GitLab

In Source/web/ there are many cyclic dependencies. These cycles make it very
difficult to discretely move class implementation from web/ to either core/ or
modules/.

An example cycle is WebViewImpl and WebLocalFrameImpl. Each refer to the other,
rather than the least derived types of WebView and WebLocalFrame (or WebFrame).
It is impossible to use the least derived type as methods have been added to the
*Impl class definitions that are consumed by other classes in Source/web.

To break the cycle we introduce the class WebViewBase, so that the inheritance
hierarchy is now WebView <- WebViewBase <- WebViewImpl (<- = implements).

Methods that were defined in WebViewImpl are made pure virtual in WebViewBase,
and now classes that were taking a dependency on WebViewImpl instead now take a
dependency on WebViewBase. This breaks the dependency chain and allows us to
start moving these classes out of Source/web into new homes in modules or core.
Note: WebViewBase is defined in core/exported/WebViewBase.h as is implements
public/web/WebView.h. Once this refactoring is complete WebViewBase will be
removed and these changes effectively reverted.

This CL moves the first portion of classes to use WebViewBase rather than
WebViewImpl, a followup CL will complete this move.

BUG=712963

Review-Url: https://codereview.chromium.org/2848513002
Cr-Commit-Position: refs/heads/master@{#467905}

Review-Url: https://codereview.chromium.org/2844253002
Cr-Commit-Position: refs/heads/master@{#467904}

The "Open" button was calling SendBannerAccepted, which made the
assumption that the app was being installed. Removed this call and
documented the method's semantics.

BUG=716313

Review-Url: https://codereview.chromium.org/2848703005
Cr-Commit-Position: refs/heads/master@{#467903}

Updates of message, proxy and agent are needed to forward capturer id to the
network process and eventually to the client.

BUG=chromium:679523, chromium:650926

Review-Url: https://codereview.chromium.org/2848753002
Cr-Commit-Position: refs/heads/master@{#467902}

The client_certs aren't actually part of the certificate request, rather the client_certs member was used to store the list of matching certificates on the client side and pass them through to the certificate selector.
Pass the list of certs through the relevant callbacks instead.

BUG=166642,394131

Review-Url: https://codereview.chromium.org/2838243002
Cr-Commit-Position: refs/heads/master@{#467901}

* Fixes an issue from crrev.com/467567, where it was possible to resize the
  sidebar without any limits on minimum or maximum width.
* Fixes an issue from crrev.com/467582, where text was colored white
  incorrectly in the bookmark add dialog.

BUG=692844
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:closure_compilation

Review-Url: https://codereview.chromium.org/2850673003
Cr-Commit-Position: refs/heads/master@{#467900}

Per https://crbug.com/336424#c31 DeviceLightEvent is no longer on track
for standardization. This CL removes its implementation.

BUG=711577

Review-Url: https://codereview.chromium.org/2845763002
Cr-Commit-Position: refs/heads/master@{#467899}

Currently, tether settings appear as part of their own subpage. In the future, these settings will be merged with cellular networks under the "Mobile" subpage.

BUG=672263
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:closure_compilation

Review-Url: https://codereview.chromium.org/2838323002
Cr-Commit-Position: refs/heads/master@{#467898}

If a download ends with CONTENT_LENGTH_MISMATCH error, chrome treat it as completed.
However, this could happen if server closes the connection too early.
And it causes the downloaded file size to be incorrect.
And with pause/resume functionalities, this could happen more frequently.
This change treats these download as interrupted if strong validators are present.
For downloads without strong validators, auto resumption could cause them to enter a restart->interrupt->restart cycle.
As a result, download is treated as completed if there are no strong validators.

BUG=453357

Review-Url: https://codereview.chromium.org/2832223004
Cr-Commit-Position: refs/heads/master@{#467897}

Currently NetworkHandler::NavigationPreloadCompleted() is using TimeTicks::Now()
to call LoadingFinished() of DevTools's Network event. But for normal requests
from the renderer, FrameFetchContext::DispatchDidFinishLoading() is using the
|finish_time| which is originated from ResourceRequestCompletionStatus's
completion_time.

NetworkHandler::NavigationPreloadCompleted() should use the same timing
information.

I found this difference while investigating the issue 712809. |completion_time|
is set when the request finished but, NavigationPreloadCompleted() isn't called
before the service worker finishes starting up.

BUG=712809

Review-Url: https://codereview.chromium.org/2844133005
Cr-Commit-Position: refs/heads/master@{#467896}

Resizing the content area no longer shows the scrollbars, so we don't need a
separate fade delay anymore.

BUG=606395
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel

Review-Url: https://codereview.chromium.org/2838053002
Cr-Commit-Position: refs/heads/master@{#467895}

Summary of changes available at:
https://chromium.googlesource.com/v8/v8/+log/71c1795a..47389137

Please follow these instructions for assigning/CC'ing issues:
https://github.com/v8/v8/wiki/Triaging%20issues

Please close rolling in case of a roll revert:
https://v8-roll.appspot.com/
This only works with a Google account.

CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel;master.tryserver.chromium.android:android_optional_gpu_tests_rel

TBR=hablich@chromium.org,machenbach@chromium.org,kozyatinskiy@chromium.org,vogelheim@chromium.org

Review-Url: https://codereview.chromium.org/2846173002
Cr-Commit-Position: refs/heads/master@{#467894}

Add sleep to SetWindowBounds to give chrome ui enough time to respond to devtools command.

BUG=chromedriver:1779

Review-Url: https://codereview.chromium.org/2836023003
Cr-Commit-Position: refs/heads/master@{#467893}

As a result, removing locks and synchronizations.
Also changing the way we store and report the OOM protected state.
We now update the OOM protected state every time it changes as long as we
are bound (that happens only on the launcher thread).
When retrieving that OOM protected state (which happens on the IO
thread), we return that state directly without the need of a lock.

BUG=714657

Review-Url: https://codereview.chromium.org/2840303002
Cr-Commit-Position: refs/heads/master@{#467892}

This is preparation for https://codereview.chromium.org/2850673002
to create a place to insert RegisterModuleScript() both for
Create() and CreateForTest().

This doesn't change the behavior.

BUG=594639, 715369

Review-Url: https://codereview.chromium.org/2844413003
Cr-Commit-Position: refs/heads/master@{#467891}

Editor used to update layout after applying commands for setting
frame selection. Now that setting frame selection no longer requires
clean layout, this patch removes these layout update calls.

BUG=590369
TEST=n/a; no behavior change

Review-Url: https://codereview.chromium.org/2729313002
Cr-Commit-Position: refs/heads/master@{#467890}

Some ozone platforms (e.g. drm) expects to be able to watch a file
descriptor, so use a TYPE_IO message-loop for the gpu process host
thread.

BUG=715957
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel

Review-Url: https://codereview.chromium.org/2852553003
Cr-Commit-Position: refs/heads/master@{#467889}

R=rockot@chromium.org
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_site_isolation

Review-Url: https://codereview.chromium.org/2844923002
Cr-Commit-Position: refs/heads/master@{#467888}

This patch prunes a call site of the deprecated function to improve code
health.

BUG=698633
TEST=n/a; no behavioral change

Review-Url: https://codereview.chromium.org/2849573002
Cr-Commit-Position: refs/heads/master@{#467887}

Revert of Roll ANGLE f584177..e8b3551 (patchset #1 id:1 of https://codereview.chromium.org/2849583003/ )

Reason for revert:

Findit(https://goo.gl/kROfz5) identified CL at revision 467881 as the
culprit for failures in the build cycles as shown on:
https://findit-for-me.appspot.com/waterfall/culprit?key=ag9zfmZpbmRpdC1mb3ItbWVyRAsSDVdmU3VzcGVjdGVkQ0wiMWNocm9taXVtL2ZhMGE3ZTY2ZGQxNTgzYWI3MTQzNmYyNmVlMzYyMjc4NDM3ZTE5NzIM

Original issue's description:
> Roll ANGLE f584177..e8b3551
>
> https://chromium.googlesource.com/angle/angle.git/+log/f584177..e8b3551
>
> BUG=715875,705865
> TBR=jmadill@chromium.org, geofflang@chromium.org
>
> TEST=bots
>
> CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.win:win_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.android:android_optional_gpu_tests_rel
>
> Review-Url: https://codereview.chromium.org/2849583003
> Cr-Commit-Position: refs/heads/master@{#467881}
> Committed: https://chromium.googlesource.com/chromium/src/+/fa0a7e66dd1583ab71436f26ee362278437e1972

TBR=jmadill@chromium.org,geofflang@chromium.org,ynovikov@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=715875,705865

Review-Url: https://codereview.chromium.org/2846763004
Cr-Commit-Position: refs/heads/master@{#467886}

This patch makes EBoxDecoration break an enum class for better type
safety.

BUG=684966
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_layout_tests_slimming_paint_v2

Review-Url: https://codereview.chromium.org/2845893002
Cr-Commit-Position: refs/heads/master@{#467885}

This change helps us avoid duplication in the MarkerType-specific list
implementations I'm going to add. Instead of having one
DocumentMarkerListEditor::AddMarker() method that works for
GenericDocumentMarkerListImpl, but then requires each of the other list
implementations to implement the method separately, we'll have one method,
AddAndMergeOverlapping(), that works for Spelling/Grammar markers (and can be
moved into SpellCheckMarkerList once we eliminate
GenericDocumentMarkerListImpl), and another method
AddWithoutMergingOverlapping(), that can be used for the other marker types.
GenericDocumentMarkerListImpl will choose which to use based on the type of the
inserted marker.

Right now both of these methods create a RenderedDocumentMarker from the
passed-in DocumentMarker. I will move this piece out of the method once not all
marker list implementations are still creating RenderedDocumentMarkers.

BUG=707867

Review-Url: https://codereview.chromium.org/2842263002
Cr-Commit-Position: refs/heads/master@{#467884}

VideoResourceUpdater uses weak pointers within the class only.
Hence, it does not have to inherit SupportsWeakPtr.

BUG=None
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel

Review-Url: https://codereview.chromium.org/2838713004
Cr-Commit-Position: refs/heads/master@{#467883}

Before this CL, when instantiation was reattempt for previously errored module,
ModuleTreeLinker::Instantiate() crashed.

This CL updates the code to follow whatwg/html update:
https://github.com/whatwg/html/pull/2559
so that we rethrow previous error when instantiation is reattempt.
TEST=webkit_unit_test --gtest_filter=ModuleTreeLinkerTest.FetchTreePreviousInstantiationFailure
BUG=594639,714521

Review-Url: https://codereview.chromium.org/2839563002
Cr-Commit-Position: refs/heads/master@{#467882}

https://chromium.googlesource.com/angle/angle.git/+log/f584177..e8b3551

BUG=715875,705865
TBR=jmadill@chromium.org, geofflang@chromium.org

TEST=bots

CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.win:win_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.android:android_optional_gpu_tests_rel

Review-Url: https://codereview.chromium.org/2849583003
Cr-Commit-Position: refs/heads/master@{#467881}

The canvas element simply mirrors the video element by grabbing frames from it.

Removing these 3 pages should reduce benchmark_duration for tough_video_cases by ~17%.

BUG=710253

Review-Url: https://codereview.chromium.org/2848723003
Cr-Commit-Position: refs/heads/master@{#467880}

This is a refry of https://crrev.com/2801803002 (as pointed out
in https://crrev.com/2836013006) that enables only VP8
hardware encoding in Android.

Tested via webrtc_media_recorder_browsertest.cc [1]

[1] https://cs.chromium.org/chromium/src/content/browser/webrtc/webrtc_media_recorder_browsertest.cc?q=webrtc+mediarecorder&sq=package:chromium&dr=CSs&l=27

BUG=638664

Review-Url: https://codereview.chromium.org/2852583002
Cr-Commit-Position: refs/heads/master@{#467879}

This roll includes:
  - [inspector_protocol] added StringUtil::toDouble method as requirement [1]

[1] https://codereview.chromium.org/2843223005/

BUG=712610
R=dgozman@chromium.org

Review-Url: https://codereview.chromium.org/2846113002
Cr-Commit-Position: refs/heads/master@{#467878}

When I moved the multi window swap interval logic, I changed a condition
by mistake. With that change the swap interval is reset to 1 for the
first swap in a new swap generation. I suspect that caused the input
latency regression although I can't reproduce it in local testing.

R=kbr
BUG=714902

CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel

Change-Id: If7f7ae4f824b39dade4cd9d81e68ce47cfc8f38d
Reviewed-on: https://chromium-review.googlesource.com/487875Reviewed-by: Kenneth Russell <kbr@chromium.org>
Commit-Queue: Sunny Sachanandani <sunnyps@chromium.org>
Cr-Commit-Position: refs/heads/master@{#467877}

https://pdfium.googlesource.com/pdfium.git/+log/f452f56021f2..66568bcd683d

$ git log f452f5602..66568bcd6 --date=short --no-merges --format='%ad %ae %s'
2017-04-26 thestig Assert CPDF_Pattern always has a valid CPDF_Document pointer.
2017-04-27 tsepez Zero 4 byte variable before reading 3 bytes into it
2017-04-27 tsepez Add barcode test skeleton
2017-04-27 thestig Remove CBC_HighLevelEncoder::illegalCharacter().

Created with:
  roll-dep src/third_party/pdfium


Documentation for the AutoRoller is here:
https://skia.googlesource.com/buildbot/+/master/autoroll/README.md

If the roll is causing failures, see:
http://www.chromium.org/developers/tree-sheriffs/sheriff-details-chromium#TOC-Failures-due-to-DEPS-rolls


TBR=dsinclair@chromium.org

Change-Id: I540cb377030a5f3f626c901af0c13d7887642217
Reviewed-on: https://chromium-review.googlesource.com/489814
Reviewed-by: <pdfium-deps-roller@chromium.org>
Commit-Queue: <pdfium-deps-roller@chromium.org>
Cr-Commit-Position: refs/heads/master@{#467876}

Since we started using LayoutOpportunity iterator for positioning BFC
fragments, we can't longer rely on the assumption that
FindLayoutOpportunityForFragment will be used for floats only.

List of changes:
- Change FindLayoutOpportunityForFragment to check if fragment's block
  size fits into a layout opportunity.
- Change NGInlineLayoutAlgorithm to avoid adding font's descent to
  line_bottom if the text only includes object replacement characters
  I need it to fix floats-wrap-top-below-002l-ref.xht.

BUG=635619

Review-Url: https://codereview.chromium.org/2840883002
Cr-Commit-Position: refs/heads/master@{#467875}

The vertical_align_ field in ComputedStyle currently has generated
storage, but not copy and compare. This patch makes it use generated
copy and compare. We are able to do this now because we have added
the 'has_custom_compare_and_copy' parameter in a previous patch.

Diff of generated files:
https://gist.github.com/183fa3073f5de347a685b1834913ee2e/revisions

BUG=628043

Review-Url: https://codereview.chromium.org/2834583004
Cr-Commit-Position: refs/heads/master@{#467874}

Because it will be used from outside ModuleScriptLoader when we support
inline scripts.

This CL
- Renames existing ModuleScript::Create() (which is a simple wrapper of
  the constructor) to CreateForTest(),
- Moves ModuleScriptLoader::CreateModuleScript() to new
  ModuleScript::Create(), and
- Reorders the parameters to match with the spec.

BUG=594639, 715369

Review-Url: https://codereview.chromium.org/2845743003
Cr-Commit-Position: refs/heads/master@{#467873}

Like the WTF::String change, this leaves the the Find() overload that
takes a StringView, start offset, and a case sensitivity option for a
followup CL.

Bug: 709815
Change-Id: I8f98fe0b41eb99a525ca608690c8ae0bdcebf6a6
Reviewed-on: https://chromium-review.googlesource.com/489368Reviewed-by: Kentaro Hara <haraken@chromium.org>
Commit-Queue: Daniel Cheng <dcheng@chromium.org>
Cr-Commit-Position: refs/heads/master@{#467872}

Native layer is already calculating some frame info about
top control (position + update flag). This CL passes
the values to Java layer so that it won't repeat those
operations again.

Review-Url: https://codereview.chromium.org/2841033002
Cr-Commit-Position: refs/heads/master@{#467871}

Revert of Only do Smart Select if one word is selected. (patchset #4 id:60001 of https://codereview.chromium.org/2838023002/ )

Reason for revert:
Null pointer crash http://crbug.com/715526

BUG=715526

Original issue's description:
> Only do Smart Select if one word is selected.
>
> There is a bug where after "Select All" Smart Select will suggest a different
> (smaller) selection. This CL checks if the selection only has one word. Since
> longpress/double tap only selects one word, if we have multiple words we can assume
> it comes from "Select All" and not do Smart Select.
>
> BUG=714106
>
> Review-Url: https://codereview.chromium.org/2838023002
> Cr-Commit-Position: refs/heads/master@{#466856}
> Committed: https://chromium.googlesource.com/chromium/src/+/03f5058e764e8b2b01d571f29426075a83f6049b

TBR=timav@chromium.org,amaralp@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=714106

Review-Url: https://codereview.chromium.org/2848793002
Cr-Commit-Position: refs/heads/master@{#467870}

Indicates whether an error has been encountered by the interface endpoint client. If true, method calls made on this interface will be dropped (and may already have been dropped).

BUG=695635

Review-Url: https://codereview.chromium.org/2840333002
Cr-Commit-Position: refs/heads/master@{#467869}

When logging a URL with a port number in the console, 'http://localhost:8000',
our linkifier incorrectly treats the '8000' port number as a line number.

Using our existing granular URL regexes, this patch makes the linkifier ignore
the URL part before the 'path' when looking for a line/column number.

BUG=701865

Review-Url: https://codereview.chromium.org/2843073003
Cr-Commit-Position: refs/heads/master@{#467868}

I think this used to be built by the old Windows official builders, but
I have no idea if it was ever even used, and AFAICT, it hasn't ever been
built by the current (PGO) builders, and hasn't been missed. It seems we
don't really need it (especially not for the official continuous
builders, which appear to be the only ones that currently build the
"chrome_official_builder" target).

R=dimu@google.com, dpranke@google.com

Review-Url: https://codereview.chromium.org/2846963002
Cr-Commit-Position: refs/heads/master@{#467867}

This enables module scripts in Blink behind a flag.

This CL implements module-script-related steps of "prepare a script"
in ScriptLoader and thus makes the module implementation code paths that
are committed previously to be executed, if the runtime flag is enabled.

This CL also reorganizes the test expectations for module-related tests.

BUG=594639

Review-Url: https://codereview.chromium.org/2781713003
Cr-Commit-Position: refs/heads/master@{#467866}