GitLab

The old implementation doesn't handle possible overflows,
when year is too large, for example. It makes a result
to be larger than 2^63 - 1, which results in overflow.

Fix Posix: use safe_math.h for multiplication and addition. If
overflow occurs, return possibly maximum platform dependent
value.

Fix Mac and Win: if safe cast is impossible, return Time(0).

Fix media and components: use day of week as well, as long
as unused variable results in undefined behavior and overflow

BUG=653445

Review-Url: https://codereview.chromium.org/2405453002
Cr-Commit-Position: refs/heads/master@{#427064}

https://chromium.googlesource.com/skia.git/+log/7a86423c4243..826167111f80

$ git log 7a86423c4..826167111 --date=short --no-merges --format='%ad %ae %s'
2016-10-24 caryclark fix one more fuzzer
2016-10-24 borenet Roll recipe DEPS

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_precise_blink_rel
TBR=reed@google.com

Review-Url: https://codereview.chromium.org/2446583002
Cr-Commit-Position: refs/heads/master@{#427063}

Remove last references to non-custom IDR_THEME_TOOLBAR.
Remove some unused SHADE images.
Remove DropdownBarView, which is nothing but a bunch of dead code.

BUG=648281
TBR=oshima@chromium.org

Review-Url: https://codereview.chromium.org/2441863002
Cr-Commit-Position: refs/heads/master@{#427062}

NOTRY=true
BUG=657839
TEST=none
TBR=pkasting@chromium.org

Review-Url: https://codereview.chromium.org/2443943002
Cr-Commit-Position: refs/heads/master@{#427061}

This makes SnippetArticle a pure model class that doesn't know about the adapter.

Goals of the overall refactoring:
1) Decoupling the various pieces of logic, by moving them from the adapter into
the tree structure. For example, dismissal is something that could be handled by
a TreeNode instead of doing the big switch statement we have right now in the
Adapter.
2) Stabilizing the structure of the tree. Right now we tend to throw away
everything and rebuild, which means we have to manually figure out what changes
exactly to notify about (cf. removeSuggestion). If we keep the tree structure
stable and just adjust the number of children below each node as necessary (like
SigninPromo already does), we get correct notifications for free, and it will
allow for some optimizations as well (we could cache the number of children for
each node and update it incrementally when a subtree changes. Right now we'd
have to contort ourselves a bit to make that work with things like
resetChildren()).

BUG=616090

Review-Url: https://codereview.chromium.org/2439683003
Cr-Commit-Position: refs/heads/master@{#427060}

BUG=556939

Review-Url: https://codereview.chromium.org/2430803003
Cr-Commit-Position: refs/heads/master@{#427059}

Removed ONC proxy policy applied check when the default network is
changed.

If proxy settings are bundled to the specific network and another
network is connected on startup, these settings are not
transferred to ARC on startup and should be transferred only when the network
becomes a default netowork.

BUG=657361
TEST=browsertest

Review-Url: https://codereview.chromium.org/2436753002
Cr-Commit-Position: refs/heads/master@{#427058}

Respect the ios_app_bundle_id_prefix and ios_automatically_manage_certs
when generating the Info.plist file for unit tests targets on iOS.

BUG=613543

Review-Url: https://codereview.chromium.org/2441793002
Cr-Commit-Position: refs/heads/master@{#427057}

- Remove all logging code from the multi-iframe NTP
- Remove NTPLoggingEventType.NTP_TILE, use NTP_*_SIDE_SUGGESTIONS instead
- Other small cleanups

BUG=656646
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:closure_compilation

Review-Url: https://codereview.chromium.org/2429523002
Cr-Commit-Position: refs/heads/master@{#427056}

It is not defined in implementation, and it is not needed to be checked.
https://chromium.googlesource.com/chromium/src/+/master/third_party/WebKit/Source/bindings/IDLExtendedAttributes.md#LogActivity_m_a

BUG=650150

Review-Url: https://codereview.chromium.org/2439413002
Cr-Commit-Position: refs/heads/master@{#427055}

BUG=475439

Review-Url: https://codereview.chromium.org/2442563002
Cr-Commit-Position: refs/heads/master@{#427054}

Update code in ui to use JavaRef when calling Java methods via JNI,
instead of passing bare jobject. Various function parameter types are
converted from jobject to JavaRef to enable calls to obj() higher up the
call chain to be removed.

BUG=506850

Review-Url: https://codereview.chromium.org/2417263002
Cr-Commit-Position: refs/heads/master@{#427053}

When the browser encounters an error while reading a response body, it should
error the body stream with a TypeError.

BUG=656601

Review-Url: https://codereview.chromium.org/2439673002
Cr-Commit-Position: refs/heads/master@{#427052}

Currently, CreateSortKey in password_manager_presenter.cc assumes that domain
and registry name obtained from a hostname string is always a suffix of the
hostname string. This is true for ASCII-based examples like this:

hostname: subdomain.example.com
domain and registry: example.com

But it is not true for non-ASCII-based examples like this:
hostname: žřč.com
domain and registry: xn--bea5m6d.com

The reason is that to obtain the domain and registry, the hostname is being
canonicalized.

The domain and registry is used to create a sort key for a credential with the
given hostname. The idea is to start with the domain and registry part, and
then append the remaining subdomains (reversed). The remaining subdomains were
computed by taking the prefix of hostname of the length of the difference of
the length of the hostname minus the length of the domain and registry part.
This number may overflow in case the domain and registry is not a suffix of the
hostname.

This CL fixes the issue by appending the whole hostname to the domain and
registry part. The results look like this:

hostname: subdomain.example.com
old key: example.com.subdomain
new key: example.com.com.example.subdomain

hostname: žřč.com
old key: undefined
new key: xn--bea5m6d.com.com.žřč

This is safe (no assumptions about suffixes) and preserves the sorting order of
any two credentials before and after this change. It is potentially duplicating
the domain and registry part. An alternative solution could just canonicalize
the hostname before composing the key, but the code would be more complex for
an unclear benefit.

The CL also uses a more efficient call to GetDomainAndRegistry (avoiding
redoing the canonicalization of a known URL). It also introduces a test and
removes an unnecessary call to ASCIIToUTF16.

BUG=655949

Review-Url: https://codereview.chromium.org/2439953004
Cr-Commit-Position: refs/heads/master@{#427051}

Also recounted, updated and corrected total bit count.

Review-Url: https://codereview.chromium.org/2442283002
Cr-Commit-Position: refs/heads/master@{#427050}

The protected virtual 'AllSensorsRemoved' method is added to PlatformSensorProviderBase class.
Unit tests are also included.

BUG=606766

Review-Url: https://codereview.chromium.org/2434073002
Cr-Commit-Position: refs/heads/master@{#427049}

This doesn't have any effect on sendBeacon's credentials handling.

This code was introduced in https://codereview.chromium.org/1016373002,
but even since at this point, AllowStoredCredentials is passed to the
BeaconLoader constructor.

Replacing with DCHECK.

R=mkwst@chromium.org
BUG=636297

Review-Url: https://codereview.chromium.org/2336273003
Cr-Commit-Position: refs/heads/master@{#427048}

We assume the caller is the first client and call StartInitialization() if
|pending_clients_| is empty, however |pending_clients_| may be cleared before
initialization finishes and cause successive StartSession() calls to call
StartInitialization() again. This leads to unwanted behavior depending on
platform-specific implementation.

Now we use an enum field to track initialization state, and flatten logic in
StartSession() to make things simple.

BUG=658662

Review-Url: https://codereview.chromium.org/2443113002
Cr-Commit-Position: refs/heads/master@{#427047}

This CL removes the sign-in menu on macOS as histogram indicate it is very rarely used.

The changes in the XIB file consists in removing this menu item from the menu nib file.

BUG=608181

Review-Url: https://codereview.chromium.org/2435053002
Cr-Commit-Position: refs/heads/master@{#427046}

HTMLElementTypeHelpers.h (generated) now has an enum, HTMLElementType, for each HTML element interface in htmltagnames.in
(Not generated for SVG elements)

New function htmlElementTypeForTag returns HTMLElementType
	Tag names that are undefined return HTMLUnknownElement

New ElementTypeHelpers.cpp.tmpl for HTML namespace with definition of function
Generates HTMLElementTypeHelpers.cpp

This function is needed in custom elements algorithms.

BUG=

Review-Url: https://codereview.chromium.org/2419383002
Cr-Commit-Position: refs/heads/master@{#427045}

BUG=642814

Review-Url: https://codereview.chromium.org/2422163003
Cr-Commit-Position: refs/heads/master@{#427044}

This patch creates a complete main for components perftests from
components unittests main and switches components perftests to using
testing/perf instead of base/perf.

BUG=648992

Review-Url: https://codereview.chromium.org/2358063002
Cr-Commit-Position: refs/heads/master@{#427043}

Include V0Cascade order to reflect that this is the legacy code for
Shadow DOM V0 cascading order in pure V0 documents.

R=kochi@chromium.org

Review-Url: https://codereview.chromium.org/2445673002
Cr-Commit-Position: refs/heads/master@{#427042}

BUG=

Review-Url: https://codereview.chromium.org/2443873002
Cr-Commit-Position: refs/heads/master@{#427041}

This was removed from CSP2, never added to CSP3, and will not ship in
any other browser. We should remove it, as it is completely redundant
with `X-XSS-Protection`, which is never going away.

Intent to Remove: https://groups.google.com/a/chromium.org/forum/#!topic/blink-dev/gjB93RpF6JY

R=jochen@chromium.org
BUG=657737

Review-Url: https://codereview.chromium.org/2428473004
Cr-Commit-Position: refs/heads/master@{#427040}

Like enums, typedefs, etc., dictionaries aren't observable unless used,
and the generated code does not change by removing [RuntimeEnabled].

R=bashi@chromium.org

Review-Url: https://codereview.chromium.org/2390483002
Cr-Commit-Position: refs/heads/master@{#427039}

Summary of changes available at:
https://chromium.googlesource.com/v8/v8/+log/e923cdb9..82319475

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_precise_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

TBR=hablich@chromium.org,machenbach@chromium.org,littledan@chromium.org,vogelheim@chromium.org

Review-Url: https://codereview.chromium.org/2440383002
Cr-Commit-Position: refs/heads/master@{#427038}

- Now that content/child can also depend on media just implement
  media related mime methods in SimpleWebMIMERegistry and remove
  renderer-only overrides.

- LinkLoader should use supportsMediaMIMEType rather than
  supportsMediaSourceMIMEType in as=media cases, as the latter
  is for MediaSource extension and always fails if empty codecs
  string is given (and that was what happening in the real renderer case)

BUG=658628
TEST=http/tests/preload/preload_with_type.html
TEST=LinkLoaderTest.Preload

Review-Url: https://codereview.chromium.org/2440163002
Cr-Commit-Position: refs/heads/master@{#427037}

Also change TransformBuilder::createTransformOperations to return the
TransformOperations rather than use an out argument.

Review-Url: https://codereview.chromium.org/2435413002
Cr-Commit-Position: refs/heads/master@{#427036}

This patch introduces |spellcheck_test| for writing asynchronous layout
tests for spellchecker in an intuitive manner, following the same spirit
of |assert_selection|.

This patch allows us to convert existing spellcheck tests easily as a
preparation for idle time spellchecker.

BUG=517298

Review-Url: https://codereview.chromium.org/2435273002
Cr-Commit-Position: refs/heads/master@{#427035}

The hardware read-only flag is obtained from device enumeration and
stored to Disk objects. Existing code overwrites that field when
applying read-only policy. However, that information will be needed when
we support remounting devices to switch between read-only and read-write
mode upon policy update.
After this change, Disk objects can distinguish whether a device is
mounted read-only because of a read-only hardware or because of the
mount option passed to Mount (for ExternalStorageReadOnly policy).

TEST=Run chromeos_unittest, unit_tests, and browser_tests.
BUG=642247,655003

Review-Url: https://codereview.chromium.org/2440443003
Cr-Commit-Position: refs/heads/master@{#427034}

searchbox_api.js uses a Chrome extension to Javascript. This is not understood by the Closure compiler, so was giving errors when grit attempted to minify it. To prevent this don't minify this file.

BUG=644392

Review-Url: https://codereview.chromium.org/2337253002
Cr-Commit-Position: refs/heads/master@{#427033}

This patch moves |setNonDirectionalSelectionIfNeeded()| to |SelectionController|
from |FrameSelection| since it is used only in |SelectionController| to simplify
|FrameSelection| for improving code health.

This patch also gets rid of |originalBase()| which had been used in
|setNonDirectionalSelectionIfNeeded()|, but it has never been used.

BUG=n/a
TEST=n/a; no behavior changes

Review-Url: https://codereview.chromium.org/2441573002
Cr-Commit-Position: refs/heads/master@{#427032}

This patch gets rid of flat tree version of |createVisibleSelection()| taking
two |VisiblePosition| by replacing with |SelectionInDOMTree| version to reduce
number of overloads for improving code health.

BUG=657237
TEST=n/a; no behavior changes

Review-Url: https://codereview.chromium.org/2437873008
Cr-Commit-Position: refs/heads/master@{#427031}

This patch changes |FrameSelection| class to use |DEFINE_TRACE()| instead of
|DEFINE_VIRTUAL_TRACE()| since |FrameSelection| is final.

|DEFINE_VIRTUAL_TRACE()| was introduced by [1] as result of mechanical replacing
of |virtual void trace()|.

[1] https://chromium.googlesource.com/chromium/src/+/0ae5a4a2770e78d7cdc0025f1ebeebd797691b16

BUG=n/a
TEST=n/a; no behavior changes.

Review-Url: https://codereview.chromium.org/2443883002
Cr-Commit-Position: refs/heads/master@{#427030}

BUG=NONE

Review-Url: https://codereview.chromium.org/2434333003
Cr-Commit-Position: refs/heads/master@{#427029}

This CL is part of the effort to improve performance of the password manager
settings. In particular, this CL fixes layout thrashing in the
updateOriginsEliding_ code. Prior to this change the code was first reading
style information from the DOM and then changing it in a loop.

This change creates a canvas element that is not part of the DOM.
It then uses the canvas to do the text measurements, thus avoiding the
repainting of the DOM. In the end, it writes the final obtained string in the
DOM, and thus does not change the display of the saved passwords.

Reference: https://developer.mozilla.org/en-US/docs/Web/API/CanvasRenderingContext2D/measureText

BUG=651049
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:closure_compilation

Review-Url: https://codereview.chromium.org/2439453005
Cr-Commit-Position: refs/heads/master@{#427028}

This CL instroduces a new class MemoryCoordinatorProxy that is the proxy
of MemoryCoordinator in content/browser. MemoryCoordinator::
GetCurrentMemoryState to get the current memory state is now not
accessible by some MemoryCoordinatorClients, and MemoryCoordinatorProxy
offers the proxy function to access this.

This is used to get the current memory state by e.g. TabLoader in
chrome/browser or components in net.

BUG=639700

Review-Url: https://codereview.chromium.org/2434353003
Cr-Commit-Position: refs/heads/master@{#427027}

* ColorInputType and RangeInputType
  They don't need to override InputType::setValue(), and should override
  InputTypeView::didSetValue() instead. It makes code simpler.

* HTMLInputElement::setValueInternal()
  Removed unused TextFieldEventBehavior argument, and rename the function to
  setNonAttributeValue().
  Make sure this is only for ValueMode::kValue.

* TextFieldInputType::setValue()
  Don't call InputType::setValue(), and call setNonAttributeValue() directly.

* InputType::setValue()
  Make sure this is only for ValueMode::kValue.

This CL has no behavior changes.

Review-Url: https://codereview.chromium.org/2435973002
Cr-Commit-Position: refs/heads/master@{#427026}

Cr-Commit-Position: refs/heads/master@{#427025}