Modify LayoutTests to prep for compositing solid background scollers

Modify LayoutTests to prep for compositing solid background scollers

Tests of scrolling behavior should allow some visual mechanism for
detecting the scroll position. Some tests have been updated to make this
so.

Other tests require that a scrolling layer not be composited. Upcoming
changes will cause solid background color scrolling content to be
composited on all displays, so such tests must be modified slightly to
prevent the scrolling layer from promotion (by making the background
partially transparent).

And one test is explicitly testing the preferCompositingToLCDText flag,
so we should rename it and adjust the background so that the test
continues to test what it claims.

R=flackr
BUG=381840
Committed: https://crrev.com/a0f3f2e7a62356e4c1a44b769e46be984366ecad
Cr-Commit-Position: refs/heads/master@{#409791}

[Stephen Chennney, 2016-08-01 18:04:58]

Patchset #1 (id:1) has been deleted

[Stephen Chennney, 2016-08-01 19:32:33]

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

[Stephen Chennney, 2016-08-01 19:32:45]

PTAL.

[commit-bot: I haz the power, 2016-08-01 19:32:57]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-08-01 20:35:06]

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

[commit-bot: I haz the power, 2016-08-01 20:35:07]

Dry run: This issue passed the CQ dry run.

[flackr, 2016-08-02 15:02:44]

https://codereview.chromium.org/2196363002/diff/40001/third_party/WebKit/Layo...
File third_party/WebKit/LayoutTests/TestExpectations (right):

https://codereview.chromium.org/2196363002/diff/40001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/TestExpectations:1364: crbug.com/381840
fast/borders/overflow-hidden-border-radius-force-backing-store.html [
NeedsRebaseline ]
For a small change like this, wouldn't it make sense to update the expectation
manually and verify the diff?

https://codereview.chromium.org/2196363002/diff/40001/third_party/WebKit/Layo...
File
third_party/WebKit/LayoutTests/compositing/overflow/clear-scroll-parent.html
(right):

https://codereview.chromium.org/2196363002/diff/40001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/compositing/overflow/clear-scroll-parent.html:11:
background-color: rgba(0,0,255,0.5);
It looks like .container had a transparent background before, so we don't need
to give it a background do we?

https://codereview.chromium.org/2196363002/diff/40001/third_party/WebKit/Sour...
File
third_party/WebKit/Source/web/tests/data/composited_selection_bounds_editable.html
(right):

https://codereview.chromium.org/2196363002/diff/40001/third_party/WebKit/Sour...
third_party/WebKit/Source/web/tests/data/composited_selection_bounds_editable.html:10:
background-color: transparent; /* Prevent scrolling layer promotion to it's own
layer */
nit: For clarity, can we mention this is for the textarea or move this
background to be specifically for textarea elements?

[Stephen Chennney, 2016-08-02 19:00:01]

New patch any second now.

https://codereview.chromium.org/2196363002/diff/40001/third_party/WebKit/Layo...
File third_party/WebKit/LayoutTests/TestExpectations (right):

https://codereview.chromium.org/2196363002/diff/40001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/TestExpectations:1364: crbug.com/381840
fast/borders/overflow-hidden-border-radius-force-backing-store.html [
NeedsRebaseline ]
On 2016/08/02 15:02:44, flackr wrote:
> For a small change like this, wouldn't it make sense to update the expectation
> manually and verify the diff?

Done.

https://codereview.chromium.org/2196363002/diff/40001/third_party/WebKit/Layo...
File
third_party/WebKit/LayoutTests/compositing/overflow/clear-scroll-parent.html
(right):

https://codereview.chromium.org/2196363002/diff/40001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/compositing/overflow/clear-scroll-parent.html:11:
background-color: rgba(0,0,255,0.5);
On 2016/08/02 15:02:44, flackr wrote:
> It looks like .container had a transparent background before, so we don't need
> to give it a background do we?

I added this to make sure that we also paint the container background. I'll make
it green though.

https://codereview.chromium.org/2196363002/diff/40001/third_party/WebKit/Sour...
File
third_party/WebKit/Source/web/tests/data/composited_selection_bounds_editable.html
(right):

https://codereview.chromium.org/2196363002/diff/40001/third_party/WebKit/Sour...
third_party/WebKit/Source/web/tests/data/composited_selection_bounds_editable.html:10:
background-color: transparent; /* Prevent scrolling layer promotion to it's own
layer */
On 2016/08/02 15:02:44, flackr wrote:
> nit: For clarity, can we mention this is for the textarea or move this
> background to be specifically for textarea elements?

Done.

[Stephen Chennney, 2016-08-02 19:00:59]

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

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

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-08-02 21:16:20]

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

[commit-bot: I haz the power, 2016-08-02 21:16:21]

Dry run: This issue passed the CQ dry run.

[flackr, 2016-08-03 20:00:01]

https://codereview.chromium.org/2196363002/diff/60001/third_party/WebKit/Layo...
File
third_party/WebKit/LayoutTests/compositing/layer-creation/fixed-position-in-fixed-overflow.html
(right):

https://codereview.chromium.org/2196363002/diff/60001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/compositing/layer-creation/fixed-position-in-fixed-overflow.html:21:
background-color: blue;
Should background color still be here? The image will completely cover it up
right?

https://codereview.chromium.org/2196363002/diff/60001/third_party/WebKit/Sour...
File
third_party/WebKit/Source/web/tests/data/composited_selection_bounds_editable.html
(right):

https://codereview.chromium.org/2196363002/diff/60001/third_party/WebKit/Sour...
third_party/WebKit/Source/web/tests/data/composited_selection_bounds_editable.html:11:
.text {
I think you meant #text

[Stephen Chennney, 2016-08-03 20:21:15]

Explained 1 and fixed 1. Thanks.

https://codereview.chromium.org/2196363002/diff/60001/third_party/WebKit/Layo...
File
third_party/WebKit/LayoutTests/compositing/layer-creation/fixed-position-in-fixed-overflow.html
(right):

https://codereview.chromium.org/2196363002/diff/60001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/compositing/layer-creation/fixed-position-in-fixed-overflow.html:21:
background-color: blue;
On 2016/08/03 20:00:01, flackr wrote:
> Should background color still be here? The image will completely cover it up
> right?

Yes, it will cover it up, but we only use background color in determining
composited scrolling, so I want to leave it in.

A gradient with opaque colors completely fills a space if there is no
background-image-size given or other things to reduce coverage, but we don't
check for that right now.

https://codereview.chromium.org/2196363002/diff/60001/third_party/WebKit/Sour...
File
third_party/WebKit/Source/web/tests/data/composited_selection_bounds_editable.html
(right):

https://codereview.chromium.org/2196363002/diff/60001/third_party/WebKit/Sour...
third_party/WebKit/Source/web/tests/data/composited_selection_bounds_editable.html:11:
.text {
On 2016/08/03 20:00:01, flackr wrote:
> I think you meant #text

Yes. Done.

[flackr, 2016-08-03 20:34:14]

https://codereview.chromium.org/2196363002/diff/60001/third_party/WebKit/Layo...
File
third_party/WebKit/LayoutTests/compositing/layer-creation/fixed-position-in-fixed-overflow.html
(right):

https://codereview.chromium.org/2196363002/diff/60001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/compositing/layer-creation/fixed-position-in-fixed-overflow.html:21:
background-color: blue;
On 2016/08/03 at 20:21:15, Stephen Chennney wrote:
> On 2016/08/03 20:00:01, flackr wrote:
> > Should background color still be here? The image will completely cover it up
> > right?
> 
> Yes, it will cover it up, but we only use background color in determining
composited scrolling, so I want to leave it in.
> 
> A gradient with opaque colors completely fills a space if there is no
background-image-size given or other things to reduce coverage, but we don't
check for that right now.

Perhaps I am missing something. .blue is not a scroller, and we won't yet check
whether the entire scrolling contents has opaque backgrounds afaik. Also, why do
we want this to become a composited scroller if it's not today - i.e. keeping it
not composited would preserve the current test result. I thought the gradient
was just to be able to visually tell where you were in the scroll.

[Stephen Chennney, 2016-08-03 20:44:26]

https://codereview.chromium.org/2196363002/diff/60001/third_party/WebKit/Layo...
File
third_party/WebKit/LayoutTests/compositing/layer-creation/fixed-position-in-fixed-overflow.html
(right):

https://codereview.chromium.org/2196363002/diff/60001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/compositing/layer-creation/fixed-position-in-fixed-overflow.html:21:
background-color: blue;
On 2016/08/03 20:34:14, flackr wrote:
> On 2016/08/03 at 20:21:15, Stephen Chennney wrote:
> > On 2016/08/03 20:00:01, flackr wrote:
> > > Should background color still be here? The image will completely cover it
up
> > > right?
> > 
> > Yes, it will cover it up, but we only use background color in determining
> composited scrolling, so I want to leave it in.
> > 
> > A gradient with opaque colors completely fills a space if there is no
> background-image-size given or other things to reduce coverage, but we don't
> check for that right now.
> 
> Perhaps I am missing something. .blue is not a scroller, and we won't yet
check
> whether the entire scrolling contents has opaque backgrounds afaik. Also, why
do
> we want this to become a composited scroller if it's not today - i.e. keeping
it
> not composited would preserve the current test result. I thought the gradient
> was just to be able to visually tell where you were in the scroll.

.blue is a scroller if the window size is less than 1000px, as it is in layout
tests. My interest was in manual testing the correctness of things when it is
composited, but in hindsight that's inconsistent unless I also change the
.overflow height to be something explicitly less than 1000px.

And I want it to be composited so that we're testing the user's experience (or
will be).

I would go with fixing the size so that the test works manually in any window
size, but I'd also be satisfied just entirely reverting this change. Do you have
any preference?

[Stephen Chennney, 2016-08-03 21:05:15]

On 2016/08/03 20:44:26, Stephen Chennney wrote:
>
https://codereview.chromium.org/2196363002/diff/60001/third_party/WebKit/Layo...
> File
>
third_party/WebKit/LayoutTests/compositing/layer-creation/fixed-position-in-fixed-overflow.html
> (right):
> 
>
https://codereview.chromium.org/2196363002/diff/60001/third_party/WebKit/Layo...
>
third_party/WebKit/LayoutTests/compositing/layer-creation/fixed-position-in-fixed-overflow.html:21:
> background-color: blue;
> On 2016/08/03 20:34:14, flackr wrote:
> > On 2016/08/03 at 20:21:15, Stephen Chennney wrote:
> > > On 2016/08/03 20:00:01, flackr wrote:
> > > > Should background color still be here? The image will completely cover
it
> up
> > > > right?
> > > 
> > > Yes, it will cover it up, but we only use background color in determining
> > composited scrolling, so I want to leave it in.
> > > 
> > > A gradient with opaque colors completely fills a space if there is no
> > background-image-size given or other things to reduce coverage, but we don't
> > check for that right now.
> > 
> > Perhaps I am missing something. .blue is not a scroller, and we won't yet
> check
> > whether the entire scrolling contents has opaque backgrounds afaik. Also,
why
> do
> > we want this to become a composited scroller if it's not today - i.e.
keeping
> it
> > not composited would preserve the current test result. I thought the
gradient
> > was just to be able to visually tell where you were in the scroll.
> 
> .blue is a scroller if the window size is less than 1000px, as it is in layout
> tests. My interest was in manual testing the correctness of things when it is
> composited, but in hindsight that's inconsistent unless I also change the
> .overflow height to be something explicitly less than 1000px.
> 
> And I want it to be composited so that we're testing the user's experience (or
> will be).
> 
> I would go with fixing the size so that the test works manually in any window
> size, but I'd also be satisfied just entirely reverting this change. Do you
have
> any preference?

That all makes no sense, because we won't be compositing anyway. Change to the
fixed-position-in-fixed-overflow.html test has been reverted.

[flackr, 2016-08-03 21:11:43]

lgtm with nits

https://codereview.chromium.org/2196363002/diff/100001/third_party/WebKit/Lay...
File
third_party/WebKit/LayoutTests/compositing/layer-creation/fixed-position-nonscrollable-body-mismatch-containers.html
(right):

https://codereview.chromium.org/2196363002/diff/100001/third_party/WebKit/Lay...
third_party/WebKit/LayoutTests/compositing/layer-creation/fixed-position-nonscrollable-body-mismatch-containers.html:70:
<div style="height: 800px; background-image: linear-gradient(to
bottom,red,orange,yellow, green, blue,indigo,violet);"></div>
Do you still want the gradient background in the other non-composited scrollers
like this one?

https://codereview.chromium.org/2196363002/diff/100001/third_party/WebKit/Lay...
File
third_party/WebKit/LayoutTests/fast/borders/overflow-hidden-border-radius-force-backing-store.html
(right):

https://codereview.chromium.org/2196363002/diff/100001/third_party/WebKit/Lay...
third_party/WebKit/LayoutTests/fast/borders/overflow-hidden-border-radius-force-backing-store.html:26:
background-image: linear-gradient(to bottom,red,orange,yellow, green,
blue,indigo,violet);
and this one?

[Stephen Chennney, 2016-08-04 14:24:11]

On 2016/08/03 21:11:43, flackr wrote:
> lgtm with nits
> 
>
https://codereview.chromium.org/2196363002/diff/100001/third_party/WebKit/Lay...
> File
>
third_party/WebKit/LayoutTests/compositing/layer-creation/fixed-position-nonscrollable-body-mismatch-containers.html
> (right):
> 
>
https://codereview.chromium.org/2196363002/diff/100001/third_party/WebKit/Lay...
>
third_party/WebKit/LayoutTests/compositing/layer-creation/fixed-position-nonscrollable-body-mismatch-containers.html:70:
> <div style="height: 800px; background-image: linear-gradient(to
> bottom,red,orange,yellow, green, blue,indigo,violet);"></div>
> Do you still want the gradient background in the other non-composited
scrollers
> like this one?
> 
>
https://codereview.chromium.org/2196363002/diff/100001/third_party/WebKit/Lay...
> File
>
third_party/WebKit/LayoutTests/fast/borders/overflow-hidden-border-radius-force-backing-store.html
> (right):
> 
>
https://codereview.chromium.org/2196363002/diff/100001/third_party/WebKit/Lay...
>
third_party/WebKit/LayoutTests/fast/borders/overflow-hidden-border-radius-force-backing-store.html:26:
> background-image: linear-gradient(to bottom,red,orange,yellow, green,
> blue,indigo,violet);
> and this one?

Yes. I do want to be able to verify manually that they still scroll correctly,
regardless of their compositing status.

[Stephen Chennney, 2016-08-04 14:24:17]

The CQ bit was checked by schenney@chromium.org

[commit-bot: I haz the power, 2016-08-04 14:24:41]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-08-04 15:29:17]

Committed patchset #5 (id:100001)

[commit-bot: I haz the power, 2016-08-04 15:31:03]

Description was changed from

==========
Modify LayoutTests to prep for compositing solid background scollers

Tests of scrolling behavior should allow some visual mechanism for
detecting the scroll position. Some tests have been updated to make this
so.

Other tests require that a scrolling layer not be composited. Upcoming
changes will cause solid background color scrolling content to be
composited on all displays, so such tests must be modified slightly to
prevent the scrolling layer from promotion (by making the background
partially transparent).

And one test is explicitly testing the preferCompositingToLCDText flag,
so we should rename it and adjust the background so that the test
continues to test what it claims.

R=flackr
BUG=381840
==========

to

==========
Modify LayoutTests to prep for compositing solid background scollers

Tests of scrolling behavior should allow some visual mechanism for
detecting the scroll position. Some tests have been updated to make this
so.

Other tests require that a scrolling layer not be composited. Upcoming
changes will cause solid background color scrolling content to be
composited on all displays, so such tests must be modified slightly to
prevent the scrolling layer from promotion (by making the background
partially transparent).

And one test is explicitly testing the preferCompositingToLCDText flag,
so we should rename it and adjust the background so that the test
continues to test what it claims.

R=flackr
BUG=381840

Committed: https://crrev.com/a0f3f2e7a62356e4c1a44b769e46be984366ecad
Cr-Commit-Position: refs/heads/master@{#409791}
==========

[commit-bot: I haz the power, 2016-08-04 15:31:04]

Patchset 5 (id:??) landed as
https://crrev.com/a0f3f2e7a62356e4c1a44b769e46be984366ecad
Cr-Commit-Position: refs/heads/master@{#409791}