GitLab

BUG=338303

Review URL: https://codereview.chromium.org/560523004

Cr-Commit-Position: refs/heads/master@{#296188}

This is particularly important on Mac, where unless we do this the Bookmarks/History
menu bar items will reflect the last active profile (which has now been locked).

On Windows this is a nice to have and to be consistent. Overall this adds an extra
layer of security that you won't accidentally open a locked profile's window.

BUG=412952

Review URL: https://codereview.chromium.org/585653002

Cr-Commit-Position: refs/heads/master@{#296187}

BUG=399456

Review URL: https://codereview.chromium.org/575213002

Cr-Commit-Position: refs/heads/master@{#296186}

The stack trace suggests LSOpenApplication fails to find the shim, and this
happens after waiting on a run loop. So a likely cause is that we're not
successfully waiting for the file operations to finish.

This CL changes the test to observe the parent directory instead of the shim
directory, and adds checks that the path exists before proceeding. If the test
still fails for this reason, at least it'll say so.

BUG=415422

Review URL: https://codereview.chromium.org/598443002

Cr-Commit-Position: refs/heads/master@{#296185}

BUG=371012

Review URL: https://codereview.chromium.org/578423002

Cr-Commit-Position: refs/heads/master@{#296184}

BUG=78512
TEST=no change

Review URL: https://codereview.chromium.org/565263005

Cr-Commit-Position: refs/heads/master@{#296183}

BUG=331829,373148,299841,103304
TEST=ninja -C out/Debug ui_unittests, then verify that
out/Debug/ui_base_unittests is there and runs. Same thing with
ui_unittests
R=sky@chromium.org

Review URL: https://codereview.chromium.org/588963002

Cr-Commit-Position: refs/heads/master@{#296182}

This adds a missing "_" suffix to |ozone_buffer_factory| member
variable in ozone implementation of GpuMemoryBufferFactory interface.

BUG=
TBR=alexst@chromium.org

Review URL: https://codereview.chromium.org/594043004

Cr-Commit-Position: refs/heads/master@{#296181}

This reverts commit 41c3d488, CL https://codereview.chromium.org/489103004.

The above change caused a couple of UI issues, as well as internationalization troubles (http://crbug.com/377410#c10). We should land it again after all the fixes get incorporated.

BUG=416754,416818,416825,416757,377410
TBR=engedy@chromium.org,pkasting@chromium.org,btapiz@gmail.com,danbeam@chromium.org

Review URL: https://codereview.chromium.org/589353003

Cr-Commit-Position: refs/heads/master@{#296180}

TBR=machenbach@chromium.org

Review URL: https://codereview.chromium.org/595883002

Cr-Commit-Position: refs/heads/master@{#296179}

Summary of changes available at:
https://chromium.googlesource.com/chromium/blink/+log/3cba268..63c0786

TBR=pdr@chromium.org,wjmaclean@chromium.org,ager@chromium.org

Review URL: https://codereview.chromium.org/600433002

Cr-Commit-Position: refs/heads/master@{#296178}

Telemetry leaves a memory hogging subprocess behind if:
1) A Monsoon Power Monitor is attached to host computer (Linux),
2) User runs a power-metric-using benchmark such as sunspider on Android
   ChromeShell, and
3) Benchmark raises an exception between power metric's Start and Stop
   calls. For example, an exception will occur if browser crashes and
   the benchmark does something like tab.WaitForJavaScriptExpression().

This patch fixes the issue by setting the daemon flag on the subprocess.
This will cause the main process to kill the subprocess on exit.

BUG=413596

R=ernstm@chromium.org

Review URL: https://codereview.chromium.org/572643002

Cr-Commit-Position: refs/heads/master@{#296177}

For DrawQuads and SharedQuadState, RenderPass used allocate them one by
one whenever needed. This new class helps RenderPass manages allocation
and iteration of those two types. This container allocates a chunk of
memory at one time and hands out raw pointers. It also provides iterator
and reverse iterators for going through its contents.

Unittest for ListContainer makes sure the raw pointers it hands out are
valid and iterator has same behavior as vector iterators.

Follows 398533002, and 404563005.

BUG=344962

Committed: https://crrev.com/6ae8c50c839a8c0d8fb399578256c754ed3ce39a
Cr-Commit-Position: refs/heads/master@{#296100}

Review URL: https://codereview.chromium.org/400463002

Cr-Commit-Position: refs/heads/master@{#296176}

It observes OnProfilesChanged, but the set of profiles changing doesn't
affect the start page. However, it probably once helped avoid crashes
when deleting profiles since the AppListViewDelegate would hold on to
the Profile* until the app list was shown again. But that was fixed in
http://crrev.com/291852 (now AppListServiceImpl::DestroyAppList() will
reliably tear down the AppListView if the profile is destroyed, without
having to wait for a ShowAppList().

When regular profile switching occurs (i.e. not when the profile is
being destroyed), the profile currently being used by the
StartPageView is also replaced. That is handled by
AppListMainView::ModelChanged() calling into
ContentsView::InitNamedPages() which makes an entirely new
StartPageView().

BUG=403647

Review URL: https://codereview.chromium.org/589333002

Cr-Commit-Position: refs/heads/master@{#296175}

BUG=416843
TBR=oshima@chromium.org

Review URL: https://codereview.chromium.org/594853002

Cr-Commit-Position: refs/heads/master@{#296174}

BUG=412402

Review URL: https://codereview.chromium.org/592353002

Cr-Commit-Position: refs/heads/master@{#296173}

Cr-Commit-Position: refs/heads/master@{#296172}

Summary of changes available at:
https://chromium.googlesource.com/chromium/blink/+log/cbea1bc..3cba268

TBR=pdr@chromium.org,wjmaclean@chromium.org,ager@chromium.org

Review URL: https://codereview.chromium.org/595753003

Cr-Commit-Position: refs/heads/master@{#296171}

BUG=none
TEST=manual

Review URL: https://codereview.chromium.org/597523002

Cr-Commit-Position: refs/heads/master@{#296170}

Revert of Update V8 to version 3.29.82 (based on bleeding_edge revision r24129). (patchset #1 id:1 of https://codereview.chromium.org/596703003/)

Reason for revert:
Unfortunately this causes layout test failures on ia32 systems. It makes:

http/tests/serviceworker/request.html

time out.

http://test-results.appspot.com/dashboards/flakiness_dashboard.html#master=ChromiumWebkit&tests=http/tests/serviceworker/request.html

Original issue's description:
> Update V8 to version 3.29.82 (based on bleeding_edge revision r24129).
>
> TBR=machenbach@chromium.org
>
> Committed: https://crrev.com/7e6bdb0cf9ef794af7ef2c76a83337d95abd4ac0
> Cr-Commit-Position: refs/heads/master@{#296166}

TBR=machenbach@chromium.org,v8-autoroll@chromium.org
NOTREECHECKS=true
NOTRY=true

Review URL: https://codereview.chromium.org/594043003

Cr-Commit-Position: refs/heads/master@{#296169}

Summary of changes available at:
https://chromium.googlesource.com/chromium/blink/+log/1fcceef..cbea1bc

TBR=pdr@chromium.org,wjmaclean@chromium.org,ager@chromium.org

Review URL: https://codereview.chromium.org/588403005

Cr-Commit-Position: refs/heads/master@{#296168}

Summary of changes available at:
https://chromium.googlesource.com/chromium/blink/+log/44b472e..1fcceef

TBR=pdr@chromium.org,wjmaclean@chromium.org,ager@chromium.org

Review URL: https://codereview.chromium.org/597513002

Cr-Commit-Position: refs/heads/master@{#296167}

TBR=machenbach@chromium.org

Review URL: https://codereview.chromium.org/596703003

Cr-Commit-Position: refs/heads/master@{#296166}

BUG=none

Review URL: https://codereview.chromium.org/592753005

Cr-Commit-Position: refs/heads/master@{#296165}

This change ensures that a WebRTC client can detect (and log) if a user has enabled the microphone mute. No platform-dependent code is involved.

BUG=NONE
TEST=Manual tests of WebRTC client on Linux, Mac and Windows.

Review URL: https://codereview.chromium.org/582733002

Cr-Commit-Position: refs/heads/master@{#296164}

r1384 Fix clang compilation error introduced in r1380.
r1381 Add inttypes for windows in dump_context
r1380 Extend mapping merge to include reserved but unused mappings.
r1379 GoogleCrashdumpUploader: adds Upload(string*) API to get the HTTP response.
r1378 Fixes common_unittests build.
r1377 Fix Windows client unittests build: add missing code files
r1376 Making the Mac build of crash_report work again after the last SVN sync

BUG=394703

Review URL: https://codereview.chromium.org/595493002

Cr-Commit-Position: refs/heads/master@{#296163}

R=isherman@chromium.org
BUG=84200

Review URL: https://codereview.chromium.org/442403002

Cr-Commit-Position: refs/heads/master@{#296162}

Summary of changes available at:
https://chromium.googlesource.com/chromium/blink/+log/082de98..44b472e

TBR=pdr@chromium.org,wjmaclean@chromium.org,ager@chromium.org

Review URL: https://codereview.chromium.org/589383002

Cr-Commit-Position: refs/heads/master@{#296161}

Propagate the user's supervised state to the ash system
so that it becomes available for views

https://chromereviews.googleplex.com/79527013/

BUG=394417

Committed: https://crrev.com/8616ac4cf7bd77f53dc001eb9d8d61c6c975eaaa
Cr-Commit-Position: refs/heads/master@{#294821}

Review URL: https://codereview.chromium.org/561713002

Cr-Commit-Position: refs/heads/master@{#296160}

BUG=115047

Review URL: https://codereview.chromium.org/592223002

Cr-Commit-Position: refs/heads/master@{#296159}

Revert of Fix bad scaling in TiledLayer (patchset #8 id:140001 of https://codereview.chromium.org/567743003/)

Reason for revert:
This break Blink layout tests. See crbug.com/416775.

Original issue's description:
> Fix bad scaling in TiledLayer
>
> On the windows side (and likely chromeos for some devices) we can end
> up in a situation where the pixel bounds is not an even multiple of the
> scale factor. This means when TiledLayer calculates the scale it ends
> up off.
>
> BUG=416655
> TEST=added test coverage
> R=danakj@chromium.org
>
> Committed: https://crrev.com/c01746b4f7d811a5fdc75ce6570a9667b9877913
> Cr-Commit-Position: refs/heads/master@{#296108}

TBR=danakj@chromium.org,sky@chromium.org
NOTREECHECKS=true
NOTRY=true
BUG=416655

Review URL: https://codereview.chromium.org/590313004

Cr-Commit-Position: refs/heads/master@{#296158}

HTML Tags: abbr, button, col, figure, meta, meter, input type(search, time)
Aria role: log, marquee
Aria state/property: aria-orientation(horizontal, vertical), aria-sort(Ascending, Descending, Other)

BUG=170579

Review URL: https://codereview.chromium.org/569203002

Cr-Commit-Position: refs/heads/master@{#296157}

Summary of changes available at:
https://chromium.googlesource.com/chromium/blink/+log/a1f6e7b..082de98

TBR=pdr@chromium.org,wjmaclean@chromium.org,ager@chromium.org

Review URL: https://codereview.chromium.org/597503002

Cr-Commit-Position: refs/heads/master@{#296156}

This patch creates new URLLoaderResource objects for each request
since those objects can't be reused.

Calling Open() on an instance of URLLoaderResource requires that mode_
is set to MODE_WAITING_TO_OPEN. After Open() has been called mode_
will never reach that state again.

BUG=None

Review URL: https://codereview.chromium.org/582623002

Cr-Commit-Position: refs/heads/master@{#296155}

Smart clip rect output needs to be offsetted when there is a change in
viewport, e.g., when location bar is shown.

BUG=414597
TEST=verified by OEM that this fixes the bug

Review URL: https://codereview.chromium.org/565383005

Cr-Commit-Position: refs/heads/master@{#296154}

Summary of changes available at:
https://chromium.googlesource.com/chromium/blink/+log/d76db5d..a1f6e7b

TBR=caseq@chromium.org,pdr@chromium.org,wjmaclean@chromium.org

Review URL: https://codereview.chromium.org/596663002

Cr-Commit-Position: refs/heads/master@{#296153}

r295646 moved ownership of the AppListViewDelegate out of the
AppListViewController which meant that the Mac AppList began removing
itself as an observer from the wrong model. This happens because, in
Chrome, the AppListViewDelegate changes the return from GetModel() when
the profile is switched, but doesn't currently have a way of notifying
the AppList that this is happening.

To fix, AppListServiceMac needs to explicitly clear the delegate on the
Chrome side before asking the view delegate to switch out its model
(which happens via AppListServiceImpl::GetViewDelegate(Profile*).

Views platforms are not affected, because they retain a raw pointer to
the AppListModel* separately from AppListViewDelegate::GetModel(). This
only works because the AppListModel* sticks around until Chrome
shutdown, when it's torn down with the other keyed services via the
~ProfileManager() destructor.

This delayed destruction means that verifying observers in a test needs
a lot of plumbing in non-test code (see patchset1). Instead, a DCHECK is
added in AppListItemList to ensure RemoveObserver always removes
something, then test coverage is given by
AppListServiceImplBrowserTest.ShowLoadedProfiles.

This CL fixes part of the flakiness in the app list interactive uitests
on Mac (http://crbug.com/415264), but I've discovered other sources of
flakiness to fix before re-enabling those tests. It should, however, fix
the recent crashes (http://crbug.com/416345).

BUG=415264, 416345
TEST=AppListServiceImplBrowserTest.ShowLoadedProfiles
TEST=crbug/416345: have multiple profiles and switch between them in the
app launcher on mac many times. Chrome shouldn't crash.

Review URL: https://codereview.chromium.org/592983002

Cr-Commit-Position: refs/heads/master@{#296152}

Cr-Commit-Position: refs/heads/master@{#296151}

By default both enabled.

TBR=gene@chromium.org

Review URL: https://codereview.chromium.org/598433002

Cr-Commit-Position: refs/heads/master@{#296150}

Summary of changes available at:
https://chromium.googlesource.com/chromium/blink/+log/356af1f..d76db5d

TBR=caseq@chromium.org,pdr@chromium.org,wjmaclean@chromium.org

Review URL: https://codereview.chromium.org/586393007

Cr-Commit-Position: refs/heads/master@{#296149}