GitLab

Some of the changes below may have a functional effect, but I don't _think_ they will, if they do it should be minor, and I don't know another way of finding out than trying to make this change and seeing what happens.  (Besides simplification, this change is also intended to unblock https://codereview.chromium.org/1139823006/ , which doesn't want to affect infobar behavior but without this change might do so.)

* Moved the check for "!details.did_replace_entry" from ConfirmInfoBarDelegate up to the base class.  Very few people subclass the base class anyway, and for those who do, there's no obvious reason not to do this.  I actually suggested doing this in a review several years ago but it never happened.

* Removed InfoBarDelegate::ShouldExpireInternal(), which didn't seem to be doing much useful.  I think the unique ID check was basically doing the same thing as the "is_navigation_to_different_page" bit that was already checked in ShouldExpire() (which didn't exist when the unique IDs were added long ago).  So if we got to ShouldExpireInternal() at all, that would be true regardless, and thus that helper would always return true.  (Incidentally, reloads _are_ treated as "navigations to a different page", so we still dismiss infobars by default on reload.  If this wasn't true we'd have already been broken anyway, since ShouldExpire() would have early-returned false.)

* Removed ProtectedMediaIdentifierInfoBarDelegate::ShouldExpireInternal().  I discussed this with kkimlabs@, the original author, and neither of us can understand why it was necessary to begin with.  We both think this should be removed.

* Removed the wonky TranslateInfoBarDelegate::ShouldExpire().  Long ago, this used to check a bit called "is_auto", because it was trying to allow the infobar to be dismissed even for automatic navigations, whereas other infobars wanted to only be dismissed for user-driven ones.  However, the "is_auto" bit was later removed, and the replacement code didn't really work the same way, and besides this, the conditional didn't really make sense, since "!details.is_main_frame" implies "!details.is_navigation_to_different_page".  So we basically had a function that would call InfoBarDelegate::ShouldExpireInternal() for all main-frame navigations.  But since, as I said above, I think the unique ID check there was functionally the same as checking for a "navigation to a different page"... I think all this ended up just doing the same thing as the original base function.  And I couldn't check because the translate infobar no longer exists for views anyway.  So whatever.

* Removed copy-and-pasted code that wasn't needed.

BUG=none
TEST=none

Review URL: https://codereview.chromium.org/1142153002

Cr-Commit-Position: refs/heads/master@{#330817}

With 1 win32 timer we can only get a callback every 10ms, and since the callback only processes one message, that means the queue could keep growing if delayed tasks are posted faster than that rate, even if they're processed very quickly. To prevent that, schedule a kMsgHaveWork if there's 0ms until the next delayed task should run, and run a delayed task in the kMsgHaveWork handler.

BUG=454333
TEST=chrome resizes smoothly

Review URL: https://codereview.chromium.org/918473002

Cr-Commit-Position: refs/heads/master@{#330816}

This patch removes lazy_background_task_queue accessor from ExtensionSystem.
It can be its own browser context keyed service.
It can be built by its new factory.

Review URL: https://codereview.chromium.org/1129063011

Cr-Commit-Position: refs/heads/master@{#330815}

Review URL: https://codereview.chromium.org/1141453005

Cr-Commit-Position: refs/heads/master@{#330814}

Add option in display settings page to switch the mode

BUG=489809
TEST=manual + DisplayManagerTest.UnifiedDesktopBasic

Review URL: https://codereview.chromium.org/1126933004

Cr-Commit-Position: refs/heads/master@{#330813}

I noticed that HistoryQuick provider's also-do-HUP-like-scoring mode
did some score-capping to prevent a non-inlineable match from showing
first.  This is no longer necessary now that we have reorder mode
(a.k.a. the allowed_to_be_default_match field).

BUG=

Review URL: https://codereview.chromium.org/1144273003

Cr-Commit-Position: refs/heads/master@{#330812}

Android emulators miss many EGL extensions required for
webview hardware accelerated rendering, so fallback to
a slow software mode when running in emulator.

BUG=490354

Review URL: https://codereview.chromium.org/1145273003

Cr-Commit-Position: refs/heads/master@{#330811}

Summary of changes available at:
https://chromium.googlesource.com/chromium/blink/+log/6bd0043..a240d64

TBR=szager@chromium.org,leviw@chromium.org,samli@chromium.org

Review URL: https://codereview.chromium.org/1149823002

Cr-Commit-Position: refs/heads/master@{#330810}

This will handle destroying the owned tab models, which will in turn
destroy the owned tabs.

BUG=473594

Review URL: https://codereview.chromium.org/1147153002

Cr-Commit-Position: refs/heads/master@{#330809}

By doing the following:
1) Avoid deferring commits until draw if not drawing soon.
2) Don't send BeginMainFrame if swap throttling will block commit.

As a side-effect, some of the changes will also help release
the main thread sooner in some cases where we were holding on to
it for too long.

Additional scheduler deadlock tests added to verify existing
and future operation.

This should also enable us to more easily avoid unconditionally
setting active_tree_needs_first_draw_ to false in
SchedulerStateMachine::UpdateStateOnInvalidateOutputSurface
in followup patches if we want to.

BUG=479671, 477082

Review URL: https://codereview.chromium.org/1139613003

Cr-Commit-Position: refs/heads/master@{#330808}

This fixes StyleLabel layout in languages that don't have word breaks
(e.g. Japanese).

BUG=479859

Review URL: https://codereview.chromium.org/1140083002

Cr-Commit-Position: refs/heads/master@{#330807}

This prevents in-process discardable shared memory instances
from requiring an open file descriptor.

BUG=489174
TEST=base_unittests --gtest_filter=DiscardableSharedMemory.Close

Review URL: https://codereview.chromium.org/1146103002

Cr-Commit-Position: refs/heads/master@{#330806}

This CL removes some unneeded //chrome dependencies in password manager utilities. The goal is to make those files movable to the password manager component. The move will happen in a separate CL.

BUG=486739

Review URL: https://codereview.chromium.org/1141413002

Cr-Commit-Position: refs/heads/master@{#330805}

Revert of Roll DEPS to class-dump 3.5 (patchset #4 id:60001 of https://codereview.chromium.org/1106993002/)

Reason for revert:
Breaking https://build.chromium.org/p/chromium.mac/builders/iOS_Device/builds/7076

Original issue's description:
> Roll DEPS to class-dump 3.5
>
> Removes some build warnings.
>
> Moves the class-dump repo to a github auto-mirror instead of a mirror
> that needs to be manually updated.
>
> This also requires that the class-dump.gyp file move from the
> third-party repo into the main source tree. Note that the file
> structure of the class-dump source changed so the .gyp had to be
> updated. For comparison, the old .class-dump.gyp can be found at
> https://chromium.googlesource.com/chromium/deps/class-dump/+/master/class-dump.gyp
>
> BUG=318160
>
> Committed: https://crrev.com/1c80791ce7c29a29fc816d89d1a5fa2ac6503b3e
> Cr-Commit-Position: refs/heads/master@{#330798}

TBR=justincohen@chromium.org,rohitrao@chromium.org,lliabraa@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=318160

Review URL: https://codereview.chromium.org/1147163003

Cr-Commit-Position: refs/heads/master@{#330804}

This forks the code from //services/files/ in the mojo repository at
commit af91be731e321b06f5f88a25ff42199feb578691.

BUG=490237

Review URL: https://codereview.chromium.org/1147083002

Cr-Commit-Position: refs/heads/master@{#330803}

This CL prevents sync from updating the "last updated time" for bookmark
favicons each time that Chrome is started. The "last updated time" determines
when the favicon is redownloaded from the web (in case it has been updated).

Sync calls HistoryBackend::MergeFavicon() for each bookmark favicon on startup.
This CL stops HistoryBackend::MergeFavicon() from updating the "last updated
time" when MergeFavicon() is called with the same bitmap data as already stored
in the database (no new bitmap data).

BUG=481414
TEST=TwoClientBookmarksSyncTest.SC_UpdatingTitleDoesNotUpdateFaviconLastUpdatedTime

Review URL: https://codereview.chromium.org/1128343004

Cr-Commit-Position: refs/heads/master@{#330802}

BUG=419410

Review URL: https://codereview.chromium.org/1142313002

Cr-Commit-Position: refs/heads/master@{#330801}

BUG=369661
TEST=none

Review URL: https://codereview.chromium.org/1143333003

Cr-Commit-Position: refs/heads/master@{#330800}

BUG=490260,490240
TEST=as in bug

Review URL: https://codereview.chromium.org/1148943002

Cr-Commit-Position: refs/heads/master@{#330799}

Removes some build warnings.

Moves the class-dump repo to a github auto-mirror instead of a mirror
that needs to be manually updated.

This also requires that the class-dump.gyp file move from the
third-party repo into the main source tree. Note that the file
structure of the class-dump source changed so the .gyp had to be
updated. For comparison, the old .class-dump.gyp can be found at
https://chromium.googlesource.com/chromium/deps/class-dump/+/master/class-dump.gyp

BUG=318160

Review URL: https://codereview.chromium.org/1106993002

Cr-Commit-Position: refs/heads/master@{#330798}

BUG=

Review URL: https://codereview.chromium.org/1147933002

Cr-Commit-Position: refs/heads/master@{#330797}

This reverts commit a2a44ba5.

BUG=316085

TBR=tdresser@chromium.org,sadrul@chromium.org,ananta@chromium.org,girard@chromium.org

Review URL: https://codereview.chromium.org/1143333002

Cr-Commit-Position: refs/heads/master@{#330796}

Blink rebaseline in progress.

BUG=490188
TBR=mtklein@chromium.org
NOTRY=true

Review URL: https://codereview.chromium.org/1148283003

Cr-Commit-Position: refs/heads/master@{#330795}

BUG=490056

Review URL: https://codereview.chromium.org/1148203002

Cr-Commit-Position: refs/heads/master@{#330794}

Flaky failures on the bots.

BUG=378574
TBR=tobiasjs@chromium.org
NOTRY=true

Review URL: https://codereview.chromium.org/1148873006

Cr-Commit-Position: refs/heads/master@{#330793}

We previously weren't.

BUG=none
TEST=none
R=msw@chromium.org

Review URL: https://codereview.chromium.org/1147093002

Cr-Commit-Position: refs/heads/master@{#330792}

Since some platforms are still using NSS for now, this change mirrors https://boringssl-review.googlesource.com/#/c/4813/ in NSS.

BUG=490240

Review URL: https://codereview.chromium.org/1143303002

Cr-Commit-Position: refs/heads/master@{#330791}

BUG=489756
TEST=Mojo media services still working in html_viewer and chrome browser.

Review URL: https://codereview.chromium.org/1134003009

Cr-Commit-Position: refs/heads/master@{#330790}

This CL is dedicated to fix above issue and achieve below behaviors:
1.  User Menu does not close during sign-in flow: Things can't steal focus; the user menu stays open while you're typing / in the signin flow.
2.  Sign-in confirmation is modal: The user menu is not closeable until the user interacts with the signin confirmation window.  Sync should be delayed until the user interacts with the dialogue.
3.  Blue information box: This shouldn't go away until the user explicitly interacts with it (not you, what's new, "x").  Doesn't affect modality/close-ability of the user menu.

BUG=341196

Review URL: https://codereview.chromium.org/1136693002

Cr-Commit-Position: refs/heads/master@{#330789}

Summary of changes available at:
https://chromium.googlesource.com/chromium/blink/+log/1d565c2..6bd0043

TBR=szager@chromium.org,leviw@chromium.org,samli@chromium.org

Review URL: https://codereview.chromium.org/1143313002

Cr-Commit-Position: refs/heads/master@{#330788}

This patch removes some redundant code (fixing a bunch of TODOs). Also
ensures that we only call pending_visible_rect once per iterator (since
the function involves a virtual call).

R=enne
BUG=488636

Review URL: https://codereview.chromium.org/1144973005

Cr-Commit-Position: refs/heads/master@{#330787}

The "LocalDelta" version of ScrollLayer in LTHI takes a scroll delta in
viewport space and scrolls a layer in its local space. This means that
the delta is scaled so that the scroll appears to be the same distance
on the user's screen, regardless of the page scale factor (pinch-zoom).
This method returns the amount consumed so that the caller can determine
how much of the original scroll delta remains to bubble. However, the
returned consumed delta is in local space, meaning we were subtracting a
layer space delta from the viewport space delta.

The practical effect of this was that when the page was scaled up the
bubbling logic in the viewport would inappropriately bubble up scrolls to
the inner (visual) viewport since the local delta is divided by the page
scale factor. This caused position:fixed elements to "paralax" scroll
with the page, cool looking but incorrect.

BUG=488549

Review URL: https://codereview.chromium.org/1143233003

Cr-Commit-Position: refs/heads/master@{#330786}

Currently TabCapturePerformanceTest records the time between "Capture"
events to determine the rate of capture. These events are issued whenever
a capture is requested, even if it is later aborted, and doesn't represent
the actual capture framerate. This CL adds an event on capture
success which will provide a more accurate idea of capture framerate.

BUG=489796

Review URL: https://codereview.chromium.org/1143723003

Cr-Commit-Position: refs/heads/master@{#330785}

ChromeSRMDelegate should not be reading the WebContents* back from
  WebContentsObserver after it has explicitly set to observe nullptr
  WebContents.
This was introduced by my earlier CL: r330213, I never tested the
  feature from NTP.

BUG=489182, 489157
Test=1) From new tab page, click on the mic beside search box,
  feed it some search query, it should not crash chrome.
2) Similar to step #1, but before the query completes, close the
  tab, observe no crash.

Review URL: https://codereview.chromium.org/1134773004

Cr-Commit-Position: refs/heads/master@{#330784}

This removes boilerplate code that was used when we supported building libpeerconnection as a shared library.  We don't support that any longer, so it's safe to delete.

BUG=463660

Review URL: https://codereview.chromium.org/1146123002

Cr-Commit-Position: refs/heads/master@{#330783}

BUG=475633

Review URL: https://codereview.chromium.org/1132353003

Cr-Commit-Position: refs/heads/master@{#330782}

BUG=475219

Review URL: https://codereview.chromium.org/1143933002

Cr-Commit-Position: refs/heads/master@{#330781}

Revert of Add SyncScheduler for scheduling CryptAuth enrollments and syncing devices. (patchset #3 id:80001 of https://codereview.chromium.org/1147563002/)

Reason for revert:
MSan and Valgrind failures (use-of-uninitialized memory).

Original issue's description:
> Add SyncScheduler for scheduling CryptAuth enrollments and syncing devices.
>
> BUG=420315
> TEST=unit test
>
> Committed: https://crrev.com/27e044cfc5c2a43b6c575508aca64763bfa2deba
> Cr-Commit-Position: refs/heads/master@{#330713}

TBR=isherman@chromium.org,tengs@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=420315

Review URL: https://codereview.chromium.org/1146193002

Cr-Commit-Position: refs/heads/master@{#330780}

R=reed@google.com,mtklein@google.com
BUG=490188

Review URL: https://codereview.chromium.org/1143923003

Cr-Commit-Position: refs/heads/master@{#330779}

Revert of Change WebContents::last_active_time_ to Time instead of Timeticks. (patchset #1 id:140001 of https://codereview.chromium.org/1140083004/)

Reason for revert:
Reverting for georgesak@ (his committer access hasn't propagated yet it looks like).

After discussion with sky@, decided to keep TimeTicks.

Original issue's description:
> Change WebContents::last_active_time_ to Time instead of Timeticks.
>
> For context, last_active_time_ is going to be used by session restore to order the loading of background tabs using MRU. In order for this to be robust, last_active_time_ must be saved and restore between sessions. Timeticks cannot be reliably restored as it's dependent on the current OS session. MRU code for session restore is being implemented in https://codereview.chromium.org/1131373003
>
> Notes:
> - In dev tools, replaced "activity" with "active" for consistency
> - In OomPriorityManagerTest.Comparator, initialized last_active_time for all tabs, as this is necessary with Time (a default TimeTicks can go back in time, not the case with Time).
>
> BUG=472772
>
> Committed: https://crrev.com/041ef9c96d9a5b6f206a24385b0e6e4b3dbf9f20
> Cr-Commit-Position: refs/heads/master@{#330029}

TBR=jamescook@chromium.org,jam@chromium.org,georgesak@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=472772

Review URL: https://codereview.chromium.org/1145233002

Cr-Commit-Position: refs/heads/master@{#330778}