Fix zoom in CSS paint worklets

Fix zoom in CSS paint worklets.

This patch does next things:

1) Propagate the zoom factor to CSSPaintDefinition, so we can calculate CSS size

2) Report correct CSS size in 'geom' parameter of CSS paint worklet.

3) Let PaintGeneratedImage report its intrinsic size, instead of improperly conforming to the size of its container.

BUG=578252
TEST=third_party/WebKit/LayoutTests/csspaint/paint2d-zoom.html

Committed: https://crrev.com/6b2b9c70636afcc1127e6b199d482a5ffee3fd84
Cr-Commit-Position: refs/heads/master@{#404891}

[Gleb Lanbin, 2016-06-29 23:42:10]

Description was changed from

==========
Propagate the zoom value into PaintRenderingContext2D's paint canvas.

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

to

==========
Propagate the zoom value into PaintRenderingContext2D's paint canvas.

BUG=567358
TEST=third_party/WebKit/LayoutTests/csspaint/paint2d-zoom.html
==========

[Gleb Lanbin, 2016-06-29 23:42:54]

Patchset #1 (id:1) has been deleted

[Gleb Lanbin, 2016-06-29 23:44:04]

glebl@chromium.org changed reviewers:
+ ikilpatrick@chromium.org

[ikilpatrick, 2016-06-30 17:24:53]

ikilpatrick@chromium.org changed reviewers:
+ chrishtr@chromium.org, junov@chromium.org

[ikilpatrick, 2016-06-30 17:24:54]

+junov & +chrishtr to sanity check (is this the best way to achieve this?)

Bug is currently pointing at the worklet bug :)

https://codereview.chromium.org/2104103003/diff/20001/third_party/WebKit/Layo...
File third_party/WebKit/LayoutTests/csspaint/paint2d-zoom-expected.txt (right):

https://codereview.chromium.org/2104103003/diff/20001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/csspaint/paint2d-zoom-expected.txt:1: CONSOLE
MESSAGE: line 30: The worklet should log: 'width: 100, height: 100'
Use the testRunner.dumpAsText here.

(will mean that we wont dump the layout tree which isn't needed).

https://codereview.chromium.org/2104103003/diff/20001/third_party/WebKit/Sour...
File third_party/WebKit/Source/modules/csspaint/CSSPaintDefinition.cpp (right):

https://codereview.chromium.org/2104103003/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/modules/csspaint/CSSPaintDefinition.cpp:55: const auto
adjustedSize = adjustSizeForZoom(size, zoom);
add a comment as to why we are doing this.

[ikilpatrick, 2016-06-30 17:25:35]

On 2016/06/30 17:24:54, ikilpatrick wrote:
> +junov & +chrishtr to sanity check (is this the best way to achieve this?)
> 
> Bug is currently pointing at the worklet bug :)
> 
>
https://codereview.chromium.org/2104103003/diff/20001/third_party/WebKit/Layo...
> File third_party/WebKit/LayoutTests/csspaint/paint2d-zoom-expected.txt
(right):
> 
>
https://codereview.chromium.org/2104103003/diff/20001/third_party/WebKit/Layo...
> third_party/WebKit/LayoutTests/csspaint/paint2d-zoom-expected.txt:1: CONSOLE
> MESSAGE: line 30: The worklet should log: 'width: 100, height: 100'
> Use the testRunner.dumpAsText here.
> 
> (will mean that we wont dump the layout tree which isn't needed).
> 
>
https://codereview.chromium.org/2104103003/diff/20001/third_party/WebKit/Sour...
> File third_party/WebKit/Source/modules/csspaint/CSSPaintDefinition.cpp
(right):
> 
>
https://codereview.chromium.org/2104103003/diff/20001/third_party/WebKit/Sour...
> third_party/WebKit/Source/modules/csspaint/CSSPaintDefinition.cpp:55: const
auto
> adjustedSize = adjustSizeForZoom(size, zoom);
> add a comment as to why we are doing this.

Also add a little more context to the patch description as why we are doing
this.

[Gleb Lanbin, 2016-06-30 20:36:54]

Description was changed from

==========
Propagate the zoom value into PaintRenderingContext2D's paint canvas.

BUG=567358
TEST=third_party/WebKit/LayoutTests/csspaint/paint2d-zoom.html
==========

to

==========
Propagate the zoom value into PaintRenderingContext2D's paint canvas.

BUG=578252
TEST=third_party/WebKit/LayoutTests/csspaint/paint2d-zoom.html
==========

[Gleb Lanbin, 2016-06-30 23:56:41]

Description was changed from

==========
Propagate the zoom value into PaintRenderingContext2D's paint canvas.

BUG=578252
TEST=third_party/WebKit/LayoutTests/csspaint/paint2d-zoom.html
==========

to

==========
Fix zoom in CSS paint worklets.

This patch does next things:

1) Propagate the zoom factor to CSSPaintDefinition, so we can calculate CSS size

2) Report correct CSS size in 'geom' parameter of CSS paint worklet.

3) Let PaintGeneratedImage report its intrinsic size, instead of improperly
conforming to the size of its container.


BUG=578252
TEST=third_party/WebKit/LayoutTests/csspaint/paint2d-zoom.html
==========

[Gleb Lanbin, 2016-06-30 23:57:20]

https://codereview.chromium.org/2104103003/diff/20001/third_party/WebKit/Layo...
File third_party/WebKit/LayoutTests/csspaint/paint2d-zoom-expected.txt (right):

https://codereview.chromium.org/2104103003/diff/20001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/csspaint/paint2d-zoom-expected.txt:1: CONSOLE
MESSAGE: line 30: The worklet should log: 'width: 100, height: 100'
On 2016/06/30 17:24:54, ikilpatrick wrote:
> Use the testRunner.dumpAsText here.
> 
> (will mean that we wont dump the layout tree which isn't needed).

Done.

https://codereview.chromium.org/2104103003/diff/20001/third_party/WebKit/Sour...
File third_party/WebKit/Source/modules/csspaint/CSSPaintDefinition.cpp (right):

https://codereview.chromium.org/2104103003/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/modules/csspaint/CSSPaintDefinition.cpp:55: const auto
adjustedSize = adjustSizeForZoom(size, zoom);
On 2016/06/30 17:24:54, ikilpatrick wrote:
> add a comment as to why we are doing this.

renamed the function and variable to make clear our intention here.

[Justin Novosad, 2016-07-04 15:45:59]

https://codereview.chromium.org/2104103003/diff/40001/third_party/WebKit/Layo...
File third_party/WebKit/LayoutTests/csspaint/paint2d-zoom.html (right):

https://codereview.chromium.org/2104103003/diff/40001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/csspaint/paint2d-zoom.html:22: ctx.strokeRect(20,
20, 60, 60);
Using a pixel-aligned rect prevents the test from verifying that content was
rasterized at the right resolution. I think it would be important to check that
the SkPicture was playedback at 300x300.

https://codereview.chromium.org/2104103003/diff/40001/third_party/WebKit/Sour...
File third_party/WebKit/Source/modules/csspaint/CSSPaintDefinition.cpp (right):

https://codereview.chromium.org/2104103003/diff/40001/third_party/WebKit/Sour...
third_party/WebKit/Source/modules/csspaint/CSSPaintDefinition.cpp:26: IntSize
adjustToCssSize(const IntSize& size, float zoom)
This name is confusing IMHO. My understanding is that the 'size' argument is the
CSS size, and that we are unzooming it to get some sort of intrinsic size.  The
name suggests the opposite.

https://codereview.chromium.org/2104103003/diff/40001/third_party/WebKit/Sour...
third_party/WebKit/Source/modules/csspaint/CSSPaintDefinition.cpp:29:
newSize.scale(1 / zoom);
Due to floating-point precision issues, this arithmetic is not guaranteed to
give the pre-zoom size.  In particular, the results are quite likely to be off
by one since the implementation of IntSize::scale truncates its results (as
opposed to rounding).

https://codereview.chromium.org/2104103003/diff/40001/third_party/WebKit/Sour...
third_party/WebKit/Source/modules/csspaint/CSSPaintDefinition.cpp:69: const auto
cssSize = adjustToCssSize(size, zoom);
abuse of "auto".

[Gleb Lanbin, 2016-07-06 20:36:18]

https://codereview.chromium.org/2104103003/diff/40001/third_party/WebKit/Layo...
File third_party/WebKit/LayoutTests/csspaint/paint2d-zoom.html (right):

https://codereview.chromium.org/2104103003/diff/40001/third_party/WebKit/Layo...
third_party/WebKit/LayoutTests/csspaint/paint2d-zoom.html:22: ctx.strokeRect(20,
20, 60, 60);
On 2016/07/04 15:45:58, Justin Novosad wrote:
> Using a pixel-aligned rect prevents the test from verifying that content was
> rasterized at the right resolution. I think it would be important to check
that
> the SkPicture was playedback at 300x300.

Justin, can you please advice how can I do that? I tried to use
internals.layerTreeAsText but it doesn't expose that data.

https://codereview.chromium.org/2104103003/diff/40001/third_party/WebKit/Sour...
File third_party/WebKit/Source/modules/csspaint/CSSPaintDefinition.cpp (right):

https://codereview.chromium.org/2104103003/diff/40001/third_party/WebKit/Sour...
third_party/WebKit/Source/modules/csspaint/CSSPaintDefinition.cpp:26: IntSize
adjustToCssSize(const IntSize& size, float zoom)
On 2016/07/04 15:45:59, Justin Novosad wrote:
> This name is confusing IMHO. My understanding is that the 'size' argument is
the
> CSS size, and that we are unzooming it to get some sort of intrinsic size. 
The
> name suggests the opposite.

done, renamed to getSpecifiedSize, according to Object-Sizing Terminology
https://drafts.csswg.org/css-images-3/#sizing-terms, i.e. "...The specified size
of an object is given by CSS, such as through the width and height..."

https://codereview.chromium.org/2104103003/diff/40001/third_party/WebKit/Sour...
third_party/WebKit/Source/modules/csspaint/CSSPaintDefinition.cpp:29:
newSize.scale(1 / zoom);
On 2016/07/04 15:45:59, Justin Novosad wrote:
> Due to floating-point precision issues, this arithmetic is not guaranteed to
> give the pre-zoom size.  In particular, the results are quite likely to be off
> by one since the implementation of IntSize::scale truncates its results (as
> opposed to rounding).

Done.

https://codereview.chromium.org/2104103003/diff/40001/third_party/WebKit/Sour...
third_party/WebKit/Source/modules/csspaint/CSSPaintDefinition.cpp:69: const auto
cssSize = adjustToCssSize(size, zoom);
On 2016/07/04 15:45:58, Justin Novosad wrote:
> abuse of "auto".

Done.

[chrishtr, 2016-07-06 21:04:01]

On 2016/06/30 at 17:24:54, ikilpatrick wrote:
> +junov & +chrishtr to sanity check (is this the best way to achieve this?)

Sorry for the late reply. Best way to achieve what?

> 
> Bug is currently pointing at the worklet bug :)
> 
>
https://codereview.chromium.org/2104103003/diff/20001/third_party/WebKit/Layo...
> File third_party/WebKit/LayoutTests/csspaint/paint2d-zoom-expected.txt
(right):
> 
>
https://codereview.chromium.org/2104103003/diff/20001/third_party/WebKit/Layo...
> third_party/WebKit/LayoutTests/csspaint/paint2d-zoom-expected.txt:1: CONSOLE
MESSAGE: line 30: The worklet should log: 'width: 100, height: 100'
> Use the testRunner.dumpAsText here.
> 
> (will mean that we wont dump the layout tree which isn't needed).
> 
>
https://codereview.chromium.org/2104103003/diff/20001/third_party/WebKit/Sour...
> File third_party/WebKit/Source/modules/csspaint/CSSPaintDefinition.cpp
(right):
> 
>
https://codereview.chromium.org/2104103003/diff/20001/third_party/WebKit/Sour...
> third_party/WebKit/Source/modules/csspaint/CSSPaintDefinition.cpp:55: const
auto adjustedSize = adjustSizeForZoom(size, zoom);
> add a comment as to why we are doing this.

[Justin Novosad, 2016-07-06 21:07:12]

On 2016/07/06 20:36:18, glebl wrote:
>
https://codereview.chromium.org/2104103003/diff/40001/third_party/WebKit/Layo...
> File third_party/WebKit/LayoutTests/csspaint/paint2d-zoom.html (right):
> 
>
https://codereview.chromium.org/2104103003/diff/40001/third_party/WebKit/Layo...
> third_party/WebKit/LayoutTests/csspaint/paint2d-zoom.html:22:
ctx.strokeRect(20,
> 20, 60, 60);
> On 2016/07/04 15:45:58, Justin Novosad wrote:
> > Using a pixel-aligned rect prevents the test from verifying that content was
> > rasterized at the right resolution. I think it would be important to check
> that
> > the SkPicture was playedback at 300x300.
> 
> Justin, can you please advice how can I do that? I tried to use
> internals.layerTreeAsText but it doesn't expose that data.

If you render a rotated rect instead of an axis-aligned rect, the graphic
results will
be dependent on rasterization resolution.  It might be hard to write an
-expected.html
that matches exactly though, so you might have to turn this into a pixel test

> 
>
https://codereview.chromium.org/2104103003/diff/40001/third_party/WebKit/Sour...
> File third_party/WebKit/Source/modules/csspaint/CSSPaintDefinition.cpp
(right):
> 
>
https://codereview.chromium.org/2104103003/diff/40001/third_party/WebKit/Sour...
> third_party/WebKit/Source/modules/csspaint/CSSPaintDefinition.cpp:26: IntSize
> adjustToCssSize(const IntSize& size, float zoom)
> On 2016/07/04 15:45:59, Justin Novosad wrote:
> > This name is confusing IMHO. My understanding is that the 'size' argument is
> the
> > CSS size, and that we are unzooming it to get some sort of intrinsic size. 
> The
> > name suggests the opposite.
> 
> done, renamed to getSpecifiedSize, according to Object-Sizing Terminology
> https://drafts.csswg.org/css-images-3/#sizing-terms, i.e. "...The specified
size
> of an object is given by CSS, such as through the width and height..."

That is excellent!

[Gleb Lanbin, 2016-07-06 21:48:26]

Patchset #4 (id:80001) has been deleted

[Gleb Lanbin, 2016-07-06 21:49:00]

On 2016/07/06 21:07:12, Justin Novosad wrote:
> On 2016/07/06 20:36:18, glebl wrote:
> >
>
https://codereview.chromium.org/2104103003/diff/40001/third_party/WebKit/Layo...
> > File third_party/WebKit/LayoutTests/csspaint/paint2d-zoom.html (right):
> > 
> >
>
https://codereview.chromium.org/2104103003/diff/40001/third_party/WebKit/Layo...
> > third_party/WebKit/LayoutTests/csspaint/paint2d-zoom.html:22:
> ctx.strokeRect(20,
> > 20, 60, 60);
> > On 2016/07/04 15:45:58, Justin Novosad wrote:
> > > Using a pixel-aligned rect prevents the test from verifying that content
was
> > > rasterized at the right resolution. I think it would be important to check
> > that
> > > the SkPicture was playedback at 300x300.
> > 
> > Justin, can you please advice how can I do that? I tried to use
> > internals.layerTreeAsText but it doesn't expose that data.
> 
> If you render a rotated rect instead of an axis-aligned rect, the graphic
> results will
> be dependent on rasterization resolution.  It might be hard to write an
> -expected.html
> that matches exactly though, so you might have to turn this into a pixel test

done. thanks.

> 
> > 
> >
>
https://codereview.chromium.org/2104103003/diff/40001/third_party/WebKit/Sour...
> > File third_party/WebKit/Source/modules/csspaint/CSSPaintDefinition.cpp
> (right):
> > 
> >
>
https://codereview.chromium.org/2104103003/diff/40001/third_party/WebKit/Sour...
> > third_party/WebKit/Source/modules/csspaint/CSSPaintDefinition.cpp:26:
IntSize
> > adjustToCssSize(const IntSize& size, float zoom)
> > On 2016/07/04 15:45:59, Justin Novosad wrote:
> > > This name is confusing IMHO. My understanding is that the 'size' argument
is
> > the
> > > CSS size, and that we are unzooming it to get some sort of intrinsic size.

> > The
> > > name suggests the opposite.
> > 
> > done, renamed to getSpecifiedSize, according to Object-Sizing Terminology
> > https://drafts.csswg.org/css-images-3/#sizing-terms, i.e. "...The specified
> size
> > of an object is given by CSS, such as through the width and height..."
> 
> That is excellent!

[Gleb Lanbin, 2016-07-08 21:57:26]

glebl@chromium.org changed reviewers:
- chrishtr@chromium.org

[Gleb Lanbin, 2016-07-08 21:57:27]

Justin,

please let me know if you have any other comments.
thanks.

[Justin Novosad, 2016-07-11 18:07:41]

lgtm

You will probably need NeedsRebaseline for paint2d-zoom.html in the
TestExpectations file since the pixel results are likely to be slightly
different from one platform to another.

Also, last time I tried, CQ does not like binary files. Since the png is just a
test result, you can remove it and NeedsRebaseline will take care of it.

[Gleb Lanbin, 2016-07-12 00:14:34]

On 2016/07/11 18:07:41, Justin Novosad wrote:
> lgtm
> 
> You will probably need NeedsRebaseline for paint2d-zoom.html in the
> TestExpectations file since the pixel results are likely to be slightly
> different from one platform to another.
> 
> Also, last time I tried, CQ does not like binary files. Since the png is just
a
> test result, you can remove it and NeedsRebaseline will take care of it.

done. thanks for the review

[Gleb Lanbin, 2016-07-12 00:14:40]

The CQ bit was checked by glebl@chromium.org

[Gleb Lanbin, 2016-07-12 00:14:40]

The patchset sent to the CQ was uploaded after l-g-t-m from junov@chromium.org
Link to the patchset: https://codereview.chromium.org/2104103003/#ps120001
(title: "mark paint2d-zoom.html as NeedsRebaseline")

[commit-bot: I haz the power, 2016-07-12 00:15:22]

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, 2016-07-12 00:18:56]

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

[commit-bot: I haz the power, 2016-07-12 00:18:58]

Try jobs failed on following builders:
  ios-device on master.tryserver.chromium.mac (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.mac/builders/ios-device/builds...)
  ios-device-gn on master.tryserver.chromium.mac (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.mac/builders/ios-device-gn/bui...)
  ios-simulator on master.tryserver.chromium.mac (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.mac/builders/ios-simulator/bui...)
  ios-simulator-gn on master.tryserver.chromium.mac (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.mac/builders/ios-simulator-gn/...)

[Gleb Lanbin, 2016-07-12 00:37:59]

The CQ bit was checked by glebl@chromium.org

[Gleb Lanbin, 2016-07-12 00:38:00]

The patchset sent to the CQ was uploaded after l-g-t-m from junov@chromium.org
Link to the patchset: https://codereview.chromium.org/2104103003/#ps140001
(title: "rebase")

[commit-bot: I haz the power, 2016-07-12 00:38:46]

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, 2016-07-12 01:33:26]

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

[commit-bot: I haz the power, 2016-07-12 01:33:28]

Try jobs failed on following builders:
  win_chromium_rel_ng on master.tryserver.chromium.win (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.win/builders/win_chromium_rel_...)

[Gleb Lanbin, 2016-07-12 06:10:05]

The CQ bit was checked by glebl@chromium.org

[Gleb Lanbin, 2016-07-12 06:10:05]

The patchset sent to the CQ was uploaded after l-g-t-m from junov@chromium.org
Link to the patchset: https://codereview.chromium.org/2104103003/#ps160001
(title: "2nd rebase")

[commit-bot: I haz the power, 2016-07-12 06:10:54]

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, 2016-07-12 10:14:22]

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

[commit-bot: I haz the power, 2016-07-12 10:14:24]

Try jobs failed on following builders:
  linux_chromium_asan_rel_ng on master.tryserver.chromium.linux (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_...)

[Gleb Lanbin, 2016-07-12 16:04:07]

The CQ bit was checked by glebl@chromium.org

[commit-bot: I haz the power, 2016-07-12 16:05:29]

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, 2016-07-12 20:16:10]

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

[commit-bot: I haz the power, 2016-07-12 20:16:11]

Try jobs failed on following builders:
  linux_chromium_asan_rel_ng on master.tryserver.chromium.linux (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_...)

[Gleb Lanbin, 2016-07-12 20:22:53]

The CQ bit was checked by glebl@chromium.org

[commit-bot: I haz the power, 2016-07-12 20:24:23]

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, 2016-07-13 02:10:05]

Description was changed from

==========
Fix zoom in CSS paint worklets.

This patch does next things:

1) Propagate the zoom factor to CSSPaintDefinition, so we can calculate CSS size

2) Report correct CSS size in 'geom' parameter of CSS paint worklet.

3) Let PaintGeneratedImage report its intrinsic size, instead of improperly
conforming to the size of its container.


BUG=578252
TEST=third_party/WebKit/LayoutTests/csspaint/paint2d-zoom.html
==========

to

==========
Fix zoom in CSS paint worklets.

This patch does next things:

1) Propagate the zoom factor to CSSPaintDefinition, so we can calculate CSS size

2) Report correct CSS size in 'geom' parameter of CSS paint worklet.

3) Let PaintGeneratedImage report its intrinsic size, instead of improperly
conforming to the size of its container.


BUG=578252
TEST=third_party/WebKit/LayoutTests/csspaint/paint2d-zoom.html
==========

[commit-bot: I haz the power, 2016-07-13 02:10:06]

Committed patchset #7 (id:160001)

[commit-bot: I haz the power, 2016-07-13 02:11:20]

CQ bit was unchecked.

[commit-bot: I haz the power, 2016-07-13 02:13:01]

Description was changed from

==========
Fix zoom in CSS paint worklets.

This patch does next things:

1) Propagate the zoom factor to CSSPaintDefinition, so we can calculate CSS size

2) Report correct CSS size in 'geom' parameter of CSS paint worklet.

3) Let PaintGeneratedImage report its intrinsic size, instead of improperly
conforming to the size of its container.


BUG=578252
TEST=third_party/WebKit/LayoutTests/csspaint/paint2d-zoom.html
==========

to

==========
Fix zoom in CSS paint worklets.

This patch does next things:

1) Propagate the zoom factor to CSSPaintDefinition, so we can calculate CSS size

2) Report correct CSS size in 'geom' parameter of CSS paint worklet.

3) Let PaintGeneratedImage report its intrinsic size, instead of improperly
conforming to the size of its container.

BUG=578252
TEST=third_party/WebKit/LayoutTests/csspaint/paint2d-zoom.html

Committed: https://crrev.com/6b2b9c70636afcc1127e6b199d482a5ffee3fd84
Cr-Commit-Position: refs/heads/master@{#404891}
==========

[commit-bot: I haz the power, 2016-07-13 02:13:03]

Patchset 7 (id:??) landed as
https://crrev.com/6b2b9c70636afcc1127e6b199d482a5ffee3fd84
Cr-Commit-Position: refs/heads/master@{#404891}