Preserve fractional scroll offset for JS scrolling API

Preserve fractional scroll offset for JS scrolling API

This patch makes that the fractional scroll offset set
by JS scrolling API: div.scrollTop, div.scrollLeft,
window.scroll, window.scrollTo and window.scrollBy do not
get truncated when passing through blink.

BUG=414283
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=183486

[Yufeng Shen (Slow to review), 2014-10-02 21:08:16]

miletus@chromium.org changed reviewers:
+ eae@chromium.org, rbyers@chromium.org

[Rick Byers, 2014-10-02 21:44:06]

Great, thanks!

I'm sure we're going to hit some subtle issues with partial conversion or
numerics problems, but I don't know any better way to proceed incrementally than
to try and fix the things we see break.

Minor nit: can you change the subject/description to make it clear this CL
"preserves" fractional scroll offsets?  The JS APIs already "support" fractional
scroll offsets - it's just that offsets get truncated to the nearest DIP (so the
fractions are only preserved for the special case of browser zoom).

eae@ do you think preserving the doubles in blink and relying on cc to
pixel-snap to physical pixels is OK ?  Or will we need to add some physical
pixel snapping logic in blink somewhere?

https://codereview.chromium.org/610423004/diff/20001/LayoutTests/fast/scrolli...
File LayoutTests/fast/scrolling/scrolling-apis-subpixel.html (right):

https://codereview.chromium.org/610423004/diff/20001/LayoutTests/fast/scrolli...
LayoutTests/fast/scrolling/scrolling-apis-subpixel.html:31: var floatPrecision =
0.01;
How close is this in practice?  Eg. would 0.0001 also work?

I'm worried that there may be web page bugs if we're off by a non-trivial
amount.  I believe we have some cases where we convert from double to LayoutUnit
and back to double again (which could be responsible for differences of ~0.01) -
those probably need to be considered bugs and fixed to consistently use double.

https://codereview.chromium.org/610423004/diff/20001/Source/core/html/ImageDo...
File Source/core/html/ImageDocument.cpp (right):

https://codereview.chromium.org/610423004/diff/20001/Source/core/html/ImageDo...
Source/core/html/ImageDocument.cpp:269: double scrollX = x / scale -
(float)frame()->view()->width() / 2;
use (double) instead of (float) to prevent loss of precision in the subtraction
(and/or make 'scale' a double?).

https://codereview.chromium.org/610423004/diff/20001/Source/core/rendering/Re...
File Source/core/rendering/RenderBox.cpp (right):

https://codereview.chromium.org/610423004/diff/20001/Source/core/rendering/Re...
Source/core/rendering/RenderBox.cpp:471: xOffset = std::max(0.0,
std::min(static_cast<double>(frameView->contentsWidth()), xOffset));
Do you really need this static_cast here?  I would have expected implicit
conversion to work.  does std::min<double>(a,b) work instead of adding a cast?

https://codereview.chromium.org/610423004/diff/20001/Source/core/rendering/Re...
Source/core/rendering/RenderBox.cpp:860: return
flooredIntSize(layer()->scrollableArea()->scrollOffset());
We need a FIXME here, right?

https://codereview.chromium.org/610423004/diff/20001/Source/core/rendering/Re...
File Source/core/rendering/RenderLayerScrollableArea.cpp (right):

https://codereview.chromium.org/610423004/diff/20001/Source/core/rendering/Re...
Source/core/rendering/RenderLayerScrollableArea.cpp:129:
toElement(node)->setSavedLayerScrollOffset(flooredIntSize(m_scrollOffset));
What's the implication of flooring here?  Do we need a FIXME?

https://codereview.chromium.org/610423004/diff/20001/Source/core/rendering/Re...
Source/core/rendering/RenderLayerScrollableArea.cpp:350: void
RenderLayerScrollableArea::setScrollOffset(const DoublePoint& newScrollOffset)
Didn't you guys agree that you'd drive to move towards using sizes for "offsets"
and points only for "positions"?  I.e. the callers should be converting to
DoubleSize here if necessary, not the implementation of this function, right?

https://codereview.chromium.org/610423004/diff/20001/Source/core/rendering/Re...
Source/core/rendering/RenderLayerScrollableArea.cpp:580:
scrollToOffsetWithoutAnimation(toFloatPoint(-origin + newScrollOffset));
scrollToOffsetWithoutAnimation should be updated (here or in a future CL) to
take a DoublePoint instead of a FloatPoint, right?

https://codereview.chromium.org/610423004/diff/20001/Source/core/rendering/co...
File Source/core/rendering/compositing/CompositedLayerMapping.cpp (right):

https://codereview.chromium.org/610423004/diff/20001/Source/core/rendering/co...
Source/core/rendering/compositing/CompositedLayerMapping.cpp:896: IntSize
adjustedScrollOffset =
flooredIntSize(m_owningLayer.scrollableArea()->adjustedScrollOffset());
Add FIXME?

https://codereview.chromium.org/610423004/diff/20001/Source/core/rendering/co...
Source/core/rendering/compositing/CompositedLayerMapping.cpp:1153: const
IntPoint position = blockSelectionGapsBounds.location() +
flooredIntSize(m_owningLayer.scrollableArea()->adjustedScrollOffset());
add FIXME?

https://codereview.chromium.org/610423004/diff/20001/Source/platform/geometry...
File Source/platform/geometry/DoubleSize.h (right):

https://codereview.chromium.org/610423004/diff/20001/Source/platform/geometry...
Source/platform/geometry/DoubleSize.h:41: inline DoubleSize operator+(const
DoubleSize& a, const IntSize& b)
are these still necessary if you have the implicit conversion operator above?  I
don't see any such methods on FloatSize.

[eae, 2014-10-02 21:52:36]

On 2014/10/02 21:44:06, Rick Byers wrote:
> Great, thanks!
> 
> I'm sure we're going to hit some subtle issues with partial conversion or
> numerics problems, but I don't know any better way to proceed incrementally
than
> to try and fix the things we see break.
> 
> Minor nit: can you change the subject/description to make it clear this CL
> "preserves" fractional scroll offsets?  The JS APIs already "support"
fractional
> scroll offsets - it's just that offsets get truncated to the nearest DIP (so
the
> fractions are only preserved for the special case of browser zoom).
> 
> eae@ do you think preserving the doubles in blink and relying on cc to
> pixel-snap to physical pixels is OK ?  Or will we need to add some physical
> pixel snapping logic in blink somewhere?

We really should do the snapping in blink, it is the only place that has enough
information to do that. If not we can end up with cases where we show a
scrollbar for content that can't scroll (i.e. 1009.32px content in a 1009.20px
container). We have a lot of existing snapping logic for this, would be great if
we could reuse some of it.

[Yufeng Shen (Slow to review), 2014-10-02 23:34:35]

https://codereview.chromium.org/610423004/diff/20001/LayoutTests/fast/scrolli...
File LayoutTests/fast/scrolling/scrolling-apis-subpixel.html (right):

https://codereview.chromium.org/610423004/diff/20001/LayoutTests/fast/scrolli...
LayoutTests/fast/scrolling/scrolling-apis-subpixel.html:31: var floatPrecision =
0.01;
On 2014/10/02 21:44:05, Rick Byers wrote:
> How close is this in practice?  Eg. would 0.0001 also work?
> 
> I'm worried that there may be web page bugs if we're off by a non-trivial
> amount.  I believe we have some cases where we convert from double to
LayoutUnit
> and back to double again (which could be responsible for differences of ~0.01)
-
> those probably need to be considered bugs and fixed to consistently use
double.

Yes. Element::setScrollLeft() has a conversion from float to LayoutUnit
LayoutUnit::fromFloatRound() which adds about ~0.003. So that's why I have to
put the precision so large, e.g. if you call setScrollLeft(4.2) and
read it back, it would be something like 4.203.

I am leaving a FIXME here to make the floatPrecision smaller.

https://codereview.chromium.org/610423004/diff/20001/Source/core/html/ImageDo...
File Source/core/html/ImageDocument.cpp (right):

https://codereview.chromium.org/610423004/diff/20001/Source/core/html/ImageDo...
Source/core/html/ImageDocument.cpp:269: double scrollX = x / scale -
(float)frame()->view()->width() / 2;
On 2014/10/02 21:44:05, Rick Byers wrote:
> use (double) instead of (float) to prevent loss of precision in the
subtraction
> (and/or make 'scale' a double?).

Done.

https://codereview.chromium.org/610423004/diff/20001/Source/core/rendering/Re...
File Source/core/rendering/RenderBox.cpp (right):

https://codereview.chromium.org/610423004/diff/20001/Source/core/rendering/Re...
Source/core/rendering/RenderBox.cpp:471: xOffset = std::max(0.0,
std::min(static_cast<double>(frameView->contentsWidth()), xOffset));
On 2014/10/02 21:44:06, Rick Byers wrote:
> Do you really need this static_cast here?  I would have expected implicit
> conversion to work.  does std::min<double>(a,b) work instead of adding a cast?

min<double>(a,b) works but not implicit conversion.

https://codereview.chromium.org/610423004/diff/20001/Source/core/rendering/Re...
Source/core/rendering/RenderBox.cpp:860: return
flooredIntSize(layer()->scrollableArea()->scrollOffset());
On 2014/10/02 21:44:05, Rick Byers wrote:
> We need a FIXME here, right?

Done.

https://codereview.chromium.org/610423004/diff/20001/Source/core/rendering/Re...
File Source/core/rendering/RenderLayerScrollableArea.cpp (right):

https://codereview.chromium.org/610423004/diff/20001/Source/core/rendering/Re...
Source/core/rendering/RenderLayerScrollableArea.cpp:129:
toElement(node)->setSavedLayerScrollOffset(flooredIntSize(m_scrollOffset));
On 2014/10/02 21:44:06, Rick Byers wrote:
> What's the implication of flooring here?  Do we need a FIXME?

No documentation about what this scrolloffet save/restore is about.
I will just add a FIXME here.

https://codereview.chromium.org/610423004/diff/20001/Source/core/rendering/Re...
Source/core/rendering/RenderLayerScrollableArea.cpp:350: void
RenderLayerScrollableArea::setScrollOffset(const DoublePoint& newScrollOffset)
On 2014/10/02 21:44:06, Rick Byers wrote:
> Didn't you guys agree that you'd drive to move towards using sizes for
"offsets"
> and points only for "positions"?  I.e. the callers should be converting to
> DoubleSize here if necessary, not the implementation of this function, right?

Ideally yes. Just that ScrollableArea::setScrollOffset is overriden by so many
subclasses that changing its signature and calling sites will touch too many
files. I added a FIXME in  ScrollableArea::setScrollOffset to make the
conversion
in a separate CL.

https://codereview.chromium.org/610423004/diff/20001/Source/core/rendering/Re...
Source/core/rendering/RenderLayerScrollableArea.cpp:580:
scrollToOffsetWithoutAnimation(toFloatPoint(-origin + newScrollOffset));
On 2014/10/02 21:44:06, Rick Byers wrote:
> scrollToOffsetWithoutAnimation should be updated (here or in a future CL) to
> take a DoublePoint instead of a FloatPoint, right?

FIXME added.

https://codereview.chromium.org/610423004/diff/20001/Source/core/rendering/co...
File Source/core/rendering/compositing/CompositedLayerMapping.cpp (right):

https://codereview.chromium.org/610423004/diff/20001/Source/core/rendering/co...
Source/core/rendering/compositing/CompositedLayerMapping.cpp:896: IntSize
adjustedScrollOffset =
flooredIntSize(m_owningLayer.scrollableArea()->adjustedScrollOffset());
On 2014/10/02 21:44:06, Rick Byers wrote:
> Add FIXME?

Done.

https://codereview.chromium.org/610423004/diff/20001/Source/core/rendering/co...
Source/core/rendering/compositing/CompositedLayerMapping.cpp:1153: const
IntPoint position = blockSelectionGapsBounds.location() +
flooredIntSize(m_owningLayer.scrollableArea()->adjustedScrollOffset());
On 2014/10/02 21:44:06, Rick Byers wrote:
> add FIXME?

Done.

https://codereview.chromium.org/610423004/diff/20001/Source/platform/geometry...
File Source/platform/geometry/DoubleSize.h (right):

https://codereview.chromium.org/610423004/diff/20001/Source/platform/geometry...
Source/platform/geometry/DoubleSize.h:41: inline DoubleSize operator+(const
DoubleSize& a, const IntSize& b)
On 2014/10/02 21:44:06, Rick Byers wrote:
> are these still necessary if you have the implicit conversion operator above? 
I
> don't see any such methods on FloatSize.

removed.

[Rick Byers, 2014-10-03 16:33:55]

On 2014/10/02 21:52:36, eae wrote:
> On 2014/10/02 21:44:06, Rick Byers wrote:
> > Great, thanks!
> > 
> > I'm sure we're going to hit some subtle issues with partial conversion or
> > numerics problems, but I don't know any better way to proceed incrementally
> than
> > to try and fix the things we see break.
> > 
> > Minor nit: can you change the subject/description to make it clear this CL
> > "preserves" fractional scroll offsets?  The JS APIs already "support"
> fractional
> > scroll offsets - it's just that offsets get truncated to the nearest DIP (so
> the
> > fractions are only preserved for the special case of browser zoom).
> > 
> > eae@ do you think preserving the doubles in blink and relying on cc to
> > pixel-snap to physical pixels is OK ?  Or will we need to add some physical
> > pixel snapping logic in blink somewhere?
> 
> We really should do the snapping in blink, it is the only place that has
enough
> information to do that. If not we can end up with cases where we show a
> scrollbar for content that can't scroll (i.e. 1009.32px content in a 1009.20px
> container). We have a lot of existing snapping logic for this, would be great
if
> we could reuse some of it.

I don't understand the scrollbar scenario.  Content and container height are
still integers - it's only the scroll offsets we're talking about here.  So this
wouldn't affect the appearance or absence of scrollbars.  But it could mean, for
example that scrollTop != scrollHeight but the user can't observe any further
scrolling.

Does blink have all the information necessary to do pixel snapping when pinch
zoomed?  Ideally when pinch-zoomed in we'd still be able to scroll the page in
physical pixel increments (as we can today on the impl thread).

[Rick Byers, 2014-10-03 16:42:46]

LGTM

I think it's fine to leave the pixel snapping question for a future CL.

https://codereview.chromium.org/610423004/diff/20001/LayoutTests/fast/scrolli...
File LayoutTests/fast/scrolling/scrolling-apis-subpixel.html (right):

https://codereview.chromium.org/610423004/diff/20001/LayoutTests/fast/scrolli...
LayoutTests/fast/scrolling/scrolling-apis-subpixel.html:31: var floatPrecision =
0.01;
On 2014/10/02 23:34:34, Yufeng Shen wrote:
> On 2014/10/02 21:44:05, Rick Byers wrote:
> > How close is this in practice?  Eg. would 0.0001 also work?
> > 
> > I'm worried that there may be web page bugs if we're off by a non-trivial
> > amount.  I believe we have some cases where we convert from double to
> LayoutUnit
> > and back to double again (which could be responsible for differences of
~0.01)
> -
> > those probably need to be considered bugs and fixed to consistently use
> double.
> 
> Yes. Element::setScrollLeft() has a conversion from float to LayoutUnit
> LayoutUnit::fromFloatRound() which adds about ~0.003. So that's why I have to
> put the precision so large, e.g. if you call setScrollLeft(4.2) and
> read it back, it would be something like 4.203.
> 
> I am leaving a FIXME here to make the floatPrecision smaller.

Ok.  Ideally we'll even be able to go back to shouldBeEqualToNumber (if all
we're doing is reading the exact double we stored, there should be no numerics
issues).

https://codereview.chromium.org/610423004/diff/20001/Source/core/rendering/Re...
File Source/core/rendering/RenderLayerScrollableArea.cpp (right):

https://codereview.chromium.org/610423004/diff/20001/Source/core/rendering/Re...
Source/core/rendering/RenderLayerScrollableArea.cpp:350: void
RenderLayerScrollableArea::setScrollOffset(const DoublePoint& newScrollOffset)
On 2014/10/02 23:34:34, Yufeng Shen wrote:
> On 2014/10/02 21:44:06, Rick Byers wrote:
> > Didn't you guys agree that you'd drive to move towards using sizes for
> "offsets"
> > and points only for "positions"?  I.e. the callers should be converting to
> > DoubleSize here if necessary, not the implementation of this function,
right?
> 
> Ideally yes. Just that ScrollableArea::setScrollOffset is overriden by so many
> subclasses that changing its signature and calling sites will touch too many
> files. I added a FIXME in  ScrollableArea::setScrollOffset to make the
> conversion
> in a separate CL.

Sounds good, thanks.

[Yufeng Shen (Slow to review), 2014-10-07 20:14:05]

Emil, can you also give a review ?

[eae, 2014-10-08 16:19:36]

https://codereview.chromium.org/610423004/diff/40001/Source/core/html/ImageDo...
File Source/core/html/ImageDocument.cpp (right):

https://codereview.chromium.org/610423004/diff/40001/Source/core/html/ImageDo...
Source/core/html/ImageDocument.cpp:269: double scrollX = x / scale -
(double)frame()->view()->width() / 2;
Use c++ style static_cast please

https://codereview.chromium.org/610423004/diff/40001/Source/core/rendering/Re...
File Source/core/rendering/RenderBox.cpp (right):

https://codereview.chromium.org/610423004/diff/40001/Source/core/rendering/Re...
Source/core/rendering/RenderBox.cpp:813: DoubleSize adjustedDelta =
adjustedScrollDelta(delta);
adjustedScrollDelta returns an IntSize.

[Yufeng Shen (Slow to review), 2014-10-08 18:12:08]

Emil, PTAL, thanks!

https://codereview.chromium.org/610423004/diff/40001/Source/core/html/ImageDo...
File Source/core/html/ImageDocument.cpp (right):

https://codereview.chromium.org/610423004/diff/40001/Source/core/html/ImageDo...
Source/core/html/ImageDocument.cpp:269: double scrollX = x / scale -
(double)frame()->view()->width() / 2;
On 2014/10/08 16:19:36, eae wrote:
> Use c++ style static_cast please

Done.

https://codereview.chromium.org/610423004/diff/40001/Source/core/rendering/Re...
File Source/core/rendering/RenderBox.cpp (right):

https://codereview.chromium.org/610423004/diff/40001/Source/core/rendering/Re...
Source/core/rendering/RenderBox.cpp:813: DoubleSize adjustedDelta =
adjustedScrollDelta(delta);
On 2014/10/08 16:19:36, eae wrote:
> adjustedScrollDelta returns an IntSize.

right.

[eae, 2014-10-08 21:10:05]

LGTM

[Yufeng Shen (Slow to review), 2014-10-08 21:20:16]

The CQ bit was checked by miletus@chromium.org

[commit-bot: I haz the power, 2014-10-08 21:21:39]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/610423004/120001

[commit-bot: I haz the power, 2014-10-09 03:11:09]

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

[commit-bot: I haz the power, 2014-10-09 03:11:10]

Exceeded time limit waiting for builds to trigger.

[Rick Byers, 2014-10-09 14:33:17]

The CQ bit was checked by rbyers@chromium.org

[commit-bot: I haz the power, 2014-10-09 14:34:05]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/610423004/120001

[commit-bot: I haz the power, 2014-10-09 14:34:24]

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

[commit-bot: I haz the power, 2014-10-09 14:34:25]

Failed to apply patch for LayoutTests/TestExpectations:
While running patch -p1 --forward --force --no-backup-if-mismatch;
  patching file LayoutTests/TestExpectations
  Hunk #2 FAILED at 1443.
  1 out of 2 hunks FAILED -- saving rejects to file
LayoutTests/TestExpectations.rej

Patch:       LayoutTests/TestExpectations
Index: LayoutTests/TestExpectations
diff --git a/LayoutTests/TestExpectations b/LayoutTests/TestExpectations
index
88972434ad7e7140a82ecfedb7a4c4c36141b796..c40b80a4cfcfe0b12a0632c7e8149b970ae265b9
100644
--- a/LayoutTests/TestExpectations
+++ b/LayoutTests/TestExpectations
@@ -84,7 +84,8 @@ crbug.com/338794 crbug.com/192172
compositing/rtl/rtl-iframe-fixed-overflow.html
 
 crbug.com/239811 crbug.com/229113
http/tests/history/redirect-js-location-assign-0-seconds.html [ Failure Pass ]
 
-crbug.com/231721 [ Win ] svg/custom/invisible-text-after-scrolling.xhtml [ Pass
Timeout ]
+# Commented out for rebaselining for crbug.com/414283.
+# crbug.com/231721 [ Win ] svg/custom/invisible-text-after-scrolling.xhtml [
Pass Timeout ]
 
 # Crashing after V8 roll to 3.28.43.1.
 crbug.com/398306 fast/events/before-unload-return-bad-value.html [ Crash
Timeout Pass ]
@@ -1442,3 +1443,55 @@ crbug.com/420008 [ Win ]
virtual/deferred/inspector/tracing [ Pass Failure ]
 crbug.com/420008 [ Win ] virtual/implsidepainting/inspector/tracing [ Pass
Failure ]
 
 crbug.com/421298
http/tests/security/subresourceIntegrity/subresource-integrity-style-blocked.html
[ Pass Failure ]
+
+crbug.com/414283 fast/overflow/scroll-nested-positioned-layer-in-overflow.html
[ NeedsRebaseline ]
+crbug.com/414283 fast/css/text-overflow-ellipsis-bidi.html [ NeedsRebaseline ]
+crbug.com/414283
ietestcenter/css3/bordersbackgrounds/background-attachment-local-scrolling.htm [
NeedsRebaseline ]
+crbug.com/414283 css3/images/cross-fade-overflow-position.html [
NeedsRebaseline ]
+crbug.com/414283 fast/css/text-overflow-ellipsis-text-align-justify.html [
NeedsRebaseline ]
+crbug.com/414283 media/track/track-cue-rendering-vertical.html [
NeedsRebaseline ]
+crbug.com/414283 fast/forms/listbox-scrollbar-incremental-load.html [
NeedsRebaseline ]
+crbug.com/414283 editing/input/reveal-caret-of-multiline-input.html [
NeedsRebaseline ]
+crbug.com/414283 fast/css/vertical-text-overflow-ellipsis-text-align-right.html
[ NeedsRebaseline ]
+crbug.com/414283 svg/custom/invisible-text-after-scrolling.xhtml [
NeedsRebaseline ]
+crbug.com/414283 fast/forms/select/menulist-appearance-basic.html [
NeedsRebaseline ]
+crbug.com/414283 fast/layers/scroll-rect-to-visible.html [ NeedsRebaseline ]
+crbug.com/414283 fast/forms/textarea-scrolled-type.html [ NeedsRebaseline ]
+crbug.com/414283 svg/custom/getscreenctm-in-scrollable-div-area.xhtml [
NeedsRebaseline ]
+crbug.com/414283 fast/overflow/paged-x-div-with-column-gap.html [
NeedsRebaseline ]
+crbug.com/414283 css3/flexbox/flexbox-baseline.html [ NeedsRebaseline ]
+crbug.com/414283 compositing/layers-inside-overflow-scroll.html [
NeedsRebaseline ]
+crbug.com/414283 fast/css/vertical-text-overflow-ellipsis-text-align-left.html
[ NeedsRebaseline ]
+crbug.com/414283 fast/repaint/block-selection-gap-stale-cache-2.html [
NeedsRebaseline ]
+crbug.com/414283 fast/overflow/scrollbar-position-update.html [ NeedsRebaseline
]
+crbug.com/414283
fast/css/text-overflow-ellipsis-block-with-border-and-padding.html [
NeedsRebaseline ]
+crbug.com/414283 fast/forms/listbox-hit-test-zoomed.html [ NeedsRebaseline ]
+crbug.com/414283 fast/overflow/image-selection-highlight.html [ NeedsRebaseline
]
+crbug.com/414283 compositing/overflow/scroll-ancestor-update.html [
NeedsRebaseline ]
+crbug.com/414283 fast/overflow/unreachable-overflow-rtl-bug.html [
NeedsRebaseline ]
+crbug.com/414283
virtual/prefer_compositing_to_lcd_text/compositing/overflow/overflow-scroll.html
[ NeedsRebaseline ]
+crbug.com/414283 fast/css/text-overflow-ellipsis.html [ NeedsRebaseline ]
+crbug.com/414283 fast/css/text-overflow-ellipsis-text-align-left.html [
NeedsRebaseline ]
+crbug.com/414283 compositing/self-painting-layers.html [ NeedsRebaseline ]
+crbug.com/414283 fast/forms/input-text-scroll-left-on-blur.html [
NeedsRebaseline ]
+crbug.com/414283 fast/overflow/overflow-rtl.html [ NeedsRebaseline ]
+crbug.com/414283
fast/css/vertical-text-overflow-ellipsis-text-align-center.html [
NeedsRebaseline ]
+crbug.com/414283 fast/overflow/overflow-text-hit-testing.html [ NeedsRebaseline
]
+crbug.com/414283 fast/css/text-overflow-ellipsis-text-align-center.html [
NeedsRebaseline ]
+crbug.com/414283 fast/overflow/008.html [ NeedsRebaseline ]
+crbug.com/414283 fast/css/text-overflow-ellipsis-text-align-right.html [
NeedsRebaseline ]
+crbug.com/414283 fast/forms/search-rtl.html [ NeedsRebaseline ]
+crbug.com/414283 compositing/overflow/overflow-scroll.html [ NeedsRebaseline ]
+crbug.com/414283 fast/css/text-overflow-ellipsis-strict.html [ NeedsRebaseline
]
+crbug.com/414283 svg/custom/getscreenctm-in-scrollable-div-area-nested.xhtml [
NeedsRebaseline ]
+crbug.com/414283 fast/table/edge-offsets.html [ NeedsRebaseline ]
+crbug.com/414283 fast/css/text-overflow-input.html [ NeedsRebaseline ]
+crbug.com/414283 fast/forms/select-initial-position.html [ NeedsRebaseline ]
+crbug.com/414283 fast/transforms/selection-bounds-in-transformed-view.html [
NeedsRebaseline ]
+crbug.com/414283 editing/input/caret-at-the-edge-of-contenteditable.html [
NeedsRebaseline ]
+crbug.com/414283 fast/forms/input-readonly-autoscroll.html [ NeedsRebaseline ]
+crbug.com/414283 fast/overflow/overflow-rtl-vertical.html [ NeedsRebaseline ]
+crbug.com/414283 editing/input/reveal-caret-of-multiline-contenteditable.html [
NeedsRebaseline ]
+crbug.com/414283 fast/overflow/scrollRevealButton.html [ NeedsRebaseline ]
+crbug.com/414283
fast/css/vertical-text-overflow-ellipsis-text-align-justify.html [
NeedsRebaseline ]
+crbug.com/414283
virtual/prefer_compositing_to_lcd_text/compositing/overflow/scroll-ancestor-update.html
[ NeedsRebaseline ]

[miletus1, 2014-10-09 16:31:34]

The CQ bit was checked by miletus@google.com

[commit-bot: I haz the power, 2014-10-09 16:32:40]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/610423004/140001

[commit-bot: I haz the power, 2014-10-09 16:32:52]

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

[commit-bot: I haz the power, 2014-10-09 16:32:54]

Failed to apply patch for LayoutTests/TestExpectations:
While running patch -p1 --forward --force --no-backup-if-mismatch;
  patching file LayoutTests/TestExpectations
  Hunk #2 FAILED at 1443.
  1 out of 2 hunks FAILED -- saving rejects to file
LayoutTests/TestExpectations.rej

Patch:       LayoutTests/TestExpectations
Index: LayoutTests/TestExpectations
diff --git a/LayoutTests/TestExpectations b/LayoutTests/TestExpectations
index
88972434ad7e7140a82ecfedb7a4c4c36141b796..0c71d3d91caac31f0ff5a3e9f29d1f491b1bf6b9
100644
--- a/LayoutTests/TestExpectations
+++ b/LayoutTests/TestExpectations
@@ -84,7 +84,8 @@ crbug.com/338794 crbug.com/192172
compositing/rtl/rtl-iframe-fixed-overflow.html
 
 crbug.com/239811 crbug.com/229113
http/tests/history/redirect-js-location-assign-0-seconds.html [ Failure Pass ]
 
-crbug.com/231721 [ Win ] svg/custom/invisible-text-after-scrolling.xhtml [ Pass
Timeout ]
+# Commented out for rebaselining for crbug.com/414283.
+# crbug.com/231721 [ Win ] svg/custom/invisible-text-after-scrolling.xhtml [
Pass Timeout ]
 
 # Crashing after V8 roll to 3.28.43.1.
 crbug.com/398306 fast/events/before-unload-return-bad-value.html [ Crash
Timeout Pass ]
@@ -1442,3 +1443,58 @@ crbug.com/420008 [ Win ]
virtual/deferred/inspector/tracing [ Pass Failure ]
 crbug.com/420008 [ Win ] virtual/implsidepainting/inspector/tracing [ Pass
Failure ]
 
 crbug.com/421298
http/tests/security/subresourceIntegrity/subresource-integrity-style-blocked.html
[ Pass Failure ]
+
+crbug.com/414283 fast/overflow/scroll-nested-positioned-layer-in-overflow.html
[ NeedsRebaseline ]
+crbug.com/414283 fast/css/text-overflow-ellipsis-bidi.html [ NeedsRebaseline ]
+crbug.com/414283
ietestcenter/css3/bordersbackgrounds/background-attachment-local-scrolling.htm [
NeedsRebaseline ]
+crbug.com/414283 css3/images/cross-fade-overflow-position.html [
NeedsRebaseline ]
+crbug.com/414283 fast/css/text-overflow-ellipsis-text-align-justify.html [
NeedsRebaseline ]
+crbug.com/414283 media/track/track-cue-rendering-vertical.html [
NeedsRebaseline ]
+crbug.com/414283 fast/forms/listbox-scrollbar-incremental-load.html [
NeedsRebaseline ]
+crbug.com/414283 editing/input/reveal-caret-of-multiline-input.html [
NeedsRebaseline ]
+crbug.com/414283 fast/css/vertical-text-overflow-ellipsis-text-align-right.html
[ NeedsRebaseline ]
+crbug.com/414283 svg/custom/invisible-text-after-scrolling.xhtml [
NeedsRebaseline ]
+crbug.com/414283 fast/forms/select/menulist-appearance-basic.html [
NeedsRebaseline ]
+crbug.com/414283 fast/layers/scroll-rect-to-visible.html [ NeedsRebaseline ]
+crbug.com/414283 fast/forms/textarea-scrolled-type.html [ NeedsRebaseline ]
+crbug.com/414283 svg/custom/getscreenctm-in-scrollable-div-area.xhtml [
NeedsRebaseline ]
+crbug.com/414283 fast/overflow/paged-x-div-with-column-gap.html [
NeedsRebaseline ]
+crbug.com/414283 css3/flexbox/flexbox-baseline.html [ NeedsRebaseline ]
+crbug.com/414283 compositing/layers-inside-overflow-scroll.html [
NeedsRebaseline ]
+crbug.com/414283 fast/css/vertical-text-overflow-ellipsis-text-align-left.html
[ NeedsRebaseline ]
+crbug.com/414283 fast/repaint/block-selection-gap-stale-cache-2.html [
NeedsRebaseline ]
+crbug.com/414283 fast/overflow/scrollbar-position-update.html [ NeedsRebaseline
]
+crbug.com/414283
fast/css/text-overflow-ellipsis-block-with-border-and-padding.html [
NeedsRebaseline ]
+crbug.com/414283 fast/forms/listbox-hit-test-zoomed.html [ NeedsRebaseline ]
+crbug.com/414283 fast/overflow/image-selection-highlight.html [ NeedsRebaseline
]
+crbug.com/414283 compositing/overflow/scroll-ancestor-update.html [
NeedsRebaseline ]
+crbug.com/414283 fast/overflow/unreachable-overflow-rtl-bug.html [
NeedsRebaseline ]
+crbug.com/414283
virtual/prefer_compositing_to_lcd_text/compositing/overflow/overflow-scroll.html
[ NeedsRebaseline ]
+crbug.com/414283 fast/css/text-overflow-ellipsis.html [ NeedsRebaseline ]
+crbug.com/414283 fast/css/text-overflow-ellipsis-text-align-left.html [
NeedsRebaseline ]
+crbug.com/414283 compositing/self-painting-layers.html [ NeedsRebaseline ]
+crbug.com/414283 fast/forms/input-text-scroll-left-on-blur.html [
NeedsRebaseline ]
+crbug.com/414283 fast/overflow/overflow-rtl.html [ NeedsRebaseline ]
+crbug.com/414283
fast/css/vertical-text-overflow-ellipsis-text-align-center.html [
NeedsRebaseline ]
+crbug.com/414283 fast/overflow/overflow-text-hit-testing.html [ NeedsRebaseline
]
+crbug.com/414283 fast/css/text-overflow-ellipsis-text-align-center.html [
NeedsRebaseline ]
+crbug.com/414283 fast/overflow/008.html [ NeedsRebaseline ]
+crbug.com/414283 fast/css/text-overflow-ellipsis-text-align-right.html [
NeedsRebaseline ]
+crbug.com/414283 fast/forms/search-rtl.html [ NeedsRebaseline ]
+crbug.com/414283 compositing/overflow/overflow-scroll.html [ NeedsRebaseline ]
+crbug.com/414283 fast/css/text-overflow-ellipsis-strict.html [ NeedsRebaseline
]
+crbug.com/414283 svg/custom/getscreenctm-in-scrollable-div-area-nested.xhtml [
NeedsRebaseline ]
+crbug.com/414283 fast/table/edge-offsets.html [ NeedsRebaseline ]
+crbug.com/414283 fast/css/text-overflow-input.html [ NeedsRebaseline ]
+crbug.com/414283 fast/forms/select-initial-position.html [ NeedsRebaseline ]
+crbug.com/414283 fast/transforms/selection-bounds-in-transformed-view.html [
NeedsRebaseline ]
+crbug.com/414283 editing/input/caret-at-the-edge-of-contenteditable.html [
NeedsRebaseline ]
+crbug.com/414283 fast/forms/input-readonly-autoscroll.html [ NeedsRebaseline ]
+crbug.com/414283 fast/overflow/overflow-rtl-vertical.html [ NeedsRebaseline ]
+crbug.com/414283 editing/input/reveal-caret-of-multiline-contenteditable.html [
NeedsRebaseline ]
+crbug.com/414283 fast/overflow/scrollRevealButton.html [ NeedsRebaseline ]
+crbug.com/414283
fast/css/vertical-text-overflow-ellipsis-text-align-justify.html [
NeedsRebaseline ]
+crbug.com/414283
virtual/prefer_compositing_to_lcd_text/compositing/overflow/scroll-ancestor-update.html
[ NeedsRebaseline ]
+crbug.com/414283 editing/input/caret-at-the-edge-of-input.html [
NeedsRebaseline ]
+crbug.com/414283 fast/events/touch/gesture/touch-gesture-scroll-listbox.html [
NeedsRebaseline ]
+

[miletus1, 2014-10-09 16:44:37]

The CQ bit was checked by miletus@google.com

[commit-bot: I haz the power, 2014-10-09 16:45:39]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/610423004/200001

[commit-bot: I haz the power, 2014-10-09 18:55:36]

Committed patchset #9 (id:200001) as 183486