Use invalidation sets for fullscreen pseudos.

Use invalidation sets for fullscreen pseudos.

Schedule invalidation sets on elements changing state for
:-webkit-full-screen and:-webkit-full-screen-ancestor.

Lazily load the UA style for fullscreen, but before we enter fullscreen
the first time to have the invalidation sets available for style
invalidation

BUG=442239
Committed: https://crrev.com/5a2027442fee79bb42fe601a608d3d5459991bc8
Cr-Commit-Position: refs/heads/master@{#382405}

[rune, 2016-01-20 00:30:46]

Description was changed from

==========
Use invalidation sets for fullscreen pseudos.

Schedule invalidation sets on elements changing state for
:-webkit-full-screen, :-webkit-full-screen-ancestor, and
:-webkit-full-screen-document.

The latter gets special treatment since it is not an element state but a
state for the whole document. That means any element may have to have its
style recalculated when the fullscreen state changes. We do that by still
doing a full document style recalc when that pseudo class is present. If
we remove the :-webkit-full-screen-document pseudo support, we'll get rid
of that full document recalc completely.

BUG=442239
==========

to

==========
Use invalidation sets for fullscreen pseudos.

Schedule invalidation sets on elements changing state for
:-webkit-full-screen and:-webkit-full-screen-ancestor.

BUG=442239
==========

[rune, 2016-01-20 00:31:07]

The CQ bit was checked by rune@opera.com to run a CQ dry run

[commit-bot: I haz the power, 2016-01-20 00:34:17]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/803133002/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/803133002/20001

[commit-bot: I haz the power, 2016-01-20 03:51:37]

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

[commit-bot: I haz the power, 2016-01-20 03:51:37]

Dry run: Try jobs failed on following builders:
  linux_android_rel_ng on tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/linux_androi...)

[rune, 2016-03-13 20:54:46]

The CQ bit was checked by rune@opera.com to run a CQ dry run

[commit-bot: I haz the power, 2016-03-13 20:54:51]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/803133002/40001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/803133002/40001

[commit-bot: I haz the power, 2016-03-13 22:11:21]

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

[commit-bot: I haz the power, 2016-03-13 22:11:21]

Dry run: This issue passed the CQ dry run.

[rune, 2016-03-13 22:29:02]

Description was changed from

==========
Use invalidation sets for fullscreen pseudos.

Schedule invalidation sets on elements changing state for
:-webkit-full-screen and:-webkit-full-screen-ancestor.

BUG=442239
==========

to

==========
Use invalidation sets for fullscreen pseudos.

Schedule invalidation sets on elements changing state for
:-webkit-full-screen and:-webkit-full-screen-ancestor.

Add the UA style for fullscreen at the same time as html.css to have
invalidation sets available first time going fullscreen.

BUG=442239
==========

[rune, 2016-03-13 22:34:53]

rune@opera.com changed reviewers:
+ esprehn@chromium.org, philipj@opera.com, timloh@chromium.org

[rune, 2016-03-13 22:34:53]

Loading fullscreen.css UA style when loading html.css means we'll have
invalidation sets available when going fullscreen the first time. After doing
that change, tests pass.

PTAL

[Timothy Loh, 2016-03-13 23:46:54]

On 2016/03/13 22:34:53, rune (slow - travelling) wrote:
> Loading fullscreen.css UA style when loading html.css means we'll have
> invalidation sets available when going fullscreen the first time. After doing
> that change, tests pass.
> 
> PTAL

lgtm

[esprehn, 2016-03-14 02:18:43]

Why have the separate sheet at all then? I think we want to break the UA sheet
into even smaller chunks not make it bigger.

[rune, 2016-03-14 17:42:40]

On 2016/03/14 02:18:43, esprehn wrote:
> Why have the separate sheet at all then? I think we want to break the UA sheet
> into even smaller chunks not make it bigger.

That looks contradictory to me. You think fullscreen.css should be merged into
html.css, or not?

[rune, 2016-03-17 08:15:40]

On 2016/03/14 17:42:40, rune wrote:
> On 2016/03/14 02:18:43, esprehn wrote:
> > Why have the separate sheet at all then? I think we want to break the UA
sheet
> > into even smaller chunks not make it bigger.
> 
> That looks contradictory to me. You think fullscreen.css should be merged into
> html.css, or not?

@esprehn?

[esprehn, 2016-03-18 05:05:58]

On 2016/03/17 at 08:15:40, rune wrote:
> On 2016/03/14 17:42:40, rune wrote:
> > On 2016/03/14 02:18:43, esprehn wrote:
> > > Why have the separate sheet at all then? I think we want to break the UA
sheet
> > > into even smaller chunks not make it bigger.
> > 
> > That looks contradictory to me. You think fullscreen.css should be merged
into
> > html.css, or not?
> 
> @esprehn?

Parsing the ua sheet isn't free, kouhei@ has commented that it takes time in the
critical path during renderer startup. I think we should consider breaking it
apart into a couple sheets, ex. forms.css, html.css, section.css so that the
common rules are in a smaller sheet.

For example there's a rule for h1 when under a article,aside,nav,section, but
those elements are pretty uncommon, we could avoid even parsing the rare css
until we see those elements avoiding the extra rules for h1 and the parsing
cost.

In this case if we're always going to add the full screen sheet, it should just
be in html.css, but really we should keep it separate and only load it when
we're in full screen mode.

[rune, 2016-03-21 10:58:34]

The CQ bit was checked by rune@opera.com to run a CQ dry run

[commit-bot: I haz the power, 2016-03-21 10:58:48]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/803133002/80001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/803133002/80001

[rune, 2016-03-21 11:32:11]

Description was changed from

==========
Use invalidation sets for fullscreen pseudos.

Schedule invalidation sets on elements changing state for
:-webkit-full-screen and:-webkit-full-screen-ancestor.

Add the UA style for fullscreen at the same time as html.css to have
invalidation sets available first time going fullscreen.

BUG=442239
==========

to

==========
Use invalidation sets for fullscreen pseudos.

Schedule invalidation sets on elements changing state for
:-webkit-full-screen and:-webkit-full-screen-ancestor.

Lazily load the UA style for fullscreen, but before we enter fullscreen
the first time to have the invalidation sets available for style
invalidation

BUG=442239
==========

[rune, 2016-03-21 11:34:15]

On 2016/03/18 05:05:58, esprehn wrote:

> In this case if we're always going to add the full screen sheet, it should
just
> be in html.css, but really we should keep it separate and only load it when
> we're in full screen mode.

I've done the fullscreen.css parsing lazy again, but moved it into a separate
method called before we need the invalidation sets for going fullscreen.

ptal.

[commit-bot: I haz the power, 2016-03-21 12:25:37]

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

[commit-bot: I haz the power, 2016-03-21 12:25:38]

Dry run: This issue passed the CQ dry run.

[esprehn, 2016-03-21 21:01:19]

lgtm

[rune, 2016-03-21 21:18:39]

The CQ bit was checked by rune@opera.com

[rune, 2016-03-21 21:18:39]

The patchset sent to the CQ was uploaded after l-g-t-m from timloh@chromium.org
Link to the patchset: https://codereview.chromium.org/803133002/#ps80001 (title:
"Load fullscreen.css on demand")

[commit-bot: I haz the power, 2016-03-21 21:19:10]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/803133002/80001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/803133002/80001

[commit-bot: I haz the power, 2016-03-21 21:57:11]

Description was changed from

==========
Use invalidation sets for fullscreen pseudos.

Schedule invalidation sets on elements changing state for
:-webkit-full-screen and:-webkit-full-screen-ancestor.

Lazily load the UA style for fullscreen, but before we enter fullscreen
the first time to have the invalidation sets available for style
invalidation

BUG=442239
==========

to

==========
Use invalidation sets for fullscreen pseudos.

Schedule invalidation sets on elements changing state for
:-webkit-full-screen and:-webkit-full-screen-ancestor.

Lazily load the UA style for fullscreen, but before we enter fullscreen
the first time to have the invalidation sets available for style
invalidation

BUG=442239
==========

[commit-bot: I haz the power, 2016-03-21 21:57:12]

Committed patchset #5 (id:80001)

[commit-bot: I haz the power, 2016-03-21 21:59:11]

Description was changed from

==========
Use invalidation sets for fullscreen pseudos.

Schedule invalidation sets on elements changing state for
:-webkit-full-screen and:-webkit-full-screen-ancestor.

Lazily load the UA style for fullscreen, but before we enter fullscreen
the first time to have the invalidation sets available for style
invalidation

BUG=442239
==========

to

==========
Use invalidation sets for fullscreen pseudos.

Schedule invalidation sets on elements changing state for
:-webkit-full-screen and:-webkit-full-screen-ancestor.

Lazily load the UA style for fullscreen, but before we enter fullscreen
the first time to have the invalidation sets available for style
invalidation

BUG=442239

Committed: https://crrev.com/5a2027442fee79bb42fe601a608d3d5459991bc8
Cr-Commit-Position: refs/heads/master@{#382405}
==========

[commit-bot: I haz the power, 2016-03-21 21:59:12]

Patchset 5 (id:??) landed as
https://crrev.com/5a2027442fee79bb42fe601a608d3d5459991bc8
Cr-Commit-Position: refs/heads/master@{#382405}