GitLab

Today the SelectorQuery code does:
isTreeScopeRoot(rootNode) || element->isDescendantOf(&rootNode)

after getting an element from the rootNode's treeScope id map as a way to
check if the element is inside the rootNode, but if isTreeScopeRoot was
true then isDescendantOf must also be true, so all this was doing was saving
a function call and a few branches for the elements with the matching id. This
doesn't seem worth the added complexity, so lets just rely on isDescendantOf.

This also lets us remove isTreeScopeRoot function, which is semantically
equivalent to Node::isTreeScope() anyway.

BUG=703900

Review-Url: https://codereview.chromium.org/2768263006
Cr-Commit-Position: refs/heads/master@{#459734}

We called pseudoStateChanged in a lot of cases where it wasn't
necessary. It should suffice to call it for the mentioned pseudo
classes when the min and max attributes changes, or the value is
changed, either as an attribute or as a result of user input.

This should fix the performance regressions in issue 704775.

R=tkent@chromium.org
BUG=704775

Review-Url: https://codereview.chromium.org/2774723004
Cr-Commit-Position: refs/heads/master@{#459733}

That's because it is used only from CompositeEditCommand subclasses.

TEST=No change in behavior

Review-Url: https://codereview.chromium.org/2772203002
Cr-Commit-Position: refs/heads/master@{#459732}

This simplifies the RemoteSuggestionsFetcher quite a bit.

BUG=609084

Review-Url: https://codereview.chromium.org/2771713003
Cr-Commit-Position: refs/heads/master@{#459731}

BUG=679905

Review-Url: https://codereview.chromium.org/2775183002
Cr-Commit-Position: refs/heads/master@{#459730}

Revert of Compile more targets in foreground/js/ui in gyp v2 (3) (patchset #5 id:80001 of https://codereview.chromium.org/2773873003/ )

Reason for revert:
It probably caused closing tree.

Original issue's description:
> Compile more targets in foreground/js/ui in gyp v2 (3)
>
> BUG=636289
> TEST=run_compiler
> CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:closure_compilation
>
> Review-Url: https://codereview.chromium.org/2773873003
> Cr-Commit-Position: refs/heads/master@{#459717}
> Committed: https://chromium.googlesource.com/chromium/src/+/00dda6dbc09aa8a14805cbba3d78b5b1569620ca

TBR=dbeam@chromium.org,fukino@chromium.org,oka@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=636289

Review-Url: https://codereview.chromium.org/2772223002
Cr-Commit-Position: refs/heads/master@{#459729}

DocumentThreadableLoader stops the timeout timer and restarts it with fresh
timeout when getting a preflight response and sending the actual request. This
fixes it so that the timeout is applied to the combined request, not to each of
the preflight and the actual request.

BUG=699985
R=tyoshino@chromium.org

Review-Url: https://codereview.chromium.org/2755483006
Cr-Commit-Position: refs/heads/master@{#459728}

BUG=625746

Review-Url: https://codereview.chromium.org/2123733002
Cr-Commit-Position: refs/heads/master@{#459727}

Each Service Worker Registration may only have one Background Fetch
active for a given tag, and has an associated origin. The new
BackgroundFetchRegistrationId type encapsulates the three values.

TBR=avi (file addition in //c/b/background_fetch/)
BUG=

Review-Url: https://codereview.chromium.org/2770393003
Cr-Commit-Position: refs/heads/master@{#459726}

From patch[1], |GranularityStrategy::updateExtent()| returns invalid
|SelectionInDOMTree|, which base is non-null and extent is null, when
|updateExtent()| is called with |IntPoint| points an element which can not
contain caret, e.g. VIDEO, SELECT, etc. This invalid selection can cause
|nullptr| reference in |SelectionInDOMtree::computeStartPositon()| in
|Position::operator<()|.

Before patch[1], |GranularityStrategy::updateExtent()| returns empty
|VisibleSelection| for invalid |SelectionInDOMTree|, base is non-null and
extent is null, by |VisibleSelection::validate()|.

This patch changes |{Character,Directional}GranularityStrategy::updateExtent()|
to handle above case by checking result of
|visiblePositionForContentsPoint()|.

[1] http://crrev.com/2735143002: Make GranularityStrategy::updateExtent() to
return SelectionInDOM instead of VisibleSelection

BUG=704529
TEST=run_webkit_unit_tests --gtest_filter=UpdateExtentWithNullPositionForCharacter
TEST=run_webkit_unit_tests --gtest_filter=UpdateExtentWithNullPositionForDirectional

Review-Url: https://codereview.chromium.org/2772163002
Cr-Commit-Position: refs/heads/master@{#459725}

Revert of Mojo JS bindings: support auto-loading mojom.js deps. (patchset #3 id:40001 of https://codereview.chromium.org/2779493002/ )

Reason for revert:
It looks we have to revert this CL too in order to revert https://codereview.chromium.org/2759563004/.

The previous revert failed in: https://codereview.chromium.org/2779533002/

Original issue's description:
> Mojo JS bindings: support auto-loading mojom.js deps.
>
> It also adds a config to turn off this behavior.
>
> BUG=699569
>
> Review-Url: https://codereview.chromium.org/2779493002
> Cr-Commit-Position: refs/heads/master@{#459695}
> Committed: https://chromium.googlesource.com/chromium/src/+/e326a5d409dadb3df6811b3dca7f1cfecb4ef520

TBR=rockot@chromium.org,yzshen@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=699569

Review-Url: https://codereview.chromium.org/2777083002
Cr-Commit-Position: refs/heads/master@{#459724}

TBR=scottmg
BUG=705367

Review-Url: https://codereview.chromium.org/2778673002
Cr-Commit-Position: refs/heads/master@{#459723}

Reasons:
* According to the comment it fails on Android only.
* There's no bug other than the Android failures.
* In https://codereview.chromium.org/2764843002/, a fix was made that solved two other failures. If this was failing on other platforms, it may well be fixed now.

BUG=NONE

Review-Url: https://codereview.chromium.org/2776433002
Cr-Commit-Position: refs/heads/master@{#459722}

I got tired of referring back to the cheat sheet, and encoding
experiments was annoying. This will hopefully be much better for
checking things.

There's a problem with --signed-in right now. Apparently the client id
and secret it auto-detects aren't usable. Providing our project's
internal one works fine, but I'll talk to Roger to see what's up with
the general client parameters.

Review-Url: https://codereview.chromium.org/2767143005
Cr-Commit-Position: refs/heads/master@{#459721}

For Kepler devices, we only use its VP8. We disable its VP9 and use Intel GPU instead.

BUG=701232
TEST=Run Thumbnail/VideoDecodeAcceleratorParamTest.TestSimpleDecode/0 of VDA
unittest for Guado.

Review-Url: https://codereview.chromium.org/2766083002
Cr-Commit-Position: refs/heads/master@{#459720}

Wake Lock presents challenges in decoupling from //content:

(1) It is naturally a per-WebContents interface, as it coalesces
requests made from frames within the same WebContents.
(2) On Android, it needs the NativeView of the WebContents in order to
block the screen from going to sleep.

This CL takes the following approach:
(1) Adds the notion of a context ID, which is set on WakeLockServiceContext
and which the Device Service can use to map contexts to NativeViews (via
a callback that is passed in at Device Service construction time by its
embedder).
(2) Adds a WakeLockContext Mojo interface. This interface can be used to
forward requests to bind wake locks from untrusted clients that are made within
the same (trusted-client specific) context.
(3) Adds a WakeLockContextProvider Mojo interface. This interface allows
for getting a WakeLockContext that is associated with a specific context ID.
The WakeLockContext impl uses that context ID on Android to obtain the
NativeView for its context.

Note that the lifetime model of WakeLockContext is somewhat complex: It must
stay alive as long as either
(1) Its Mojo connection is still valid (as the client might make future
    GetWakeLock() calls) OR
(2) There are still live WakeLock instances that it has instantiated (since
    they call into it when they receive Mojo requests from *their* clients).

Consequently, WakeLockContext monitors the state of the connections described
in (1) and (2), dying only when *all* of those connections go away. As a final
subtlety, WakeLockServiceImpl calls back into WakeLockContext from its
destructor, which executes immediately *after* WakeLockContext receives the
callback that that WakeLockServiceImpl instance has experienced a connection
error. WakeLockContext thus posts a task to delete itself rather than deleting
itself synchronously.

BUG=689403
TEST=On Chrome on Android, go to about://flags and turn on "Experimental Web
Platform features." Restart Chrome. Go to
https://colinblundell.github.io/wake_lock.html and verify that the page
says "Wake Lock API found". Verify that if you wait ~40 seconds, the device
goes to sleep. Tap "Block sleep." Verify that if you wait longer than the
above period, the device does not go to sleep. Tap "Unblock sleep" and
verify again that the device goes to sleep after waiting.
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_site_isolation

Review-Url: https://codereview.chromium.org/2734943003
Cr-Commit-Position: refs/heads/master@{#459719}

This ensures that bookmarks can only be saved when they have a valid,
non-empty URL.

BUG=692821
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:closure_compilation

Review-Url: https://codereview.chromium.org/2769863002
Cr-Commit-Position: refs/heads/master@{#459718}

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

Review-Url: https://codereview.chromium.org/2773873003
Cr-Commit-Position: refs/heads/master@{#459717}

This commit moves HalfFloatMaker and its implementation from cc/resources
to media/video.

BUG=None
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel

Review-Url: https://codereview.chromium.org/2763503002
Cr-Commit-Position: refs/heads/master@{#459716}

This patch moves all the cpp files for wtf_unittests to their new home
platform/wtf. The target "wtf_unittests" is removed from wtf/BUILD.gn, and
it's added in platform/wtf/BUILD.gn. Due to this change, a few other files
need to be updated, too.

BUG=691465

Review-Url: https://codereview.chromium.org/2771783003
Cr-Commit-Position: refs/heads/master@{#459715}

Expand firstRange() to simplify code.

BUG=693975
TEST=No change in behavior

Review-Url: https://codereview.chromium.org/2776983003
Cr-Commit-Position: refs/heads/master@{#459714}

In a previous patch https://codereview.chromium.org/2739553003, the
constructor and destructor of ComputedStyle were moved to be protected,
but it was not clear in the code why this was done. This patch adds a
comment explain it.

BUG=628043

Review-Url: https://codereview.chromium.org/2775153002
Cr-Commit-Position: refs/heads/master@{#459713}

Now that Frame stores a Page, this just updates the initializeCoreFrame
method to take a Page&, since it can't be null. This is pre-work to
remove FrameHost.

BUG=688614

Review-Url: https://codereview.chromium.org/2764773003
Cr-Commit-Position: refs/heads/master@{#459712}

TEST=No change in behavior

Review-Url: https://codereview.chromium.org/2773923002
Cr-Commit-Position: refs/heads/master@{#459711}

This CL adds a linear state machine to the Frame to ensure that it
follows an orderly progression through its state machine. The plan is to
use this state machine to make an isAttached() method which can check
this state, and later make page(), client(), host() etc return
references that DCHECK on the correct state. Also removed m_isDetaching
and replaced it with checks on the lifecycle.

Based on abarth's CL for adding a state machine for Document in:
https://chromiumcodereview.appspot.com/25036004

BUG=700783

Review-Url: https://codereview.chromium.org/2748603002
Cr-Commit-Position: refs/heads/master@{#459710}

Refactored out the need to use CSSPropertyID in the parsing logic of the
CSSPropertyCounterIncrement and CSSPropertyCounterReset properties. This
is pre work to allow the parseSingleValue method of these properties to
be implemented in their property APIs.

BUG=668012

Review-Url: https://codereview.chromium.org/2777793002
Cr-Commit-Position: refs/heads/master@{#459709}

Now newly introduced FrameLoadTypeReloadMainResource is used in major
reload code paths by default. To avoid adding wrong usage of Reload, and
to clean up loading code, remove existing FrameLoadTypeReload behaviors
completely, and change the type to take FrameLoadTypeReloadMainResource's
behavior instead. This will result in removing the type
FrameLoadTypeReloadMainResource, but it means removing legacy reload
behavior actually.

BUG=674404
TBR=thakis@chromium.org

Review-Url: https://codereview.chromium.org/2772753002
Cr-Commit-Position: refs/heads/master@{#459708}

Replace all direct references to constants in ui/views/layout/layout_constants.h
under chrome/browser/ui/views/extensions/* with LayoutDelegate metrics.

BUG=691897

Review-Url: https://codereview.chromium.org/2753243002
Cr-Commit-Position: refs/heads/master@{#459707}

BUG=692821
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:closure_compilation

Review-Url: https://codereview.chromium.org/2769103002
Cr-Commit-Position: refs/heads/master@{#459706}

Currently we have two UMA to measure time spent in window proxies initialization
- Blink.Binding.InitializeMainWindowProxy
- Blink.Binding.InitializeNonMainWindowProxy
but now WindowProxy was split into LocalWP and RemoteWP.
https://chromium.googlesource.com/chromium/src/+/690c3cf203b94623e5adbece7d6b4afc3501cd94

So I would like to split those UMAs to track performance more precisely.
- Blink.Binding.InitializeMainLocalWindowProxy
- Blink.Binding.InitializeMainRemoteWindowProxy
- Blink.Binding.InitializeNonMainLocalWindowProxy
- Blink.Binding.InitializeNonMainRemoteWindowProxy

BUG=588893

Review-Url: https://codereview.chromium.org/2771723005
Cr-Commit-Position: refs/heads/master@{#459705}

Some files are missing in BUILD.gn.
This CL updates BUILD.gn to include them,
or removes the file if it is just stale.

BUG=none
TEST=Ran bots.

Review-Url: https://codereview.chromium.org/2769343002
Cr-Commit-Position: refs/heads/master@{#459704}

surface_manager.cc/h and modify calls to surface_manager accoridingly.

BUG=

R=fsamuel@chromium.org
R=enne@chromium.org
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel;master.tryserver.chromium.linux:linux_site_isolation

Review-Url: https://codereview.chromium.org/2684933003
Cr-Commit-Position: refs/heads/master@{#459703}

https://pdfium.googlesource.com/pdfium.git/+log/b8227824c221..e4b035b722ad

$ git log b8227824c..e4b035b72 --date=short --no-merges --format='%ad %ae %s'
2017-03-26 palmer Use PartitionAlloc for JavaScript ArrayBuffers and strings.

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

Review-Url: https://codereview.chromium.org/2776923004
Cr-Commit-Position: refs/heads/master@{#459702}

This CL fixes a bug that a validation bubble is not closed by page navigation in
some cases.

We close a validation message on Page::documentDetached(). However it seems it was
too late to communicate with the browser process in some cases. So, this CL moves
it to Document unload timing.

 * Add ValidationMessage::willUnloadDocument(), which closes a validation bubble,
  and Document::dispatchUnloadEvents() calls it indirectly through Page.

 * HTMLFormControlElement prevents from showing a validation message after the
  unload processing.

BUG=704560

Review-Url: https://codereview.chromium.org/2771193002
Cr-Commit-Position: refs/heads/master@{#459701}

This patch makes |Selection#modify()| with word granularity to work as same
as mouse double-clicking to avoid skipping punctuation characters.

Example: Ctrl+ArrowRight or selection.modify('move', 'forward', 'word')
Before this patch: |ab := bc -> ab := |bc
After this patch:  |ab := bc -> ab |:= bc
Where "|" represent caret.

Note: this is Windows only behavior.

BUG=699903
TEST=attached test especially for "first-letter-word-boundary.html"

Review-Url: https://codereview.chromium.org/2754543004
Cr-Commit-Position: refs/heads/master@{#459700}

This patch simply adds SHA-1 string hashing for product registry lookup.

R=alph,caseq
BUG=703762

Review-Url: https://codereview.chromium.org/2766823002
Cr-Commit-Position: refs/heads/master@{#459699}

Revert of Remove sandbox/win/tools. (patchset #1 id:1 of https://codereview.chromium.org/2773333002/ )

Reason for revert:
Compile failed on WinClang64.

https://uberchromegw.corp.google.com/i/chromium.win/builders/WinClang64%20%28dbg%29/builds/11236

https://luci-logdog.appspot.com/v/?s=chromium%2Fbb%2Fchromium.win%2FWinClang64__dbg_%2F11236%2F%2B%2Frecipes%2Fsteps%2Fcompile%2F0%2Fstdout

[479/12131] CXX obj/sandbox/win/pocdll/handles.obj
FAILED: obj/sandbox/win/pocdll/handles.obj
C:\b\c\goma_client/gomacc.exe ../../third_party/llvm-build/Release+Asserts/bin/clang-cl.exe /nologo /showIncludes /FC @obj/sandbox/win/pocdll/handles.obj.rsp /c ../../sandbox/win/sandbox_poc/pocdll/handles.cc /Foobj/sandbox/win/pocdll/handles.obj /Fd"obj/sandbox/win/pocdll_cc.pdb"
C:\b\c\b\win_clang\src\sandbox\win\sandbox_poc\pocdll\handles.cc(7,10):  fatal error: 'sandbox/win/tools/finder/ntundoc.h' file not found
#include "sandbox/win/tools/finder/ntundoc.h"
         ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1 error generated.

Original issue's description:
> Remove sandbox/win/tools.
>
> The files therein look unused, haven't been really touched since 2012, aren't
> referenced in any build files, and have some maintenance cost.
>
> BUG=661774
> CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.win:win10_chromium_x64_rel_ng
>
> Review-Url: https://codereview.chromium.org/2773333002
> Cr-Commit-Position: refs/heads/master@{#459696}
> Committed: https://chromium.googlesource.com/chromium/src/+/6625b2f764ae4efb0637215660500ba721b8a43d

TBR=wfh@chromium.org,thakis@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=661774

Review-Url: https://codereview.chromium.org/2775123002
Cr-Commit-Position: refs/heads/master@{#459698}

List of changes:
1) IsNewFormattingContextForInFlowBlockLevelChild returns true for
   out-of-flow positioned child's style.
2) Add is_parent_new_fc_ to NGConstraintSpaceBuilder.
   We need need this bit to determine whether we need to pass exclusions
   to child's CS. Consider an example:
   <div id="container" style="width/height: 200px;">
     <float1><float-child></float1>
     <float2></float2>
   </div>
   float1 and float2 CSs will have available size == 200x200 and that
   size will be used to search for layout opportunities. So instead of
   passing arround the list of exclusions this patch introduces
   is_parent_new_fc that tells space_builder whether we need to pass
   exclusions to a child's CS.
3) Delete test lines that check NGLayoutOpportunityTree.
   That's because we can't access it anymore and we check its correctness
   in other places.

BUG=635619

Review-Url: https://codereview.chromium.org/2770483002
Cr-Commit-Position: refs/heads/master@{#459697}

The files therein look unused, haven't been really touched since 2012, aren't
referenced in any build files, and have some maintenance cost.

BUG=661774
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.win:win10_chromium_x64_rel_ng

Review-Url: https://codereview.chromium.org/2773333002
Cr-Commit-Position: refs/heads/master@{#459696}

It also adds a config to turn off this behavior.

BUG=699569

Review-Url: https://codereview.chromium.org/2779493002
Cr-Commit-Position: refs/heads/master@{#459695}