GitLab

In r404402 I created the outline of the UpdateTestExpectations script that
removes lines from TestExpectations that are no longer flaky according to
results on the build waterfall. This CL actually implements the functionality
in the script and makes it basically functional.

BUG=595414

Review-Url: https://codereview.chromium.org/2125633002
Cr-Commit-Position: refs/heads/master@{#404515}

It's because Linux ATI bot fails
WebglConformance.conformance_textures_misc_tex_image_webgl
When finding the solution for ATI driver, rollback this patch.

BUG=557848
TEST=WebglConformance.conformance_textures_misc_tex_image_webgl
CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_rel;tryserver.chromium.win:win_optional_gpu_tests_rel

Review-Url: https://codereview.chromium.org/2131863003
Cr-Commit-Position: refs/heads/master@{#404514}

When CMA media pipeline is started we post a task that will continuouly
update media time. But since that task is posted, it won't run right
away. In the meantime somebody might call GetMediaTime and will get
an invalid value (kNoTimestamp). To avoid this we can just set the
last_media_time_ right away for the first time when starting playback.
This will ensure that we have valid media time even before the first
UpdateMediaTime task invocation.

BUG=internal b/30035825

Review-Url: https://codereview.chromium.org/2131413002
Cr-Commit-Position: refs/heads/master@{#404513}

Modify the aggregate_vector_icons step in
ui/gfx/BUILD.gn to accept a response file
containing only the desired .icon files to
process rather than have every build configuration
process and include all .icon files. This will
make it possible to separate icons into
subdirectories and conditionally process them
on a per-platform basis.

BUG=535386
TEST=builds with GN and GYP

Review-Url: https://codereview.chromium.org/2129683002
Cr-Commit-Position: refs/heads/master@{#404512}

https://chromium.googlesource.com/external/khronosgroup/webgl.git/+log/72e7ed9..4da7838

BUG=
TBR=cwallez@chromium.org
TEST=bots
NOTRY=true

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

Review-Url: https://codereview.chromium.org/2133693004
Cr-Commit-Position: refs/heads/master@{#404511}

This also removes the special case of ignoring verdicts for CRX, which
by my read is not possible, and UMA says has been hit once in the last
week (weird).

Without this, debug builds will DCHECK for some downloads soon.

BUG=581044

Review-Url: https://codereview.chromium.org/2128583005
Cr-Commit-Position: refs/heads/master@{#404510}

Sprinkle in some const&s and explicit ctors in a few places that jumped
out.

BUG=None

Review-Url: https://codereview.chromium.org/2133203002
Cr-Commit-Position: refs/heads/master@{#404509}

ScrollAnchor can now have a RootFrameViewport as its scroller so
that an anchoring adjustment can be distributed between layout
and visual viewports.

BUG=594876

Review-Url: https://codereview.chromium.org/2123053002
Cr-Commit-Position: refs/heads/master@{#404508}

Currently, the test runner only takes action on SIGTERM on test execution
threads and only stops at the end of a test shard. In some cases, this isn't
fast enough. This CL makes the main thread drop test execution.

This CL also tweaks the handling of test results to better handle UNKNOWNs.

BUG=

Review-Url: https://codereview.chromium.org/2134523002
Cr-Commit-Position: refs/heads/master@{#404507}

StrictMode violation was from Calendar.getInstance() reading
the /etc/timezone file. Whitelist the StrictMode violation
to be fixed.

Tests were flaky because the UI thread would not always launch
the InfoBar before calling addInfoBarAnimationFinished(). Run
the UI thread blocking so that the InfoBar has always been
launched.

BUG=625038, 577185

Review-Url: https://codereview.chromium.org/2130083002
Cr-Commit-Position: refs/heads/master@{#404506}

Record user action when:
- User clicks the update menu item in the system menu

BUG=620508
TEST=manual

Review-Url: https://codereview.chromium.org/2125403002
Cr-Commit-Position: refs/heads/master@{#404505}

https://chromium.googlesource.com/angle/angle.git/+log/8bad46d..e8d5c5c

BUG=angleproject:1439

TEST=bots

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

Review-Url: https://codereview.chromium.org/2133183003
Cr-Commit-Position: refs/heads/master@{#404504}

Also move the syzygy_optimize and syzygy_asan GN templates to //build/win/syzygy so they can be reused in //content (without adding a dependency with //chrome).

BUG=619086

Review-Url: https://codereview.chromium.org/2126673002
Cr-Commit-Position: refs/heads/master@{#404503}

It is no longer necessary, since the browser compositor no longer needs access
to the underlying GpuMemoryBuffers.

This CL is a revert of "Expose GpuMemoryBufferId through glGetImageivCHROMIUM."
> WebKit widgets that provide textures to TextureLayer need to also provide the
> GpuMemoryBufferId if the texture is backed by a GpuMemoryBuffer. This CL
> modifies the CHROMIUM Image extension to allow the GpuMemoryBuffer id to be
> obtained by a new function glGetImageivCHROMIUM.

BUG=608026
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel

Review-Url: https://codereview.chromium.org/2134793002
Cr-Commit-Position: refs/heads/master@{#404502}

BUG=

Review-Url: https://codereview.chromium.org/2135443002
Cr-Commit-Position: refs/heads/master@{#404501}

Mousewheel-based scroll bubbling from OOPIFs to parent frames was broken
in a couple of ways, due to it improperly accounting for the mechanics
of synthetic gesture scroll events from the InputRouter class. This
refactors bubbling to do the following:
- remove bubbling of unused scroll delta from WheelEvent acks, which was
  wrong,
- pass acks from GestureScrollUpdate and GestureScrollEnd events to
  the CrossProcessFrameConnector and then have RWHInputEventRouter do
  most of the reasoning about where to target unused scroll,
- track the current scroll bubbling target, which can change due to
  nested OOPIFs, and manage GestureScrollBegin and GestureScrollEnds
  to each renderer (the previous approach of wrapping each
  GestureScrollUpdate led to bubbled scrolls feeling really slow).

BUG=621624
CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_site_isolation

Review-Url: https://codereview.chromium.org/2123843002
Cr-Commit-Position: refs/heads/master@{#404500}

BUG=607682

Review-Url: https://codereview.chromium.org/2120613002
Cr-Commit-Position: refs/heads/master@{#404499}

This CL times deserialization of the CompositorFrame generated by cc_serialization_perftest using StructTraits
versus ParamTraits.

This CL also makes some mojo changes to ensure we don't need to move the mojo::Array we are deserializing from.

BUG=624459
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel

Review-Url: https://codereview.chromium.org/2136593002
Cr-Commit-Position: refs/heads/master@{#404498}

In the next CL, GetECT API which computes Effective
Connection Type (ECT) based on transport RTT and downlink
throughput will be exposed to Cronet.

BUG=625305

Review-Url: https://codereview.chromium.org/2113363002
Cr-Commit-Position: refs/heads/master@{#404497}

BUG=602484

Review-Url: https://codereview.chromium.org/2102273002
Cr-Commit-Position: refs/heads/master@{#404496}

At this point, VirtualGLApi only hooks the extension string query, and for not a very good reason, since it ends up serving the extension string of the underlying context.
OTOH it adds a bunch of complexity to require the switch between RealGLApi and the VirtualGLApi, so instead fold the logic into GLContext.

BUG=None
CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_rel;tryserver.chromium.win:win_optional_gpu_tests_rel

Review-Url: https://codereview.chromium.org/2132913002
Cr-Commit-Position: refs/heads/master@{#404495}

Now that these builders are using ninja, there isn't a compelling
reason not to also be using goma.

R=smut@chromium.org
NOTRY=true
BUG=626005

Review-Url: https://codereview.chromium.org/2127963003
Cr-Commit-Position: refs/heads/master@{#404494}

Until now, we've used base::TimeDelta.is_zero() as a signal that
a given timing is unset.

This is technically incorrect, as a TimeDelta can be set to the
zero value.

This mostly presented issues that made writing tests more complex,
but also addresses real issues where a metric reports a zero time
delta that would have previously been ignored but will now be
included in our metrics.

BUG=623556

Review-Url: https://codereview.chromium.org/2111073003
Cr-Commit-Position: refs/heads/master@{#404493}

Remove public methods from Layer and LayerTreeHost that are just a pass
through to the AnimationHost. These are used by the PropertyTreeBuilder,
have it query the AnimationHost directly.

BUG=625384
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel

Review-Url: https://codereview.chromium.org/2130053002
Cr-Commit-Position: refs/heads/master@{#404492}

Currently the image pattern shader created is only padded
with 1 black transparent pixel right and bottom. We should
pad it on the left and top as well so that when pattern is
transformed, the color on its left and top will not bleed.

BUG=602509

Review-Url: https://codereview.chromium.org/2130063002
Cr-Commit-Position: refs/heads/master@{#404491}

Revert of Add support for installing ARM64 sysroot (patchset #1 id:1 of https://codereview.chromium.org/2078293002/ )

Reason for revert:
https://bugs.chromium.org/p/chromium/issues/detail?id=626784

Original issue's description:
> Add support for installing ARM64 sysroot
>
> This change modifies install-sysroot.py to add support for
> installing ARM64 sysroot. Also update the jessie arm64 package list.
>
> BUG=613452
>
> Committed: https://crrev.com/ad891a49ee7aeff1a42559c0a059cb51f053e1fb
> Cr-Commit-Position: refs/heads/master@{#404323}

TBR=thestig@chromium.org,mmoss@chromium.org,thakis@chromium.org,apatole@nvidia.com
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=613452

Review-Url: https://codereview.chromium.org/2132153002
Cr-Commit-Position: refs/heads/master@{#404490}

Also shuffle around some JSDoc @tag-related code

R=dpapad@chromium.org
BUG=567770
NOTRY=true

Review-Url: https://codereview.chromium.org/2131373002
Cr-Commit-Position: refs/heads/master@{#404489}

All suites now use isolate files that are either generated from gn
data / data_deps or checked in to chromium.

BUG=

Review-Url: https://codereview.chromium.org/2131083003
Cr-Commit-Position: refs/heads/master@{#404488}

This will be used to periodically gather per-test run times for various
test suites, to improve the sharding distribution.

Check in the data for a recent run of webgl2_conformance_tests.
Reference this from the Swarming invocations.

BUG=352807
CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_rel;tryserver.chromium.win:win_optional_gpu_tests_rel
NOTRY=true

Review-Url: https://codereview.chromium.org/2116593004
Cr-Commit-Position: refs/heads/master@{#404487}

Add OnStatisticsUpdate and OnWaitingForDecryptionKey to
mojom::RendererClient, and call these callbacks from
MojoRendererService. These are needed for Cast Apps like Netflix that
depend on decoding statistics from the media pipeline.

BUG=585287
BUG= internal b/29390639

Review-Url: https://codereview.chromium.org/2101043003
Cr-Commit-Position: refs/heads/master@{#404486}

The resize lock completely throws off tests that generate
events. These tests aren't meant to exercise the resize lock, so
disabling seems reasonable. This is just the plumbing for the disable,
will start enabling tests separately.

BUG=331924
TEST=none
R=piman@chromium.org

Review-Url: https://codereview.chromium.org/2135743002
Cr-Commit-Position: refs/heads/master@{#404485}

TBR=zmo@chromium.org
BUG=626524
BUG=626742
CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_optional_gpu_tests_rel;tryserver.chromium.mac:mac_optional_gpu_tests_rel;tryserver.chromium.win:win_optional_gpu_tests_rel

Review-Url: https://codereview.chromium.org/2134643004
Cr-Commit-Position: refs/heads/master@{#404484}

This patch fixes a crash on https://codepen.io where the paint offset
in PaintPropertyTreeBuilder::updateSvgLocalToBorderBoxTransform would
not be reset if transformToPixelSnappedBorderBox was identity.

The new testcase, SvgPixelSnappingShouldResetPaintOffset, both ensures
the DCHECK in PaintPropertyTreeBuilder::updateTransform is not hit, and
verifies the transform property tree is correct in this case.

BUG=600618

Review-Url: https://codereview.chromium.org/2131803002
Cr-Commit-Position: refs/heads/master@{#404483}

Bug was caused by having transitions for the image and sign in indicator for the user pod, while the image pane stays 100%. So when we are in the public account user pod the image will be centered.

BUG=626157

Review-Url: https://codereview.chromium.org/2130143002
Cr-Commit-Position: refs/heads/master@{#404482}

* Introduce lazy allocation of Storage.  This in particular allows moved-from
SpdyHeaderBlock instances to be valid without having to allocate Storage for
them.
* Use swap in move constructor and move assignment operator to make sure
moved-from instance is valid.
* Add SpdyHeaderBlockTest.MovedFromIsValid test.
* s/block1/block/ in SpdyHeaderBlockTest.AddHeaders.

This CL lands server change 125821596 by bnc.

BUG=488484

Review-Url: https://codereview.chromium.org/2125403003
Cr-Commit-Position: refs/heads/master@{#404481}

TBR=brettw@chromium.org

Review-Url: https://codereview.chromium.org/2133083002
Cr-Commit-Position: refs/heads/master@{#404480}

BUG=624651

Review-Url: https://codereview.chromium.org/2124143004
Cr-Commit-Position: refs/heads/master@{#404479}

The children of non-editable combo boxes, search boxes and text fields specified via e.g. aria-owns should not be hidden from platform trees.

Children of combo boxes, search boxes and text fields might not always be part of the shadow DOM, especially when no <input> or contentEditable . tags have been usedThey might be a list of combo box items specified via aria-owns.
data:text/html,<div role="combobox" aria-owns="list" onKeyDown="this.setAttribute('aria-activedescendant', 'item');">content</div><ul id="list"><li id="item">item</li></ul>
BUG=620432
R=dmazzoni@chromium.org
TESTED=manually with Jaws NVDA and Voiceover

Review-Url: https://codereview.chromium.org/2120523002
Cr-Commit-Position: refs/heads/master@{#404478}

BUG=616044
R=kozyatinskiy

Review-Url: https://codereview.chromium.org/2135703003
Cr-Commit-Position: refs/heads/master@{#404477}

BUG=625704

Review-Url: https://codereview.chromium.org/2126353002
Cr-Commit-Position: refs/heads/master@{#404476}