GitLab

Summary of changes available at:
https://chromium.googlesource.com/skia/+log/35f5576..fbcf0bd

CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
TBR=msarett@google.com

Review URL: https://codereview.chromium.org/1077583002

Cr-Commit-Position: refs/heads/master@{#324426}

We were leaking activity Contexts in ResourcesContextWrapperFactory's
cache of wrapped contexts, because WeakHashMap only references its
*keys* weakly, and the ContextWrapper objects used as values in the
hashmap have a strong reference to the Context being used as a key, so
nothing was ever removed from the map.

Fix this by deleting the cache entirely, as it's not really necessary;
the number of wrapped contexts we create is already bounded to two per
WebView (one for the activity context and one for the application
context), and the objects are very small, so it's not worth the
complexity of trying to avoid lifetime issues while reusing them.

We do still need to ensure that WebViewChromiumFactoryProvider always
uses the same wrapper for the application context, though, since in some
initialisation paths we set the application context more than once and
the underlying code only permits that if the same context object is used
each time.

BUG=473146

Review URL: https://codereview.chromium.org/1078693003

Cr-Commit-Position: refs/heads/master@{#324425}

WebRTC 8955:8956
Changes: https://chromium.googlesource.com/external/webrtc/trunk/webrtc.git/+log/a4d326e..9e4853a

Libjingle 8954:8956
Changes: https://chromium.googlesource.com/external/webrtc/trunk/talk.git/+log/57c1e6f..32d6efb

TBR=perkj

Review URL: https://codereview.chromium.org/1071123002

Cr-Commit-Position: refs/heads/master@{#324424}

New bubble implementation was done in https://crrev.com/319892

BUG=454815
R=vabr@chromium.org

Review URL: https://codereview.chromium.org/1078723002

Cr-Commit-Position: refs/heads/master@{#324423}

Cr-Commit-Position: refs/heads/master@{#324422}

Drop "offline sign in" help article reference since HelpApp doesn't have that content.

BUG=468257,469459

Review URL: https://codereview.chromium.org/1058433002

Cr-Commit-Position: refs/heads/master@{#324421}

Summary of changes available at:
https://chromium.googlesource.com/v8/v8/+log/1cb04a8e..ca308522

Please follow these instructions for assigning/CC'ing issues:
https://code.google.com/p/v8-wiki/wiki/TriagingIssues

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

Review URL: https://codereview.chromium.org/1070053002

Cr-Commit-Position: refs/heads/master@{#324420}

Review URL: https://codereview.chromium.org/1072803002

Cr-Commit-Position: refs/heads/master@{#324419}

Add a flag to the WebView manifest to ensure that webview apks are
compatible with the new mechanism for loading webview.

BUG=

Review URL: https://codereview.chromium.org/1075883002

Cr-Commit-Position: refs/heads/master@{#324418}

- use ALIGNOF macro
- use variadic templates

Review URL: https://codereview.chromium.org/1070603003

Cr-Commit-Position: refs/heads/master@{#324417}

Summary of changes available at:
https://chromium.googlesource.com/chromium/blink/+log/eb1578b..b559582

TBR=keishi@chromium.org

Review URL: https://codereview.chromium.org/1071913003

Cr-Commit-Position: refs/heads/master@{#324416}

Revert of Refactor GestureNavigation to eliminate code redundancy (patchset #32 id:620001 of https://codereview.chromium.org/895543005/)

Reason for revert:
After this CL, content browsertest WebContentsViewAuraTest.RepeatedQuickOverscrollGestures started failing with segmentation fault on almost every run on Linux ChromiumOS Tests (dbg)(1).

See: http://build.chromium.org/p/chromium.chromiumos/builders/Linux%20ChromiumOS%20Tests%20%28dbg%29%281%29/builds/850

Original issue's description:
> Refactor GestureNavigation to eliminate code redundancy
>
> Right now, Gesture Navigation has two very distinct phases: the initial
> one, where the web contents window receives events and is moved, and the
> follow up navigation, where the page still hasn't loaded and the user
> scrolls on the overlay. The meat of this patch is to unify as much of
> the logic as possible without sacrificing features. A new class is
> introduced, OverscrollWindowAnimation, and the responsibility of the
> existing OverscrollNavigationOverlay is steered into performing the
> actual navigation.
>
> TEST=Overscroll*
> BUG=467692,464532,420121
>
> Committed: https://crrev.com/5384f002f839439ef666ed9688069e42ca5ccdca
> Cr-Commit-Position: refs/heads/master@{#324275}

TBR=mfomitchev@chromium.org,sadrul@chromium.org,danakj@chromium.org,avi@chromium.org,nsatragno@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=467692,464532,420121

Review URL: https://codereview.chromium.org/1076743003

Cr-Commit-Position: refs/heads/master@{#324415}

This CL makes it so that the helper functions used in RenderViewImpltest work
with browser-side navigation enabled. This makes 20 browser tests pass.

BUG=475027

Review URL: https://codereview.chromium.org/1064413003

Cr-Commit-Position: refs/heads/master@{#324414}

In application_lifetime.h, it is said that what ExitCleanly() does is "Shutdown chrome cleanly without blocking".
However, it actually calls BrowserProcessImpl::EndSession(), which blocks the UI thread waiting for Preferences file to be written.

This EndSession() call was added by https://codereview.chromium.org/7273038.

By getting rid of this EndSession() call in ExitCleanly(), we can avoid blocking the UI thread waiting for the disk IO.
Also, we'll be able to avoid writing the Preferences file multiple times during shutdown.
This should make it less likely to hit SessionManager's 3-second timeout.

BUG=470501

Review URL: https://codereview.chromium.org/1057533003

Cr-Commit-Position: refs/heads/master@{#324413}

The android_aosp bot is gone and so these files are no longer used.

BUG=440793

Review URL: https://codereview.chromium.org/1071873003

Cr-Commit-Position: refs/heads/master@{#324412}

WebRTC 8945:8955
Changes: https://chromium.googlesource.com/external/webrtc/trunk/webrtc.git/+log/f814f6d..a4d326e

Libjingle 8941:8954
Changes: https://chromium.googlesource.com/external/webrtc/trunk/talk.git/+log/82bd4f4..57c1e6f

TBR=perkj

Review URL: https://codereview.chromium.org/1078643002

Cr-Commit-Position: refs/heads/master@{#324411}

Summary of changes available at:
https://chromium.googlesource.com/v8/v8/+log/ca9f7433..1cb04a8e

Please follow these instructions for assigning/CC'ing issues:
https://code.google.com/p/v8-wiki/wiki/TriagingIssues

TBR=hablich@chromium.org,machenbach@chromium.org,yangguo@chromium.org,vogelheim@chromium.org
BUG=

Review URL: https://codereview.chromium.org/1076733002

Cr-Commit-Position: refs/heads/master@{#324410}

Property tree verification was inadvertently disabled by
http://crrev.com/976553002.

This CL re-enables it.

Review URL: https://codereview.chromium.org/1068383002

Cr-Commit-Position: refs/heads/master@{#324409}

Enterprise.DevicePolicyInvalidations can take the values [0, 1, 3, 4],
not [0, 1, 2, 3] as erroneously specified in histograms.xml.

BUG=None
TEST=None

Review URL: https://codereview.chromium.org/1067323002

Cr-Commit-Position: refs/heads/master@{#324408}

BUG=444203

Review URL: https://codereview.chromium.org/1063413002

Cr-Commit-Position: refs/heads/master@{#324407}

Summary of changes available at:
https://chromium.googlesource.com/chromium/blink/+log/ec58772..eb1578b

TBR=keishi@chromium.org

Review URL: https://codereview.chromium.org/1069983003

Cr-Commit-Position: refs/heads/master@{#324406}

R=posciak@chromium.org, wuchengli@chromium.org
BUG=none
TEST=play h264 html5 video

Review URL: https://codereview.chromium.org/1048773002

Cr-Commit-Position: refs/heads/master@{#324405}

Revert of Make ui/gl depend on the necessary runtime targets on Windows. (patchset #3 id:40001 of https://codereview.chromium.org/1070933002/)

Reason for revert:
This CL introduces a typo to the GN build file, so that Windows GN builds fail.

It should be possible to reland after below line is changed from appending to assignment.

ERROR at //ui/gl/BUILD.gn:203:5: Undefined variable for +=.
    data_deps += [
    ^--------
I don't have something with this name in scope now.

See: http://build.chromium.org/p/chromium.win/builders/Win8%20GN

Original issue's description:
> Make ui/gl depend on the necessary runtime targets on Windows.
>
> Committed: https://crrev.com/f4a6e44ff20c29a6410a698a06a45dd11666ef53
> Cr-Commit-Position: refs/heads/master@{#324320}

TBR=brettw@chromium.org,scottmg@chromium.org,jam@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

Review URL: https://codereview.chromium.org/1071113002

Cr-Commit-Position: refs/heads/master@{#324404}

Addressing comments from CL 1004733003.

Review URL: https://codereview.chromium.org/1038203003

Cr-Commit-Position: refs/heads/master@{#324403}

We now stream deferred and parsing blocking scripts too, so some obsolete
histograms are no longer obsolete. Added new histograms.

R=isherman@chromium.org
BUG=

Review URL: https://codereview.chromium.org/1068203002

Cr-Commit-Position: refs/heads/master@{#324402}

This moves the image IDRs used on the chrome side back from IDR2_ to the
IDR_ versions. (The next and final patch will remove the IDR2_s from
webkit.)

BUG=471609, 467176

Review URL: https://codereview.chromium.org/1073523004

Cr-Commit-Position: refs/heads/master@{#324401}

In some cases we encounter situation when in processing onSubmit of password form we didn't receive information from password store yet. For example this can happen when JavaScript creates password form after submission and submit it (as for nytimes.com).

To process such situation this CL postpones checking if password form fetched from store until moment when we are going to decide if we should save it.

In order to process this correctly creating pending credentials in PasswordManager was moved to separate method and it is called only when both events happen - form submission and fetching from store finished.

Tests are not added yet

BUG=470322

Review URL: https://codereview.chromium.org/1050903002

Cr-Commit-Position: refs/heads/master@{#324400}

Replace the struct workaround for forward-declaring ChromeViewHostMsg_GetPluginInfo_Status with a proper enum class.

Review URL: https://codereview.chromium.org/1069703002

Cr-Commit-Position: refs/heads/master@{#324399}

It should mention that Windows is the only supported platform.

BUG=474961

Review URL: https://codereview.chromium.org/1069753002

Cr-Commit-Position: refs/heads/master@{#324398}

BUG=475377
TEST=manually

Review URL: https://codereview.chromium.org/1071033003

Cr-Commit-Position: refs/heads/master@{#324397}

BUG=475328

Review URL: https://codereview.chromium.org/1071603006

Cr-Commit-Position: refs/heads/master@{#324396}

Immediately select through the edge on external drags.
Fix minor OBO for external bounds checking.

BUG=415907
TEST=Quickly dragging outside the textfield immediately selects visible text.
R=pkasting@chromium.org

Review URL: https://codereview.chromium.org/1073523005

Cr-Commit-Position: refs/heads/master@{#324395}

Summary of changes available at:
https://chromium.googlesource.com/chromium/blink/+log/7557b3f..ec58772

TBR=cbiesinger@chromium.org,keishi@chromium.org

Review URL: https://codereview.chromium.org/1071103002

Cr-Commit-Position: refs/heads/master@{#324394}

The cryptauth component also uses base64url encoding, but cannot depend on
proximity_auth.

BUG=420316

Review URL: https://codereview.chromium.org/1063323002

Cr-Commit-Position: refs/heads/master@{#324393}

The patterns are internally sorted from the lowest precedence to highest (i.e.
from the less specific pattern to the most specific, though the rules are more
complicated than that), and they were shown to the user in this order too,
which was confusing and making them think it's not sorted at all. Showing the
highest precedence patterns first should be more intuitive and also sort the
equivalent patterns more-or-less alphabetically.

BUG=450580

Review URL: https://codereview.chromium.org/1047153002

Cr-Commit-Position: refs/heads/master@{#324392}

BUG=475313
TEST=manually

Review URL: https://codereview.chromium.org/1077513002

Cr-Commit-Position: refs/heads/master@{#324391}

Follow-up CL for https://codereview.chromium.org/1045173003/

BUG=472006
R=vabr@chromium.org

Review URL: https://codereview.chromium.org/1066423002

Cr-Commit-Position: refs/heads/master@{#324390}

Currently LTHI "background ticks" animations once every second. This background
ticking doesn't have any js detectable effect, nor user visible effects, which
means it is just wasting CPU cycles (and thus power).

BUG=371747
R=jdduke,brianderson,danakj

Review URL: https://codereview.chromium.org/1050833002

Cr-Commit-Position: refs/heads/master@{#324389}

Cr-Commit-Position: refs/heads/master@{#324388}

This test (for unknown reasons) depends on the frame being visible and
currently only passes because of an artificial delay in the compositor when
going invisible. This is probably way it is already disabled on Mac and
Windows.

The patch at http://crrev.com/1051123002 removes this delay causing this test
to become extremely flaky. See http://crbug.com/282116

BUG=282116
R=lazyboy,fsamuel

Review URL: https://codereview.chromium.org/1069803003

Cr-Commit-Position: refs/heads/master@{#324387}