GitLab

Revert of Re-land 'Fold out-of-flow objects into anonymous blocks when removing children' (patchset #2 id:20001 of https://codereview.chromium.org/1543283002/ )

Reason for revert:
Ton of use-after-free security regressions.
BUG=573281

TBR=mstensho@opera.com

Review URL: https://codereview.chromium.org/1553853002

Cr-Commit-Position: refs/heads/master@{#367233}

TBR=ananta@chromium.org
BUG=573352
NOTRY=true

Review URL: https://codereview.chromium.org/1552913002

Cr-Commit-Position: refs/heads/master@{#367232}

(╯^□^)╯︵ ❄☃❄

BUG=557422
R=avi@chromium.org
TBR=atwilson@chromium.org

Review URL: https://codereview.chromium.org/1557693002

Cr-Commit-Position: refs/heads/master@{#367231}

(╯^□^)╯︵ ❄☃❄

BUG=557422
R=avi@chromium.org
TBR=stevenjb@chromium.org

Review URL: https://codereview.chromium.org/1556773002

Cr-Commit-Position: refs/heads/master@{#367230}

(╯^□^)╯︵ ❄☃❄

BUG=557422
R=avi@chromium.org

Review URL: https://codereview.chromium.org/1557683002

Cr-Commit-Position: refs/heads/master@{#367229}

(╯^□^)╯︵ ❄☃❄

BUG=557422
R=avi@chromium.org
TBR=caitkp@chromium.org

Review URL: https://codereview.chromium.org/1552023002

Cr-Commit-Position: refs/heads/master@{#367228}

After refactoring needed to support OOPIFs in save-page-as feature, the
html serializers in blink serialize frames, not pages and therefore
their old names (WebPageSerializer, PageSerializer, etc) do not apply
anymore.  Therefore this CL renames PageSerializer to FrameSerializer
where applicable.

Note that this rename not only covers PageSerializer class, but also
WebPageSerializer, WebPageSerializerClient, PageSerializerTest, test
data files, etc.

Note that FrameSerializer is more correct than HtmlSerializer, because
- FrameSerializer also serializes subresources (i.e. images)
- Even though WebFrameSerializerImpl deals only with HTML documents,
  for now it might make sense to leave it named as WebFrameSerializerImpl:
  - it still takes a Frame as input
  - it implements part of the public WebFrameSerializer class
    (and therefore it is desirable to have a similar name here as well)

BUG=538766,526786

Review URL: https://codereview.chromium.org/1541463002

Cr-Commit-Position: refs/heads/master@{#367227}

use_vulcanize is on by default. This means that release, official,
and debug builds that don't need debug versions of webui pages will
get the fastest experience.

To turn off vulcanization for local development, set

  use_vulcanize=0

in GYP, or

  use_vulcanize = false

in GN.

To conditionally include files in .grd files, do:

  <if expr="use_vulcanize">

and in C++:

  #include "chrome/common/features.h"
  #if BUILDFLAG(USE_VULCANIZE)

Currently, only the new Material Design downloads page is vulcanized,
but there will soon be others.

R=groby@chromium.org,thakis@chromium.org
BUG=541455
TEST=smaller binary

Review URL: https://codereview.chromium.org/1494253003

Cr-Commit-Position: refs/heads/master@{#367226}

Cr-Commit-Position: refs/heads/master@{#367225}

These buttons are implemented as a new class separate from LabelButton, which we'll need to retain even when MD is default for certain use cases such as the bookmarks bar.

This new class is used on MD infobars.

This first rough cut doesn't support
a) different button attributes that might affect rendering such as being a Call to Action or default in a dialog. TODO: estade
b) hover or press effects. TODO: bruthig
c) images on the button. I doubt we'll ever want to support images on material buttons, but it's possible we'll decide to add them later.

BUG=520266,571500

Review URL: https://codereview.chromium.org/1525163004

Cr-Commit-Position: refs/heads/master@{#367224}

(╯^□^)╯︵ ❄☃❄

BUG=557422
R=derat@chromium.org

Review URL: https://codereview.chromium.org/1551133002

Cr-Commit-Position: refs/heads/master@{#367223}

BUG=none

Review URL: https://codereview.chromium.org/1554823002

Cr-Commit-Position: refs/heads/master@{#367222}

Part 1 of several. This is currently set up to be landed as a standalone part of the first run flow work.

This change includes:
- Add strings for Chromium specific first run flow.
- Add first run flow nux to the WebUI (but stay unstamped for now). Used redlines and mocks as guidelines.

BUG=560457

Review URL: https://codereview.chromium.org/1554663002

Cr-Commit-Position: refs/heads/master@{#367221}

Also stop building and running this test on the Debug bots. This mirrors
how it's run in the GPU recipe.

BUG=542370
TBR=phajdan.jr@chromium.org,dpranke@chromium.org
NOTRY=true

Review URL: https://codereview.chromium.org/1559663002

Cr-Commit-Position: refs/heads/master@{#367220}

Picks up https://codereview.chromium.org/1551983002

BUG=573250
TBR=thestig

Review URL: https://codereview.chromium.org/1556753002

Cr-Commit-Position: refs/heads/master@{#367219}

BUG=542370
TBR=dalecurtis@chromium.org

Review URL: https://codereview.chromium.org/1555753002

Cr-Commit-Position: refs/heads/master@{#367218}

BUG=542370
TBR=dpranke@chromium.org,phajdan.jr@chromium.org

Review URL: https://codereview.chromium.org/1552903002

Cr-Commit-Position: refs/heads/master@{#367217}

This CL unifies validation for TexSubImage2D/3D and fixes texSubimage3D
cases in deqp/functional/gles3/negativetextureapi.html.

BUG=295792
TEST=deqp/functional/gles3/negativetextureapi.html

Review URL: https://codereview.chromium.org/1557603002

Cr-Commit-Position: refs/heads/master@{#367216}

Summary of changes available at:
https://chromium.googlesource.com/v8/v8/+log/3a586236..1b50b896

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.

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

Review URL: https://codereview.chromium.org/1559643002

Cr-Commit-Position: refs/heads/master@{#367215}

BUG=537412

Review URL: https://codereview.chromium.org/1550113003

Cr-Commit-Position: refs/heads/master@{#367214}

BUG=535323,572777

Review URL: https://codereview.chromium.org/1548403002

Cr-Commit-Position: refs/heads/master@{#367213}

Review URL: https://codereview.chromium.org/1554793002

Cr-Commit-Position: refs/heads/master@{#367212}

https://chromium.googlesource.com/chromium/src/+/c2a7942a1

BUG=572822
TBR=xidachen@chromium.org

Review URL: https://codereview.chromium.org/1555743002 .

Cr-Commit-Position: refs/heads/master@{#367211}

gles2_conform_test isn't present yet.

BUG=542370
TBR=dpranke@chromium.org,phajdan.jr@chromium.org

Review URL: https://codereview.chromium.org/1559613002

Cr-Commit-Position: refs/heads/master@{#367210}

Review URL: https://codereview.chromium.org/1552833002

Cr-Commit-Position: refs/heads/master@{#367209}

Picks up https://codereview.chromium.org/1557633002/

BUG=573250
TBR=jiayl@chromium.org

Review URL: https://codereview.chromium.org/1553793002

Cr-Commit-Position: refs/heads/master@{#367208}

BUG=566085

Review URL: https://codereview.chromium.org/1531013006

Cr-Commit-Position: refs/heads/master@{#367207}

Previous CL (crrev.com/1386873003) made MHTML serialization work in
presence of out-of-process iframes, but can potentially include the same
MHTML part twice.  MHTML with duplicated parts can be rendered by Chrome
and Internet Explorer without problems (at least according to the
limited ad-hoc testing I've done), but the unnecessary increase of the
file size is a performance regression.

After the current CL, the browser keeps track of already serialized
MHTML parts and communicates them to the renderer/serializer for
each frame (so they can be deduped).  This fixes the issue pointed
out above.

BUG=538766

Review URL: https://codereview.chromium.org/1417323006

Cr-Commit-Position: refs/heads/master@{#367206}

1. Use int instead of int64.
2. Define ENABLE_BIDIRECTIONAL_STREAM on net_unittests target.

BUG=516342

Review URL: https://codereview.chromium.org/1557643002

Cr-Commit-Position: refs/heads/master@{#367205}

When we transition to RunningOnCompositorButNeedsUpdate, we need to call
registerForAnimation again to ensure that updateCompositorAnimations is called.
We also cannot rely on m_lastTickTime since we do not get tickAnimation for
composited scrolls.

BUG=571680

Review URL: https://codereview.chromium.org/1551933002

Cr-Commit-Position: refs/heads/master@{#367204}

This change adds the following metrics:
- Count of times the help page is opened.
- Count of times the user explicitly navigates from the route details to sink list.
- Count of times the user explicitly navigates from the sink list to route details.
- Count of times the user explicitly navigates from the sink list to cast mode picker.

BUG=547977

Review URL: https://codereview.chromium.org/1548123002

Cr-Commit-Position: refs/heads/master@{#367203}

(╯^□^)╯︵ ❄☃❄

BUG=557422
R=avi@chromium.org
TBR=pasko@chromium.org

Review URL: https://codereview.chromium.org/1554733003

Cr-Commit-Position: refs/heads/master@{#367202}

Similar to r367111, send the reply to software draw in swap as well, and
allow PrepareTiles to run in parallel.

BUG=545633

Review URL: https://codereview.chromium.org/1553723003

Cr-Commit-Position: refs/heads/master@{#367201}

The existing division-by-0 guard was not correct, since MaximumDimension can
return negative values.  This led to occasional wild jumps when scrolling up and
down quickly (because new_velocity had the wrong sign).

This patch addresses the sign issue and improves handling of cases where we have
a large current velocity and a small delta to the new target.

BUG=575
CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel

Review URL: https://codereview.chromium.org/1553603003

Cr-Commit-Position: refs/heads/master@{#367200}

https://chromium.googlesource.com/skia.git/+log/e3c2f80006e7..f880e45e92fe

$ git log e3c2f8000..f880e45e9 --date=short --no-merges --format='%ad %ae %s'
2015-12-30 reed change factory to method, to parallel pattern in imagefilters. no functionality change.

CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel
TBR=mtklein@google.com

Review URL: https://codereview.chromium.org/1552873002

Cr-Commit-Position: refs/heads/master@{#367199}

ALIGN_CENTER is already default for STYLE_BUTTON.

BUG=none

Review URL: https://codereview.chromium.org/1556543002

Cr-Commit-Position: refs/heads/master@{#367198}

BUG=570113

Review URL: https://codereview.chromium.org/1552813002

Cr-Commit-Position: refs/heads/master@{#367197}

Should fix:
../../buildtools/third_party/libc++/trunk/include/map:843:5: error: static_assert failed
    "Allocator::value_type must be same type as value_type"
    static_assert((is_same<typename allocator_type::value_type, value_type>::value),
    ^             ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
../../components/metrics/leak_detector/leak_analyzer.h:69:55: note: requested here
           Allocator<std::pair<ValueType, uint32_t>>> suspected_histogram_;

BUG=573250

Review URL: https://codereview.chromium.org/1557663002 .

Cr-Commit-Position: refs/heads/master@{#367196}

Picks up https://codereview.chromium.org/1557653002/

BUG=573250
TBR=thestig

Review URL: https://codereview.chromium.org/1552883002 .

Cr-Commit-Position: refs/heads/master@{#367195}

This occurred because a ChildBrokerHost noticed that its connection to the child process died and so it nulled out its channel. On POSIX, this leads to the destruction of the ChildBrokerHost immediately and so the singleton BrokerState is notified synchronously. On Windows, there's another channel for sandboxing related sync IPCs which uses IO completition ports. To ease lifetime issues because of that, the ChildBrokerHost is only deleted in response to errors from that second channel. So if the first channel dies first, it is possible that BrokerState tries to connect to a message pipe in that ChildBrokerState with a null channel.

The fix is to null check the channel and to inform the other side of the message pipe that its peer has died so that it can fire peer closed notifications.

BUG=573244

Review URL: https://codereview.chromium.org/1558643002

Cr-Commit-Position: refs/heads/master@{#367194}