GitLab

Implements new public-key-pinning Java API (see https://tools.ietf.org/html/rfc7469). The API is accessible through CronetEngine.Builder.addPublicKeyPins() method. The method accepts the host, the a collection of pin hashes, the flag that indicates whether the pinning policy should be applied to the host subdomains and the pin expiration date. The pins are not persisted between client restarts.

BUG=522275

Review URL: https://codereview.chromium.org/1407263010

Cr-Commit-Position: refs/heads/master@{#363017}

Cut the max text length to 200 characters.
Use Newton method rather than b-search to find the text length.

Review URL: https://codereview.chromium.org/1493073002

Cr-Commit-Position: refs/heads/master@{#363016}

testharness.js - https://github.com/w3c/testharness.js
commit: 8658d6ca31d086e695b333aef55bb388091c7049
files: testharness.js idlharness.js

webidl2.js - https://github.com/darobin/webidl2.js
commit: bd216bcd5596d60734450adc938155deab1e1a80
files: WebIDLParser.js (orig: webidl2.js)

Related files testharness.css and vendor-prefix.js did not require
updates - see LayoutTests/resources/README.txt

BUG=561708
R=domenic@chromium.org

Review URL: https://codereview.chromium.org/1493473002

Cr-Commit-Position: refs/heads/master@{#363015}

RenderWidgetHostViewMus::CopyFromCompositingSurfaceToVideoFrame had a NOTREACHED()
that we were hitting on navigation. This CL removes that and adds a TODO.

BUG=none

Review URL: https://codereview.chromium.org/1493823002

Cr-Commit-Position: refs/heads/master@{#363014}

Review URL: https://codereview.chromium.org/1493083002

Cr-Commit-Position: refs/heads/master@{#363013}

Blink now implements the same thing in its interest rect code.

CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel

Review URL: https://codereview.chromium.org/1485823005

Cr-Commit-Position: refs/heads/master@{#363012}

Ultimately it will be done for all platforms and text selection changed will be deprecated.
TESTED=manual testing with Jaws and NVDA using textareas, input and content editables
BUG=558710
R=dmazzoni@chromium.org

Review URL: https://codereview.chromium.org/1490953002

Cr-Commit-Position: refs/heads/master@{#363011}

Pull in some of the upstream changes made to the Service Worker Cache API tests.

BUG=476814,482241,429832,564245

Review URL: https://codereview.chromium.org/1485283002

Cr-Commit-Position: refs/heads/master@{#363010}

BUG=565438
TEST=GPU FYI bots
TBR=bajones@chromium.org
NOTRY=true

Review URL: https://codereview.chromium.org/1493553008

Cr-Commit-Position: refs/heads/master@{#363009}

Let bindings generate a TypeError, rather than the module code
producing a TypeMismatchError or InvalidModificationError, if null is
passed for non-nullable interface arguments.

This was previously untested, so new tests were added for the async
API and sync API (in worker) cases.

(Some async cases are not modified in this CL, as there is a
behavior change we want to tackle separately.)

BUG=561338
R=philipj@opera.com

Review URL: https://codereview.chromium.org/1480363003

Cr-Commit-Position: refs/heads/master@{#363008}

Under pinch-to-zoom, there's no extra redraw event. However, the old
WindowView.resetGeometry() code was sensitive to dimensions
(window.innerWidth/innerHeight) that change with pinch-to-zoom. So when
page is pinch-to-zoomed in, refreshing or resizing triggers redraw,
causing inconsistent element sizing.

Our fix is to use dimensions that are insensitive to pinch-to-zoom,
i.e., document.documentElement.clientWidth/clientHeight. Under 100%
pinch-to-zoom, these are identical to the old dimensions except when
scroll bars are present.

BUG=481842

Review URL: https://codereview.chromium.org/1494783002

Cr-Commit-Position: refs/heads/master@{#363007}

Summary of changes available at:
https://chromium.googlesource.com/v8/v8/+log/a2c55ca8..baf869df

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/1491853004

Cr-Commit-Position: refs/heads/master@{#363006}

The API now specifies that ContentLayerClient decides its own PaintableRegion().

CQ_INCLUDE_TRYBOTS=tryserver.blink:linux_blink_rel

Review URL: https://codereview.chromium.org/1489713004

Cr-Commit-Position: refs/heads/master@{#363005}

This so far can happen in the test. The CL https://codereview.chromium.org/1483083003/ changes the timing of animation completion, which led to this state.

BUG=None
TEST=RootWindowLayoutManagerTest.DeleteChildDuringResize
R=sky@chromium.org

Review URL: https://codereview.chromium.org/1484383003

Cr-Commit-Position: refs/heads/master@{#363004}

Looks like everything that is done in ChromeMainDelegate::SandboxInitialized
is not applicable to WebView.

Review URL: https://codereview.chromium.org/1491743007

Cr-Commit-Position: refs/heads/master@{#363003}

Previously MSE GC algorithm didn't take into account properly the
duration of the last appended buffer. As a result GC would fail when the
media_time was beyond the timestamp of the last appended buffer. This CL
fixes that and adds a new unit test case to cover this situation.

BUG=563292

Review URL: https://codereview.chromium.org/1491513002

Cr-Commit-Position: refs/heads/master@{#363002}

In document mode, there are multiple activities and therefore multiple
Reader Mode managers. The UI utils that allow the in-page
javascript to work needs the active manager. This change updates
the reference in UI utils whenever a different tab is shown.

BUG=562360

Review URL: https://codereview.chromium.org/1497483003

Cr-Commit-Position: refs/heads/master@{#363001}

Pass all object parameters to JNI methods in JavaParamRef<> wrappers.
This matches previous changes made to do this for JNI non-method
functions.

BUG=519562
R=solb@chromium.org

Review URL: https://codereview.chromium.org/1472053003

Cr-Commit-Position: refs/heads/master@{#363000}

https://html.spec.whatwg.org/multipage/webstorage.html

The only change to the generated code is the data->value rename.

BUG=460722
R=jsbell@chromium.org

Review URL: https://codereview.chromium.org/1498823002

Cr-Commit-Position: refs/heads/master@{#362999}

Currently, functions like (for example) HTMLElement.toString() invoke a callback that obtains HTMLElement.toString's toString function and applies it to HTMLElement.

This leads to some weird behaviour, like:

> HTMLDocument.toString()
  "function HTMLDocument() { [native code] }"
> HTMLHRElement.toString()
  "function HTMLHRElement() { [native code] }"
> HTMLDocument.toString.toString = function() { return "foobar"; }
  function () { return "foobar"; }
> HTMLHRElement.toString()
  "foobar"
> HTMLDivElement
  function "foobar"

If we don't install the custom toString callback, everything works as expected.

BUG=

Review URL: https://codereview.chromium.org/1497503002

Cr-Commit-Position: refs/heads/master@{#362998}

Send localized group names to frontend and reuse some strings from
the old WebUi.
(Prerequisite for CL 1371073003)

BUG=134849

Review URL: https://codereview.chromium.org/1395073002

Cr-Commit-Position: refs/heads/master@{#362997}

In general the framebuffer has to be complete to claim
supporting RED on FBO.

If textureId is invalid glFramebufferTexture2DEXT returns
GL_FRAMEBUFFER_INCOMPLETE_ATTACHMENT.
So before this patch, IsGL_REDSupportedOnFBOs would have
returned true.

This leads to green frames when playing video.

This CL does fix bug 545904 but it is an indirect consequence.

We still need a new gpu driver bug workaround because this
CL does not avoid the DCHECK(glGetError() == GL_NO_ERROR);
at the end of IsGL_REDSupportedOnFBOs. Indeed the error is
set from the glTexImage2D call and glFramebufferTexture2DEXT
does not reset this gl error.

BUG=545904

R=bajones@chromium.org, hendrikw@chromium.org, kbr@chromium.org, zmo@chromium.org

TEST= --use-gl=egl on Mesa drivers

Review URL: https://codereview.chromium.org/1465223010

Cr-Commit-Position: refs/heads/master@{#362996}

BUG=138542

Review URL: https://codereview.chromium.org/1472083005

Cr-Commit-Position: refs/heads/master@{#362995}

It was not possible to determine how InputEvents were handled by
scripts or the default handlers.

The determination is required to add uma metrics for determining
the use of passive event listeners.

BUG=543611

Review URL: https://codereview.chromium.org/1463823003

Cr-Commit-Position: refs/heads/master@{#362994}

https://chromium.googlesource.com/skia.git/+log/c60b310af763..53c5d5fb795f

$ git log c60b310af..53c5d5fb7 --date=short --no-merges --format='%ad %ae %s'
2015-12-03 egdaniel Revert of Add transfer buffer support. (patchset #9 id:160001 of https://codereview.chromium.org/1490473003/ )
2015-12-03 benjaminwagner Fix ODR violation.
2015-12-02 reed API to support native scaling by image-generator
2015-12-02 joshualitt Wire up resource cache stats dumping in nanobench
2015-12-02 joshualitt Wire up new define in buildbot_spec
2015-12-02 joshualitt reset shader compilations for stats
2015-12-02 jvanverth Add transfer buffer support.
2015-12-02 joshualitt Create a define for nanobench builds to dump stats
2015-12-02 reed drawSprite and drawImage+clip should draw the same w/ an imagefilter applied. This gm exercises that invariant.
2015-12-02 mtklein Revert of skstd -> std for unique_ptr (patchset #17 id:320001 of https://codereview.chromium.org/1436033003/ )
2015-12-02 joshualitt Add gpuStatsDump to nanobench GPU bots
2015-12-02 mtklein skstd -> std for unique_ptr
2015-12-02 robertphillips Reduce use of GrRenderTarget in GrTextContexts

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

Review URL: https://codereview.chromium.org/1495943002

Cr-Commit-Position: refs/heads/master@{#362993}

Don't require users who have entered the control group to be eligible
for the trial in subsequent launches in order to stay in the control
group. This makes the control group persistent for users who enter
it (just as the experiment group is persistent for users who are
migrated, although the mechanism there is very different) so there
won't be different biasing around crashing or incognito use.

Also fixes a recent regression that would cause force-enabled or
force-disabled users not to report the group.

TBR=rohitrao

BUG=557115
TEST=For a control group user, either terminate in incognito, or
crash. On the next launch, the user should still be in the control
group.

Review URL: https://codereview.chromium.org/1495443006

Cr-Commit-Position: refs/heads/master@{#362992}

This Cl fixes a crash happening in 192 unit_tests with PlzNavigate enabled.

BUG=510836

Review URL: https://codereview.chromium.org/1498663003

Cr-Commit-Position: refs/heads/master@{#362991}

With --site-per-process and --isolate-extensions, cross-process
transfers were broken inside DevTools pages, since DevTools changes
the WebContents being navigated as it executes OpenURLFromTab, and the
RFHM that's eventually navigated can't find its cross-site transfer
request, since it exists on a different RFHM in a different
WebContents.  This breaks DevTools extensions.  See bug for more
details.

This CL disables process transfers in devtools pages to fix this.
According to dgozman@, this should be safe, since only extensions are
ever embedded in devtools pages.

BUG=564216, 532666

Review URL: https://codereview.chromium.org/1491043002

Cr-Commit-Position: refs/heads/master@{#362990}

TBR=oilpan-reviews
BUG=499780
NOTRY=true

Review URL: https://codereview.chromium.org/1494563005

Cr-Commit-Position: refs/heads/master@{#362989}

BUG=542921

Review URL: https://codereview.chromium.org/1492033002

Cr-Commit-Position: refs/heads/master@{#362988}

This change comes along with:
- http://crrev.com/1475583002 (ArcBridge part, this patch)
- http://crrev.com/1477733002 (UI part)
- http://ag/824594
- http://ag/824724
- http://ag/818913

These patch adds minimal support of new ARC notification:
- Showing notifications
- Closing ARC notifications from either side

This chromium-side change should do nothing until the android side change
is landed.

BUG=560682

Review URL: https://codereview.chromium.org/1475583002

Cr-Commit-Position: refs/heads/master@{#362987}

Because it wrongly claims supporting GL_EXT_texture_rg.
The bug in Mesa has been fixed in 11.1:
https://bugs.freedesktop.org/show_bug.cgi?id=93126

Video playback rendering is done differently in Chromium
depending if this extension is supported or not. If it
uses the extension whereas the driver actually does not
support it, this leads to green video frames.

BUG=545904

R=bajones@chromium.org, hendrikw@chromium.org, kbr@chromium.org, piman@chromium.org, zmo@chromium.org

TEST= --use-gl=egl on Mesa drivers

Review URL: https://codereview.chromium.org/1493813002

Cr-Commit-Position: refs/heads/master@{#362986}

I did a big renaming of long ago in r190693 that replaced uses of 'window'
with 'viewport' and some other terminology cleanups (see
http://www.chromium.org/developers/design-documents/blink-coordinate-spaces
for more details). This CL finishes off what was started there:

Renamed Widget's conversion methods from 'window' to 'root frame'. These
convert up to the coordinate space of the root widget, which is always the
root FrameView AFAIK.

Renamed Widget's conversion methods from 'view' to 'widget'. No need to
have multiple names for the same thing.

Clarified the coordinate space on a few ambiguous variables. And added
some clarifying comments and TODOs.

BUG=371902

Review URL: https://codereview.chromium.org/1486743002

Cr-Commit-Position: refs/heads/master@{#362985}

The SVGImage class is marked final and SVGImage::usesContainerSize()
returns true.  This means setContainerSize should never have to bail
out early over not using container size.

Review URL: https://codereview.chromium.org/1500573003

Cr-Commit-Position: refs/heads/master@{#362984}

This CL generalizes link rewriting done by PageSerializer
and gives control over it to users of PageSerializer through
the new PageSerializer::Delegate::rewriteLink virtual method.

The link rewriting is used to write cid:... uris in place of original
iframe.src attributes.  This is a more generic approach for solving the
about:blank cross-referencing (the old approach of using artificial
wyciwyg://frame/X URLs is being removed from PageSerializer).  The new
approach also solves bugs in MHTMLs containing subframes with a
different original-vs-current URIs (i.e. bugs 537047, 539936).
Additionally the new approach is compatible with OOPIFs (which don't
necessarily know the URLs of remote child frames and therefore wouldn't
be able to handle about:blank frames in the old way).  To see other
reasons, why Content-IDs are desirable, please see the CL description of
crrev.com/1418653009.

The old link rewriting code in PageSerializer (i.e.
LinkChangeSerializerMarkupAccumulator) is being removed - this code was
part of the merge of PageSerializer and WebPageSerializerImpl that was
reverted back in July 2015.  This code was not being used currently
(i.e. the registerRewriteURL and setRewriteURLFolder methods were not
being called right now outside of the test code).
The "rewritelinks_base.html" testcase is not applicable anymore, because
the code it was testing (base rewriting code in
LinkChangeSerializerMarkupAccumulator) is now removed.  The other
PageSerializerTest testcase related to LinkChangeSerializerMarkupAccumulator
(RewriteLinksSimple) has been rewritten on top of the new
PageSerializer::Delegate::rewriteLink API.

ContentIDs for subframes are for now generated in WebPageSerializer.cpp
(eventually this will have to be done in the browser process to be
compatible with --site-per-process flag and OOPIFs).
convertContentIDToURI method is being moved from
ArchiveResourceCollection.cpp to MHTMLParser, to enable reusing this
method by Content-ID-generating code in WebPageSerializer.

Pages with subframes that have been saved as MHTML after this CL lands will
contain Content-ID.  This means that newly generated MHTML files will not
render properly in older versions of Chromium (i.e. before crrev.com/1418653009
which first landed in 48.0.2555.0) and will instead say "[Chromium / Google
Chrome] needs to launch an external application to handle cid: links".

After building Chromium ToT + this CL and using it to save an MHTML of a test
page from http://anforowicz.github.io/nested-frames/index.htm, I've tried to
open the resulting test.mht file in various other, non-chromium-based browsers:
- Chromium ToT, Chrome 48.0.2564.8 and Internet Explorer 11.0.9600.18097
  opened and rendered the file correctly.
- Both Firefox 42.0 and Safari 9.0 (11601.1.56) seem to not support MHTML at all:
  - Firefox renders MHTML files as if they were text/plain (i.e. it renders
    MHTML headers and boundary markers and escaped/encoded MHTML parts).
  - Safari simply opens the Mac Finder with the location of the MHTML file.

BUG=538766, 537047, 539936

Review URL: https://codereview.chromium.org/1441553002

Cr-Commit-Position: refs/heads/master@{#362983}

BUG=138542

Review URL: https://codereview.chromium.org/1477433005

Cr-Commit-Position: refs/heads/master@{#362982}

Remove code supporting the migration of "session.urls_to_restore_on_startup"
to "session.startup_urls", the corresponding constants, histograms and unit
tests. This migration was there to support pre M-32 profiles.

This is a preparatory step before componentization of //chrome/browser/prefs
as keeping the migration code in the component is not worth the complexity.

Mark histograms as obsolete as of 2015/12 due to the removal of the migration
code they were mesuring.

BUG=525079

Review URL: https://codereview.chromium.org/1302303002

Cr-Commit-Position: refs/heads/master@{#362981}

This CL sets the HTTP method before starting a request in
CronetHttpURLConnection.

R=pauljensen@chromium.org

BUG=561678

Review URL: https://codereview.chromium.org/1490203003

Cr-Commit-Position: refs/heads/master@{#362980}

Summary of changes available at:
https://chromium.googlesource.com/v8/v8/+log/02a7a63b..a2c55ca8

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/1493343002

Cr-Commit-Position: refs/heads/master@{#362979}

Extension is not supported for Aura Android yet. So exclude some of them
to avoid pull unnecessary dependancy.

BUG=548797

Review URL: https://codereview.chromium.org/1482643002

Cr-Commit-Position: refs/heads/master@{#362978}