GitLab

Summary of changes available at:
https://chromium.googlesource.com/v8/v8/+log/489561d2..df06cc9f

Please follow these instructions for assigning/CC'ing issues:
https://github.com/v8/v8/wiki/Triaging%20issues

Please close rolling in case of a roll revert:
https://v8-roll.appspot.com/
This only works with a Google account.

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=hablich@chromium.org,machenbach@chromium.org,kozyatinskiy@chromium.org,vogelheim@chromium.org

Review-Url: https://codereview.chromium.org/2852943002
Cr-Commit-Position: refs/heads/master@{#468299}

Revert of Changing default Windows compiler to VS 2017 (patchset #1 id:1 of https://codereview.chromium.org/2852433005/ )

Reason for revert:
Reverting as planned - it was just a test switch to VS 2017, which seems to have gone smoothly. I'm hopeful that canary 60.0.3086.0 is being built with VS 2017, to
complete the test.

Original issue's description:
> Changing default Windows compiler to VS 2017
>
> This is a single character change to temporarily switch Chrome to
> building with VC++ 2017. This CL is currently purely for testing
> purposes and will be landed for the weekend in order to flesh out any
> remaining bugs.
>
> This is a retry of crrev.com/2762093003 which failed due to a compiler
> warning which was missed in local testing and the try bots. That warning
> has been resolved and a local build of 'all' completed cleanly.
>
> BUG=683729
>
> Review-Url: https://codereview.chromium.org/2852433005
> Cr-Commit-Position: refs/heads/master@{#468239}
> Committed: https://chromium.googlesource.com/chromium/src/+/060d293f1d87d450a824dc46f564792272ed34a3

TBR=thakis@chromium.org
# Not skipping CQ checks because original CL landed more than 1 days ago.
BUG=683729

Review-Url: https://codereview.chromium.org/2851143002
Cr-Commit-Position: refs/heads/master@{#468298}

This fix will set translate infobar with a translating status in the beginning, then highlighted the targeted language tab after translated.

Tested on my device

BUG=703887

Review-Url: https://codereview.chromium.org/2845013002
Cr-Commit-Position: refs/heads/master@{#468297}

- Apply color and line-height to the whole page
- Set iron-icon-width/height globally, reducing the number of places
  where specific sizing is necessary
- Import cr-shared-style into Bookmarks' shared style module, which
  removes the need to specify some button styles
- Use paper-icon-button-light in folder-node, which simplifies CSS and
  slightly improves performance
- Make item highlight colors consistent

BUG=715394
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:closure_compilation

Review-Url: https://codereview.chromium.org/2846673003
Cr-Commit-Position: refs/heads/master@{#468296}

AnimationEffectTiming.idl and AnimationEffectTimingReadOnly.idl
now use
attribute PlaybackDirection direction;
for consistency with the spec.

https://w3c.github.io/web-animations/#the-animationeffecttiming-interface
https://w3c.github.io/web-animations/#the-animationeffecttimingreadonly-interface

BUG=716963

Review-Url: https://codereview.chromium.org/2856503002
Cr-Commit-Position: refs/heads/master@{#468295}

The parser will in some cases create new elements in documents that
have become frame-detached. Account for that -- no execution context
due to the document having become detached -- when processing the
event listeners of an event attribute.

R=haraken
BUG=714353

Review-Url: https://codereview.chromium.org/2855443002
Cr-Commit-Position: refs/heads/master@{#468294}

Currently, when we generate non bit fields, we order them in
alphabetical order. This may result in unnecessary padding, e.g.

struct {
  char a;
  /* char padding[3] */
  int b;
  char c;
  /* char padding[3] */
};

See [1] for an explanation.

This patch defines an ordering on field types so that we can sort
non bit fields from largest alignment size to smallest. This heuristic
minimises the padding most of the time:

struct {
  int b;
  char a;
  char c;
  /* char padding[2] */
};

This is purely a memory optimisation, not speed. It does not take into
account data access patterns and cache locality. We use the data from
[2] to define the order of types.

With this heuristic, some of the generated code has changed, but it is
unlikely to have changed the size of any structs.

Diff of generated files:
https://gist.github.com/darrnshn/81d146c2696ef100d56309e5b4b529ff/revisions

[1] http://www.drdobbs.com/cpp/padding-and-rearranging-structure-member/240007649?pgno=2
[2] https://codereview.chromium.org/2841413002

BUG=628043

Review-Url: https://codereview.chromium.org/2844223002
Cr-Commit-Position: refs/heads/master@{#468293}

BUG=715387
TEST=LayoutTests/editing/selection/selectstart_detaches_frame.html

Review-Url: https://codereview.chromium.org/2848993002
Cr-Commit-Position: refs/heads/master@{#468292}

Revert of CSS Motion Path: delete implementation of motion-offset property (patchset #2 id:40001 of https://codereview.chromium.org/2841863005/ )

Reason for revert:
Speculatively reverting due to build failure on Mac https://build.chromium.org/p/chromium/builders/Mac/builds/26841

Original issue's description:
> CSS Motion Path: delete implementation of motion-offset property
>
> The 'motion-offset' property has been replaced by 'offset-distance'.
>
> 'motion-offset' is among the properties disabled in M58.
>
> Intent to Remove: https://groups.google.com/a/chromium.org/d/msg/blink-dev/o1C5NzGf9Q0/sbaLF1MbAgAJ
>
> BUG=711937
>
> Review-Url: https://codereview.chromium.org/2841863005
> Cr-Commit-Position: refs/heads/master@{#468287}
> Committed: https://chromium.googlesource.com/chromium/src/+/612ac411762985a97166c4144ff0d049716a80db

TBR=bugsnash@chromium.org,meade@chromium.org,ericwilligers@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=711937

Review-Url: https://codereview.chromium.org/2856513002
Cr-Commit-Position: refs/heads/master@{#468291}

Revert of Disable flaky CommandLineExtensionsDontLoad test (patchset #1 id:1 of https://codereview.chromium.org/2852933002/ )

Reason for revert:
Speculatively reverting due to build failure on Mac https://build.chromium.org/p/chromium/builders/Mac/builds/26841

Original issue's description:
> Disable flaky CommandLineExtensionsDontLoad test
>
> TBR=emaxx@chromium.org
> BUG=716727
>
> Review-Url: https://codereview.chromium.org/2852933002
> Cr-Commit-Position: refs/heads/master@{#468286}
> Committed: https://chromium.googlesource.com/chromium/src/+/91f64acccbfe592af367f79ae74649d520528f0b

TBR=emaxx@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=716727

Review-Url: https://codereview.chromium.org/2855553002
Cr-Commit-Position: refs/heads/master@{#468290}

Revert of Disable flaky ExperimentalAppWindowApiTest.SetIcon (patchset #1 id:1 of https://codereview.chromium.org/2849073003/ )

Reason for revert:
Speculatively reverting due to build failure on Mac https://build.chromium.org/p/chromium/builders/Mac/builds/26841

Original issue's description:
> Disable flaky ExperimentalAppWindowApiTest.SetIcon
>
> TBR=benwells@chromium.org
> BUG=716726
>
> Review-Url: https://codereview.chromium.org/2849073003
> Cr-Commit-Position: refs/heads/master@{#468285}
> Committed: https://chromium.googlesource.com/chromium/src/+/78552aab8b749f33cd8176a977bdff4fa7d5de50

TBR=benwells@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=716726

Review-Url: https://codereview.chromium.org/2855563002
Cr-Commit-Position: refs/heads/master@{#468289}

BUG=695635

Review-Url: https://codereview.chromium.org/2844133003
Cr-Commit-Position: refs/heads/master@{#468288}

The 'motion-offset' property has been replaced by 'offset-distance'.

'motion-offset' is among the properties disabled in M58.

Intent to Remove: https://groups.google.com/a/chromium.org/d/msg/blink-dev/o1C5NzGf9Q0/sbaLF1MbAgAJ

BUG=711937

Review-Url: https://codereview.chromium.org/2841863005
Cr-Commit-Position: refs/heads/master@{#468287}

TBR=emaxx@chromium.org
BUG=716727

Review-Url: https://codereview.chromium.org/2852933002
Cr-Commit-Position: refs/heads/master@{#468286}

TBR=benwells@chromium.org
BUG=716726

Review-Url: https://codereview.chromium.org/2849073003
Cr-Commit-Position: refs/heads/master@{#468285}

Recently imported
external/wpt/WebCryptoAPI/generateKey/test_successes_RSA-OAEP.https.html
is occasionally timing out, so adding to TestExpectations.

TBR=eroman@chromium.org
BUG=716729

Review-Url: https://codereview.chromium.org/2852923002
Cr-Commit-Position: refs/heads/master@{#468284}

This adds tests to PermissionDialogTests that test Media requests with
the kUsePermissionManagerForMediaRequests feature enabled. When this
feature is enabled, media permissions can be requested in the same way
as other permissions.

Once the feature flag is removed, the old behavior that is being tested
can also be removed.

BUG=596786

Review-Url: https://codereview.chromium.org/2838873004
Cr-Commit-Position: refs/heads/master@{#468283}

IndexedDBBrowserTest.CursorTestIncognito is flaky on
linux_android_rel_ng.

TBR=jsbell@chromium.org
BUG=716728

Review-Url: https://codereview.chromium.org/2853753002
Cr-Commit-Position: refs/heads/master@{#468282}

This CL fixes an issue where drag and drop wouldn't work on Mac because
the drop data was being cleared by the bookmark manager API's drop
handler before the bookmark manager's handler had handled the drop. This
has been fixed by adding a setTimeout to clearing the data, which was
also in the old bookmark manager, giving the bookmark manager a chance
to deal with the drop.

BUG=716257
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:closure_compilation

Review-Url: https://codereview.chromium.org/2845243003
Cr-Commit-Position: refs/heads/master@{#468281}

added a headlessresourcedispatcherhostdelegate so that we do not crash when a website requests a client cert store

BUG=714534

Review-Url: https://codereview.chromium.org/2852633002
Cr-Commit-Position: refs/heads/master@{#468280}

When we generate ComputedStyle, we differentiate between properties
(which are meant to correspond to CSS properties) and nonproperties
(everything else). However, this terminology is confusing:

- Some nonproperties are derived from real CSS properties (e.g.
  hasSimpleUnderline).
- The generator code contains a lot of "for property_ in nonproperties"
  which is quite confusing.
- The name does not reveal the actual difference between properties
  and nonproperties in terms of code generation.

This patch removes the field role of 'nonproperty' in favour of the
parameter 'has_custom_compare_and_copy'. So there are only two field
roles: 'property' (which includes flags like 'unique' and 'emptyState')
and 'inherited_flag'.

BUG=628043

Review-Url: https://codereview.chromium.org/2830983003
Cr-Commit-Position: refs/heads/master@{#468279}

Clusterfuzz found a sample where
CompositeEditCommand::MoveParagraphWithClones() moves |before_paragraph|
out of the DOM tree after moving the paragraph. This patch uses
RelocatablePosition to track |before_paragraph| and |after_paragraph| so
that the positions we track are always connected.

Note: CompositeEditCommand::MoveParagraphs() is already using
RelocatablePosition to track positions before and after the moved
paragraphs.

BUG=712510
TEST=ApplyBlockElementCommandTest.IndentHeadingIntoBlockquote

Review-Url: https://codereview.chromium.org/2850773003
Cr-Commit-Position: refs/heads/master@{#468278}

This patch prunes some call sites of the deprecated function to improve
code health. The change is safe because layout has been updated a few
lines before each call site.

BUG=698633
TEST=n/a; no behavioral change

Review-Url: https://codereview.chromium.org/2847813005
Cr-Commit-Position: refs/heads/master@{#468277}

Add a reftest using an additional axis VM2B in the variabletest_box
font. The VM2B axis shifts the mark anchor to the left, leading to the
combining mark below box shifting right when the axis parameter is
adjusted from 0 to 1000. As the reference in the ref test, the font now
contains two additional glyphs N and O, which have fixed base anchors at
the middle and right position. A rendering of the M character showing
"m2b pos", followed by COMBINING SQUARE BELOW with VM2B 0, 500, 1000
applied must thus be equal to renderings of the N and O glyph with no
VM2B variation values applied.

BUG=693065
R=behdad, kojii

Review-Url: https://codereview.chromium.org/2850743002
Cr-Commit-Position: refs/heads/master@{#468276}

This CL resolves TODO(kouhei) in ScriptModuleResolverImpl to rethrow instantiation error when
resolved module has failed to instantiate.

TEST=ScriptModuleResolverImplTest.ResolveInstantiationFailedModule
BUG=594639

Review-Url: https://codereview.chromium.org/2837253002
Cr-Commit-Position: refs/heads/master@{#468275}

SharedWorkerRepositoryClientImpl implements classes defined in public/web/
(WebSharedWorkerConnectListener, WebSharedWorkerRepositoryClient) so it is being
moved to core/exported.

SharedWorkerRepositoryClientImpl.cpp did include WebLocalFrameImpl but luckily
enough didn't use it so we could just break this dependency.

BUG=708879

Review-Url: https://codereview.chromium.org/2833693002
Cr-Commit-Position: refs/heads/master@{#468274}

Add @RetryOnFailure annotation for
testVisualStateCallbackNotCalledAfterRendererGone

TBR=tobiasjs@chromium.org
BUG=716763

Review-Url: https://codereview.chromium.org/2851133002
Cr-Commit-Position: refs/heads/master@{#468273}

which are only called from one place:

  RecordAndCloseSessionOnError
  CloseSessionOnErrorInner

Review-Url: https://codereview.chromium.org/2849823005
Cr-Commit-Position: refs/heads/master@{#468272}

Cr-Commit-Position: refs/heads/master@{#468271}

When the Chrome installation has a brand code, the SettingsResetPromptModel may try to fetch default settings over the network. If the fetch fails, the ProfileResetter is passed a null pointer and any attempt to reset a setting will crash.

BUG=716372

Review-Url: https://codereview.chromium.org/2852883002
Cr-Commit-Position: refs/heads/master@{#468270}

TEST=run_compiler
BUG=636289
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:closure_compilation

Review-Url: https://codereview.chromium.org/2846783006
Cr-Commit-Position: refs/heads/master@{#468269}

The interface taking a raw pointer is deprecated and being removed.

BUG=690937

Review-Url: https://codereview.chromium.org/2847353002
Cr-Commit-Position: refs/heads/master@{#468268}

- NSS/ClientCertStoreTest/CertAuthorityFiltering
- DeviceDescriptionServiceTest.TestCleanUpCacheEntries

BUG=716730,716810
TBR=mattm@chromium.org,zhaobin@chromium.org

Review-Url: https://codereview.chromium.org/2850083002
Cr-Commit-Position: refs/heads/master@{#468267}

This CL adds SwapMetricsObserver which records swap rates and reports
them as platform-specific metrics. The purpose of adding these metrics
is to measure the impact of actions we take to free up memory. If we
aggressively free up memory under high memory pressure, we may touch
pages which are already swapped out (or compressed). In such cases
thrashing could happen and we are making the situation worse.
Swap rates and page compression/decompression rate would be good
indicators of thrashing. Metrics will be recorded every 60 seconds.

This CL only adds metrics for Android/ChromeOS/Linux. Follow-up CLs
will add them for macOS.

BUG=617492

Review-Url: https://codereview.chromium.org/2824133002
Cr-Commit-Position: refs/heads/master@{#468266}

This CL adds a feature flag which, when enabled, causes media requests to call
into PermissionManager::RequestPermissions in order to display a permission
prompt (among other things). The feature flag is added because this results in
the UI changing, and also won't work on Android until we have unified the
permission request codepaths (see https://crbug.com/606138).

BUG=596786

Review-Url: https://codereview.chromium.org/2829123003
Cr-Commit-Position: refs/heads/master@{#468265}

This CL fixes a regression since r467828. It changed the working directory, but
it missed to update relative path in "git add" argument.

BUG=710535
TBR=jeffcarp@chromium.org
NOTRY=true

Review-Url: https://codereview.chromium.org/2848183002
Cr-Commit-Position: refs/heads/master@{#468264}

Revert of Change ScopedFeatureList to overrides FeatureList not reset (patchset #6 id:140001 of https://codereview.chromium.org/2834583002/ )

Reason for revert:
Mac ASAN bots reporting use-after-free errors.

Original issue's description:
> Change ScopedFeatureList to overrides FeatureList not reset
>
> The current situation is that using ScopedFeatureList resets to an
> empty feature list and then enables/disables an explicit list of
> features.
>
> That's never what you want for browser tests (or other higher-level
> tests) since it effectively overrides higher-level test configurations
> (e.g. those in fieldtrial_testing_config.json, or a bot set up to
> specifically test a feature).
>
> In this patch:
>
> 1. Keep SFL::Init, SFL::InitWithFeatureList,
>    SFL::InitFromCommandLine reset to empty list but add warning to
>    remind developer should use them with care.
> 2. Change SFL::InitAndEnableFeature, SFL::InitAndDisableFeature and
>    SFL::InitWithFeatures to not reset but override current FeatureList
>    with given enables/disables.
>
> We also add unit tests for ScopedFeatureList.
>
> BUG=713390
>
> Review-Url: https://codereview.chromium.org/2834583002
> Cr-Commit-Position: refs/heads/master@{#468210}
> Committed: https://chromium.googlesource.com/chromium/src/+/9c04ed553bd7abe820a6a93c5e8981e6738881a9

TBR=isherman@chromium.org,asvitkine@chromium.org,thakis@chromium.org,chaopeng@chromium.org
# Not skipping CQ checks because original CL landed more than 1 days ago.
BUG=713390

Review-Url: https://codereview.chromium.org/2850073002
Cr-Commit-Position: refs/heads/master@{#468263}

1. A local variable |copied_files| is never read. Remove it.
2. So, copy_file() needs to return nothing.
3. Then, _webkit_root is removable.

Review-Url: https://codereview.chromium.org/2852673003
Cr-Commit-Position: refs/heads/master@{#468262}

BUG=710535

Review-Url: https://codereview.chromium.org/2848853002
Cr-Commit-Position: refs/heads/master@{#468261}

It can be replaced with path_from_tools_scripts().

BUG=710535

Review-Url: https://codereview.chromium.org/2849743003
Cr-Commit-Position: refs/heads/master@{#468260}