GitLab

Single sample metrics offer a mechanism for clients to modify
a histogram sample repeatedly while having it only reported once;
either upon destruction of a helper class or at process termination.

This is helpful since the fast shutdown path simply kills renderer
processes without going through any destructors. In this case we
lose histogram values that might otherwise be reported at that
time.

In media/ code we've created a cumbersome proxy mechanism which
sends histogram values like this to the browser process and records
them there (see WatchTimeReporter and MediaInternals interactions).

Single sample histograms are implemented through a new mojo service
hosted by the browser process that receives samples and logs the
last received sample upon mojo channel closure.

base:: shims are provided so that these metrics can be created by
anyone in the renderer process. For non-renderer processes a default
implementation is provided without the mojo channel behavior.

Usage looks like the following:
  std::unique_ptr<base::SingleSampleMetric> metric(
      base::SingleSampleMetricsFactory::Get()->CreateCustomCountsMetric(
          "Media.VideoRenderer.CadenceChanges", 1, 10, 10));
  metric->SetSample(1);
  metric->SetSample(2);
  metric->SetSample(3);
  metric.reset();

Only the last sample (3) would end up being reported to the histogram.

BUG=689751
TEST=new tests

Review-Url: https://codereview.chromium.org/2687583002
Cr-Commit-Position: refs/heads/master@{#469524}

In https://codereview.chromium.org/2665913002 we set some default
network properties in currentRouteChanged. This was triggering
networkPropertiesChanged_ which udates autoConnect_ and
preferNetwork_ which triggers calls to setNetworkProperties.

To prevent this sort of error, set networkPropertiesReceived_
once properties are received from Chrome and do not call
networkingPrivate.setProperties until that has been set.

BUG=718389

Review-Url: https://codereview.chromium.org/2863713004
Cr-Commit-Position: refs/heads/master@{#469523}

This is a prerequisite of whitelisting //third_party/WebKit/public
for header checking in GN.

BUG=716359

Review-Url: https://codereview.chromium.org/2851953002
Cr-Commit-Position: refs/heads/master@{#469522}

This CL changes the logic for choosing the loud vs. quiet interstitial.
Instead of comparing against screen dimensions, we instead use window
dimensions, which makes more sense when Android Multi-Window is enabled.

BUG=718195

Review-Url: https://codereview.chromium.org/2861713005
Cr-Commit-Position: refs/heads/master@{#469521}

Manually convert tests to use new Audit

BUG=704967
TEST=
Panner/panner-automation-basic.html
Panner/panner-automation-position.html
Panner/panner-distance-clamping.html
Panner/panner-rolloff-clamping.html
Panner/pannernode-basic.html

Review-Url: https://codereview.chromium.org/2799793003
Cr-Commit-Position: refs/heads/master@{#469520}

The actual filenames are all lowercase, but src attributes in
execorder.html contains uppercase letters. This causes the scripts not
found and thus timeouts on Linux (which runs on a case-sensitive
filesystem).

This CL fixes the timeout by making the src attributes all lowercased.

BUG=594639

Review-Url: https://codereview.chromium.org/2856053005
Cr-Commit-Position: refs/heads/master@{#469519}

This CL removes a hack that adds /usr/lib/gcc/x86_64-linux-gnu/4.6 to
ld.so.conf.d/zz_hack.conf.  The only sysroots that exist currently are
Jessie and Trusty which both use gcc 4.8.

This caused the following flags to be passed to the linker:
-L/path/to/chromium/src/build/linux/debian_jessie_amd64-sysroot/usr/lib/gcc/x86_64-linux-gnu/4.6
-Wl,-rpath-link=/path/to/chromium/src/build/linux/debian_jessie_amd64-sysroot/usr/lib/gcc/x86_64-linux-gnu/4.6
While harmless, it is confusing and should be removed.

R=thestig@chromium.org

Review-Url: https://codereview.chromium.org/2863663003
Cr-Commit-Position: refs/heads/master@{#469518}

Not sure why/how this happens, but I was seeing this infrequently on  https://with.in/watch/under-neon-lights/

BUG=718501

Review-Url: https://codereview.chromium.org/2862883002
Cr-Commit-Position: refs/heads/master@{#469517}

And improve information provided by out-of-order destruction of task
scheduler and its provided constructs in unit tests.

(comes out of discussion on https://codereview.chromium.org/2852333004/)

BUG=708584, 689520
TBR=avi@chromium.org (documentation update in test_browser_thread_bundle.h)

Review-Url: https://codereview.chromium.org/2860063003
Cr-Commit-Position: refs/heads/master@{#469516}

ASAN wraps the kernel dylib on mac. Special case this to avoid tripping
a DCHECK.

BUG=718628
TBR=avi@chromium.org

Review-Url: https://codereview.chromium.org/2859303002
Cr-Commit-Position: refs/heads/master@{#469515}

Multi-display support in ARC was punted to the next milestone. This CL
temporarily prevents windows from being dragged to secondary displays
by using display rather than screen coordinates for window bounds.

BUG=714832
TEST=ARC apps cannot be dragged to secondary displays.

Review-Url: https://codereview.chromium.org/2860863004
Cr-Commit-Position: refs/heads/master@{#469514}

This is currently a blacklist entry, which means that it's calculated in
the browser and the resulting decision could be incorrect if the GPU
process uses a different GPU from what the browser expects. If this is
a GPU driver bug workaround entry then it'll be determined with more
information in the GPU process.

BUG=687004
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/2682723003
Cr-Commit-Position: refs/heads/master@{#469513}

The required texture formats (GL_R16F, etc) are not in the core until
3.0.

We could, in principle, query the particular extensions required for
these formats (GL_ARB_texture_rg, etc), but this query is simpler and
equally effective in practice.

BUG=699278
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/2863773002
Cr-Commit-Position: refs/heads/master@{#469512}

https://skia.googlesource.com/skia.git/+log/18f9602094ac..783c43658d91

$ git log 18f960209..783c43658 --date=short --no-merges --format='%ad %ae %s'
2017-05-04 egdaniel Fix comma in vulkan

Created with:
  roll-dep src/third_party/skia


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


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=scroggo@chromium.org

Change-Id: I12cb923ca36afc50220d839b0c13483ad2f616a1
Reviewed-on: https://chromium-review.googlesource.com/496887Reviewed-by: Skia Deps Roller <skia-deps-roller@chromium.org>
Commit-Queue: Skia Deps Roller <skia-deps-roller@chromium.org>
Cr-Commit-Position: refs/heads/master@{#469511}

BUG=706802

Review-Url: https://codereview.chromium.org/2846993002
Cr-Commit-Position: refs/heads/master@{#469510}

Now that we use the PaintOpBuffer, analyze whether a DisplayItemList
has any discardable images to skip metadata generation if unnecessary.

CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel

Review-Url: https://codereview.chromium.org/2830243002
Cr-Commit-Position: refs/heads/master@{#469509}

Make the order of setting parameters in ConstructSystemRequestContext
exactly match the order in URLRequestContextBuilder.  Also add a few
(trivial) setters to URLRequestContextBuilder for things that
ConstructSystemRequestContext sets that URLRequestContextBuilder does
not support.

This is part of a multi-stage operation to use URLRequestContextBuilder
to create the system URLRequestContext:

1)  Defrag (This CL).
2)  Switch to having a URLRequestContextStorage own everything, and create
everything inline with system URLRequestContext configuration.
3)  Switch the proxy service to using the system URLRequestContext.
4)  Switch to using URLRequestContextBuilder, but having IOThread create
  everything (This could be combined with 2, but 2 will affect a lot of
  accessors, so will probably to do it incrementally)
5)  Switch to having URLRequestContextBuilder create everything it
  reasonably can.

Hopefully the incremental approach will help to protect against regressions.

BUG=715695

Review-Url: https://codereview.chromium.org/2860813002
Cr-Commit-Position: refs/heads/master@{#469508}

R=kpaulhamus@chromium.org

Review-Url: https://codereview.chromium.org/2857813002
Cr-Commit-Position: refs/heads/master@{#469507}

BUG=647036
TBR=yzshen,dpranke,jbroman

Review-Url: https://codereview.chromium.org/2853293003
Cr-Commit-Position: refs/heads/master@{#469506}

Screenshot in bug.

R=dschuyler@chromium.org
BUG=718210

Review-Url: https://codereview.chromium.org/2862773003
Cr-Commit-Position: refs/heads/master@{#469505}

https://pdfium.googlesource.com/pdfium.git/+log/d41f4752c81d..b8a8c43f3942

$ git log d41f4752c..b8a8c43f3 --date=short --no-merges --format='%ad %ae %s'
2017-05-04 thestig CPDF_Document::GetPageData() normally does not return NULL.
2017-05-04 thestig Make CPDF_Font member variables protected.

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: I62224ef89307f373da6699d28dad3df96e6e658d
Reviewed-on: https://chromium-review.googlesource.com/496926
Reviewed-by: <pdfium-deps-roller@chromium.org>
Commit-Queue: <pdfium-deps-roller@chromium.org>
Cr-Commit-Position: refs/heads/master@{#469504}

CPU frequency file exists in [2] for a majority of devices. It can
also exist in [1] for certain devices like chell, cave, caroline,
veyron_speedy, etc. Add a parser to read data from [1] which has
different format from [2].

[1] /sys/devices/system/cpu/cpufreq/all_time_in_state
[2] /sys/devices/system/cpu/cpu*/cpufreq/stats/time_in_state

BUG=548510

Review-Url: https://codereview.chromium.org/2853863002
Cr-Commit-Position: refs/heads/master@{#469503}

This CL prevents flex pressure from squeezing paper-buttons. E.g. prevent
a long account name from making the "sign in to chrome" button into a
column of text.

BUG=718318
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:closure_compilation

Review-Url: https://codereview.chromium.org/2864483006
Cr-Commit-Position: refs/heads/master@{#469502}

BUG=718448

Review-Url: https://codereview.chromium.org/2857413002
Cr-Commit-Position: refs/heads/master@{#469501}

- move more things to the interface
- make offscreencanvas() only available for bindings

BUG=

Review-Url: https://codereview.chromium.org/2862923002
Cr-Commit-Position: refs/heads/master@{#469500}

This allows the hardware_class field in UMA to be populated on
iOS (with values like "iPhone9,3"), which will allow us to stop
sending those strings in the cpu_architecture field which is
what's currently being done.

BUG=370104

Review-Url: https://codereview.chromium.org/2860663005
Cr-Commit-Position: refs/heads/master@{#469499}

cc will soon reject frames with no render passes. Submit a frame with
a solid color quad of right size instead of an empty frame in
SurfacesInstance. Also, we just get rid of the empty frame submission in
HardwareRenderer because when SurfacesInstance updates its
references_surfaces the surface of HardwareRenderer will be garbage
collected and resources will be returned.

CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel

Review-Url: https://codereview.chromium.org/2856913002
Cr-Commit-Position: refs/heads/master@{#469498}

This patch updates the global interface listing after PaymentRequest was
enabled on additional platforms: Windows, Mac, ChromeOS, Linux.

blink-dev@ announcement:
https://groups.google.com/a/chromium.org/d/msg/blink-dev/p1DYoxHlkKg/g_ZW1YLDCgAJ

This patch also re-enables global interface listing tests, which should
never be disabled.

BUG=717914

Review-Url: https://codereview.chromium.org/2856193002
Cr-Commit-Position: refs/heads/master@{#469497}

BUG=718185

Review-Url: https://codereview.chromium.org/2859533004
Cr-Commit-Position: refs/heads/master@{#469496}

Adding #if and #endif to match our coding style.

BUG=

Review-Url: https://codereview.chromium.org/2863643003
Cr-Commit-Position: refs/heads/master@{#469495}

R=dschuyler@chromium.org
BUG=718345
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:closure_compilation

Review-Url: https://codereview.chromium.org/2862963002
Cr-Commit-Position: refs/heads/master@{#469494}

There is no need to discover services when receive an extra didConnectPeripheral
event from macOS.

BUG=681414

Review-Url: https://codereview.chromium.org/2853933002
Cr-Commit-Position: refs/heads/master@{#469493}

MRWebUIMessageHandler

This is a reland of http://crrev/2731033002/. There were DCHECKs with side
effects that should have been CHECKs.

BUG=684638

Review-Url: https://codereview.chromium.org/2859953002
Cr-Commit-Position: refs/heads/master@{#469492}

BUG=640116

Review-Url: https://codereview.chromium.org/2864483004
Cr-Commit-Position: refs/heads/master@{#469491}

BUG=717322

Review-Url: https://codereview.chromium.org/2862913002
Cr-Commit-Position: refs/heads/master@{#469490}

Manually convert constructor/iirfilter.html to use new Audit.

Updated new-audionodeoptions to allow specifying constructor option
when testing the default constructor.

BUG=704976
TEST=constructor/iirfilter.html

Review-Url: https://codereview.chromium.org/2859193003
Cr-Commit-Position: refs/heads/master@{#469489}

No change in logic.

This creates an initial implementation of a new WebView-specific
blocking page. This has no change in logic, but paves the way for
choosing between two different blocking pages (loud and quiet).

Currently, all this does is specify SBErrorDisplayOptions without
depending on the defaults set in BaseBlockingPage.

This exposes CreateControllerClient() in BaseBlockingPage as protected
so that AwSafeBrowsingBlockingPage can take advantage of it.

BUG=718545

Review-Url: https://codereview.chromium.org/2863723002
Cr-Commit-Position: refs/heads/master@{#469488}

In Blink Renaming, Resource::error() was incorrectly renamed to GetError().
This CL re-renames it to FinishAsError().

This CL doesn't rename it to Error() because Error() looks like the getter of
|Resource::error_|.

BUG=675877

Review-Url: https://codereview.chromium.org/2855883004
Cr-Commit-Position: refs/heads/master@{#469487}

https://pdfium.googlesource.com/pdfium.git/+log/cf53b788ca1c..d41f4752c81d

$ git log cf53b788c..d41f4752c --date=short --no-merges --format='%ad %ae %s'
2017-05-04 dsinclair Remove unused IFX_Pause member
2017-05-04 thestig Mention pdfium_all target in README.md.
2017-05-04 npm Fix undefined shift in opj_get_all_encoding_parameters

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: Iaebce2e41bf2d1afe065f38102647a2e25985ddd
Reviewed-on: https://chromium-review.googlesource.com/496646
Reviewed-by: <pdfium-deps-roller@chromium.org>
Commit-Queue: <pdfium-deps-roller@chromium.org>
Cr-Commit-Position: refs/heads/master@{#469486}

Android GN builds have been unable to build raw "unit_tests" as a target,
instead relying on unit_tests_apk or chrome/test:unit_tests. Linux builds,
in contrast, can do so without issue. I had thought that the issue was
that a "unit_tests" gn target in another directory was preventing gn
from creating a top-level alias. While this may be true, it's not the
reason for the discrepancy between Android and Linux builds. Instead,
the difference is that Linux builds create a unit_tests executable in
the top-level output directory, while Android builds create an APK and
a wrapper script in child directories.

This CL writes an additional wrapper script for gtest targets on Android
in the top-level directory that matches the suite's output_name (usually
its target name), allowing us to successfully build raw "unit_tests" on
Android. As a bonus, we can now run

  $ out/Release/unit_tests

(etc) on Android.

BUG=716117

Review-Url: https://codereview.chromium.org/2851523002
Cr-Commit-Position: refs/heads/master@{#469485}