase::IsCompressible

This CL introduces ExternalStringResourceBase::IsCompressible.

This CL is a preparation for CompressibleString, which can
be compressed for memory reduction in Blink. We've found that
JavaScript strings account for a relatively large part of Blink
memory usage, and we are now trying to replace JavaScript String/
AtomicString with CompressibleString.

When a string is compressed, the original char data is deleted
and V8 pointer cache becomes invalid. This CL introduces
isCompressible  property and if an external string's isCompressble
return true, this is stored short_external_*_map instead of
external_*_map so that V8 always requires the char pointer whenever
V8 needs the string data.

BUG=chromium:574317
LOG=n
==========

[commit-bot: I haz the power, 2016-01-12 06:28:13]

Committed patchset #5 (id:80001)

[commit-bot: I haz the power, 2016-01-12 06:29:07]

Description was changed from

==========
Add Add ExternalStringResourceBase::IsCompressible

This CL introduces ExternalStringResourceBase::IsCompressible.

This CL is a preparation for CompressibleString, which can
be compressed for memory reduction in Blink. We've found that
JavaScript strings account for a relatively large part of Blink
memory usage, and we are now trying to replace JavaScript String/
AtomicString with CompressibleString.

When a string is compressed, the original char data is deleted
and V8 pointer cache becomes invalid. This CL introduces
isCompressible  property and if an external string's isCompressble
return true, this is stored short_external_*_map instead of
external_*_map so that V8 always requires the char pointer whenever
V8 needs the string data.

BUG=chromium:574317
LOG=n
==========

to

==========
Add Add ExternalStringResourceBase::IsCompressible

This CL introduces ExternalStringResourceBase::IsCompressible.

This CL is a preparation for CompressibleString, which can
be compressed for memory reduction in Blink. We've found that
JavaScript strings account for a relatively large part of Blink
memory usage, and we are now trying to replace JavaScript String/
AtomicString with CompressibleString.

When a string is compressed, the original char data is deleted
and V8 pointer cache becomes invalid. This CL introduces
isCompressible  property and if an external string's isCompressble
return true, this is stored short_external_*_map instead of
external_*_map so that V8 always requires the char pointer whenever
V8 needs the string data.

BUG=chromium:574317
LOG=n

Committed: https://crrev.com/150887a13c303665103aed8aa97c604c462e725c
Cr-Commit-Position: refs/heads/master@{#33224}
==========

[commit-bot: I haz the power, 2016-01-12 06:29:08]

Patchset 5 (id:??) landed as
https://crrev.com/150887a13c303665103aed8aa97c604c462e725c
Cr-Commit-Position: refs/heads/master@{#33224}

[Yang, 2016-01-25 08:31:09]

https://codereview.chromium.org/1490193002/diff/80001/src/factory.cc
File src/factory.cc (right):

https://codereview.chromium.org/1490193002/diff/80001/src/factory.cc#newcode704
src/factory.cc:704: // TODO(hajimehoshi): Rename these to
'uncached_external_string_...'.
Please do not forget this TODO!

[hajimehoshi, 2016-01-25 10:10:19]

On 2016/01/25 08:31:09, Yang wrote:
> https://codereview.chromium.org/1490193002/diff/80001/src/factory.cc
> File src/factory.cc (right):
> 
>
https://codereview.chromium.org/1490193002/diff/80001/src/factory.cc#newcode704
> src/factory.cc:704: // TODO(hajimehoshi): Rename these to
> 'uncached_external_string_...'.
> Please do not forget this TODO!

Thank you for reminding me :-)

I should rename not only variable names but also constants like
kShortExternal..., shouldn't I?

[Yang, 2016-01-25 10:11:41]

On 2016/01/25 10:10:19, hajimehoshi wrote:
> On 2016/01/25 08:31:09, Yang wrote:
> > https://codereview.chromium.org/1490193002/diff/80001/src/factory.cc
> > File src/factory.cc (right):
> > 
> >
>
https://codereview.chromium.org/1490193002/diff/80001/src/factory.cc#newcode704
> > src/factory.cc:704: // TODO(hajimehoshi): Rename these to
> > 'uncached_external_string_...'.
> > Please do not forget this TODO!
> 
> Thank you for reminding me :-)
> 
> I should rename not only variable names but also constants like
> kShortExternal..., shouldn't I?

Yes. And enum types like SHORT_EXTERNAL_STRING_TYPE.