GitLab

Attempt to fallback to some matching font in the GetMatchingDirectWriteFontForTypeface function in case we fail to find a best match via IDWriteFontFamily::GetFirstMatchingFont.

The IDWriteFontFamily::GetFirstMatchingFont call fails on some machines with E_FAIL with well known fonts like MS UI Gothic,
Segoe UI, etc. It is not clear as to why a font would be accessible to GDI and not to DirectWrite.

Added some debug alias variables to help track down these failures. We also attempt fallback to some font in the IDWriteFontList
for the font family in question. If all fail, we CHECK as before.

Next step would be to remove the CHECKS and fallback to GDI if CreateFontFromSkia fails.

BUG=434425

Review URL: https://codereview.chromium.org/737103002

Cr-Commit-Position: refs/heads/master@{#304981}

The previous version failed because we seem to have rolled in a new
version of breakpad that had an unused function error in a target
that was previously compiling fine w/ the chromium_code config.

This patch marks the target (symupload) as no_chromium_code.

TBR=brettw@chromium.org
BUG=431177

Review URL: https://codereview.chromium.org/743813002

Cr-Commit-Position: refs/heads/master@{#304980}

Revert of ChromeBrowserProvider shouldn't own the FaviconService (patchset #1 id:1 of https://codereview.chromium.org/731873003/)

Reason for revert:
Broke the branch build. https://uberchromegw.corp.google.com/i/chrome.mobile/builders/arm%20stable/builds/7097

Original issue's description:
> ChromeBrowserProvider shouldn't own the FaviconService
>
> As FaviconService is KeyedService and will be deleted along with profile.
>
> BUG=411435
> TBR=Tedchoc
>
> Review URL: https://codereview.chromium.org/739713002
>
> Cr-Commit-Position: refs/heads/master@{#304690}
>
> Committed: https://chromium.googlesource.com/chromium/src/+/8c5b8c707daa9c026540a63f6f1e1c5981f5909c

TBR=michaelbai@chromium.org
NOTREECHECKS=true
NOTRY=true
BUG=411435

Review URL: https://codereview.chromium.org/740853002

Cr-Commit-Position: refs/heads/master@{#304979}

Review URL: https://codereview.chromium.org/719383004

Cr-Commit-Position: refs/heads/master@{#304978}

This is a complete rewrite of CodeGen from scratch, taking advantage
of the fact that a while back we changed PolicyCompiler to not require
modifying instructions after they're created.  In particular, we no
longer need topological sorting because the MakeInstruction call order
implicitly sorts instructions for us.  Additionally, we no longer need
to compute basic blocks even because they were only needed for
sorting.

It turns out that a simple greedy algorithm with basic memoization
suffices for flattening the instruction graph, and we only really need
to worry about inserting "jump" instructions when the targets are too
far away.  With just this, we're able to emit programs that are just
as efficient as the old CodeGen implementation did.

As a bonus, the new code generator is now fully deterministic because
it doesn't create any sets or maps using pointer values as keys.

BUG=414363

Review URL: https://codereview.chromium.org/699633003

Cr-Commit-Position: refs/heads/master@{#304977}

This CL adds the BlueZ Media Endpoint service provider class and its
fake class, and add the files in chromeos/chromeos.gyp.

BUG=429016

Review URL: https://codereview.chromium.org/725383003

Cr-Commit-Position: refs/heads/master@{#304976}

Loading the chime causes the page to create an audio output stream.
Setting preload=none on the audio tag prevents the chime from being
loaded on startup, and instead causes it to be loaded on demand when the
chime is played.

BUG=432699,397019

Review URL: https://codereview.chromium.org/741873002

Cr-Commit-Position: refs/heads/master@{#304975}

Previously the display rectangles are placed by display object's
x and y, with scaling. However, due to scaling the value and
flooring the scaled result, there could be 1px gap from the
expected position, therefore just clicking may cause sending
the display layout change which causes fade-out and fade-in.

This 1px gap has been there, it's tiny, but it's noticeable.
Therefore this CL rewrites the layouting of display rectangles as
follows:
- first, place the primary display
- then look up the secondary display which is sticking to the
  primary display
This way there cannot be any gaps, therefore doesn't cause any
weird fade-outs.

BUG=386401
R=xiyuan@chromium.org, oshima@chromium.org
TEST=manually

Review URL: https://codereview.chromium.org/739413002

Cr-Commit-Position: refs/heads/master@{#304974}

Certain encoder implementations may provide memory buffers with special
properties in which clients are expected to write input frames. This
patch introduces a simple new interface allowing a VideoEncoder to
provide a factory object from which clients may obtain input video
frames. Encoder implementations are free to return null, in which case
VideoSender provides a default implementation based on VideoFramePool.

The upcoming VideoToolbox encoder implementation for OS X and iOS will
provide a custom implementation of this interface that will use the
encoder's underlying CVPixelBufferPool. The buffers vended by that pool
are backed by memory shared between the encoding process and the daemon
interfacing with the hardware video encoder. Clients that write into
these buffers will avoid a full frame copy of every input frame, which
is a substantial saving in memory bandwidth and latency for 1080p@60
video.

BUG=429057
R=miu, hclam, hubbe

Review URL: https://codereview.chromium.org/688423003

Cr-Commit-Position: refs/heads/master@{#304973}

This patch changes how PLI cleans tilings up. Previously, we'd iterate
all tilings to find the ones that match the criteria, then iterate
those and remove them in PLI.

With this patch, we ask PLTS to do most of the work. We calculate
scales that use layer information and pass it down to PLTS to
do the cleanup. PLTS, in turn, constructs a list of things to remove
and removes it, also removing it from the passed twin sets.

This allows us to remove one function from PLI and one from PLTS,
as well as hides more of the implementation details of PLTS from PLI.

R=danakj, enne
BUG=433048

Review URL: https://codereview.chromium.org/742903002

Cr-Commit-Position: refs/heads/master@{#304972}

Create a new class that takes in a list of directories, recursively scans them for image and video files, and returns the resulting list.

This is a re-do of 730353002, which fixes a flakey test.

BUG=420680
TEST=browser_test: FileManagerJsTest.MediaScannerTest

Review URL: https://codereview.chromium.org/739263003

Cr-Commit-Position: refs/heads/master@{#304971}

This patch adds a new SecurityInterstitialPage class and derives SSLBlockingPage
and SafeBrowsingBlockingPage from it. In the future, a captive portal
interstitial will also derive from this new class. In particular,
SecurityInterstitial does:

- Unify HTML generating code.
- Differentiate between different types of security interstitials for tests.
- Contain common code for showing and creating the interstitial page.

BUG=349737

Review URL: https://codereview.chromium.org/622683006

Cr-Commit-Position: refs/heads/master@{#304970}

The reboot for the devices was added because after a wipe certain settings
require a reboot to take effect. When the device is not wiped, there is no
reason to reboot it. We need the ability to have bots that will not reboot
devices, because some devices have boot-screen prompt as a security feature.

BUG=

Review URL: https://codereview.chromium.org/737033003

Cr-Commit-Position: refs/heads/master@{#304969}

Version 0.8.2.0

[ae952911d4] Change the default date used in zap_timestamp.
[0c61db4c14] SyzyASan - Fix a potential deadlock.

R=chrisha@chromium.org
BUG=

Review URL: https://codereview.chromium.org/743753003

Cr-Commit-Position: refs/heads/master@{#304968}

Introduce ProxyResolverDelegate to get rid of dependencies to content/ and chrome/

BUG=429354

Review URL: https://codereview.chromium.org/743623002

Cr-Commit-Position: refs/heads/master@{#304967}

Fix the broken page_cycler.netsim.top_10 and remove the rather ugly
clear_cache_before_ache_run class member. The fixed version should
not have the issue of affecting subsequent page_cycler runs.

R=tonyg@chromium.org
BUG=434927

Review URL: https://codereview.chromium.org/733463005

Cr-Commit-Position: refs/heads/master@{#304966}

This change gets nexe file identity while doing work in the browser process,
preventing an unnecessary plugin->browser IPC round, and simplifying this
codepath in the process.

BUG=394130
TEST=NonSfiPackagedAppTest.SuccessfulLoad covers the Non-SFI case.
CQ_EXTRA_TRYBOTS=tryserver.chromium.linux:linux_rel_precise32

Review URL: https://codereview.chromium.org/572973002

Cr-Commit-Position: refs/heads/master@{#304965}

This version should fix all the issues with chrome_proxy that thwarted
versions #3 and #4 ...

TBR=dtu@chromium.org
BUG=388256

Review URL: https://codereview.chromium.org/741903002

Cr-Commit-Position: refs/heads/master@{#304964}

Adds GYP targets that dictate which tests should be run on the
builders, and scripts to generate lists of the targets and commands
to run (with filters or other options as needed).

R=lcwu@chromium.org,gusfernandez@chromium.org
BUG=409163

Review URL: https://codereview.chromium.org/719353005

Cr-Commit-Position: refs/heads/master@{#304963}

BUG=434478

Review URL: https://codereview.chromium.org/726373005

Cr-Commit-Position: refs/heads/master@{#304962}

We currently use a mixture of ints and uint16s for port numbers.  This triggers a variety of "value possibly truncated" warnings on MSVC (currently disabled) wherever we implicitly truncate.  Fix this by using uint16 consistently through more functions.  (Using int consistently would be more problematic as the majority of third-party/system APIs that use port numbers use 16-bit types for them.)

By far the majority of these changes are fallout from changing IPEndPoint, which is widely used; it'd be difficult to split this CL up into smaller pieces :(

Note that I didn't use uint16_t to avoid introducing inconsistencies between existing uint16 usage and new uint16_t usage.  Conversion of everything to uint16_t can happen later.

This CL was reviewed and approved in pieces in the following CLs:
https://codereview.chromium.org/716223002/
https://codereview.chromium.org/717263003/
https://codereview.chromium.org/717373002/
https://codereview.chromium.org/718273002/
https://codereview.chromium.org/722503002/
Committing as TBR to the original reviewers.

BUG=81439
TEST=none
TBR=gunsch,cpu,jhawkins,davidben,jyasskin,mmenke

Review URL: https://codereview.chromium.org/655063002

Cr-Commit-Position: refs/heads/master@{#304961}

Change from setting QUIC's ack timer to 0ms when any handshake packet is
received to setting it when the handshake is confirmed on the client
side.

Merge internal change: 80306847
https://codereview.chromium.org/742823005/

Preparation for Chrome experiment to enable bandwidth resumption.
Protected behind FLAGS_quic_enable_bandwidth_resumption_experiment

Once this CL is in, I intend to add a Chrome experiment to add the
connection option kBWRE to a small percentage of connections to test the
effect of turning on BW resumption.

If a client provides a previously cached bandwidth estimate (that is
recent, and from the same serving region), then set initial CWND based
on this data.

Behavior change only when both the flag is true, and the Chrome
experiment adds the connection option to client handshakes.

A followup CL (78446219) will add varz to track the distribution of
resulting initial CWNDs.

Merge internal change: 80239310
https://codereview.chromium.org/737153002/

Changes QUIC's SentPacketManager to pass in HAS_RETRANSMITTABLE_DATA to
the SendAlgorithm for FEC packets, so that the SendAlgorithm can count
FEC packets towards congestion control.

Merge internal change: 80171666
https://codereview.chromium.org/731863007/

Record the last packet send time before we start sending the packet

Problem:

Every time QuicConnection sends a new packet, it records the time that
the packet was sent.  It passes this information to the
QuicSentPacketManager, which ultimately stores the information in the
TransmissionInfo.sent_time for the packet.

Later, when the QuicSentPacketManager receives an ack for the packet,
it retrieves its TransmissionInfo and takes a sample of the current
RTT as follows:

rtt_sample = ack_receive_time - transmission_info.sent_time

Previously, QuicConnection was recording the packet "sent_time" as the
time that WritePacket completes.  The problem with this approach is
that the write itself may pause the thread or take a long time.  In
this case, transmission_info.sent_time will be artificially inflated.
When that inflated value is subtracted from the ack time, as above, it
will cause the current RTT sample to become artificially small.

An artificially small RTT will affect our current estimate of the min
RTT, which we currently cannot recover from.  The min_rtt will be
pinned to the aberrant value.

Solution:

Changed the code to record the sent_time as the time that the write
begins.  The drawback of this approach is that any extra send time
will be temporarily factored into our smoothed-RTT calculations.  The
advantage is that it will prevent artificially small RTTs from setting
the min_rtt forever.

Added a corresponding test, which exercises this scenario by
artificially advancing the clock during the write.

Change QUIC to record send timestamp prior to write. Protected by
FLAGS_quic_record_send_time_before_write.

Merge internal change: 80138676
https://codereview.chromium.org/740793002/

Remove a QUIC LOG(DFATAL) that validly occurs when two packets are
queued.

This may occur when the internal server is write blocked when an two
packet REJ is written.

Merge internal change: 80124025
https://codereview.chromium.org/736053002/

Tighten up a QUIC LOG(DFATAL) for packets expected to have a non-zero
nack count.

If the packet is never sent, it never gets nacked, but that wasn't taken
into account.

Merge internal change: 80120917
https://codereview.chromium.org/735353002/

Use override instead of virtual..override in ./net/quic/... and
./net/tools/quic/...

C++11 in Chromium!

$ find .../quic/ \( -iname \*.h -o -iname \*.cc \) -print -exec clang_tidy {} -checks='-*,misc-use-override' -fix \;

Merge internal change: 80111599
https://codereview.chromium.org/741773002/

Log the SNI and UAID fields in handshake messages as quoted strings.

Merge internal change: 80005106
https://codereview.chromium.org/735933002/

R=rch@chromium.org

Review URL: https://codereview.chromium.org/744433003

Cr-Commit-Position: refs/heads/master@{#304960}

The failure seems to have been triggered from the recent
gtest update.  The exact revision seems to be:
https://code.google.com/p/googletest/source/detail?r=693

The issue only effects builds made with our older
gcc 4.4 toolchain. Neither asan nor valgrind report any
issues with the linux run of this test, and PNaCl build
also passes just fine so I'm assuming is a compiler bug
in the older toolchain.

The fix I found was to use "const char*" over "char []"
for local string constants.  In fact just adding a const
alone and leaving the [] syntax also fixes the issue.

Most likely there is an underlying compiler bug that
still needs to be addressed.

BUG=434821

Review URL: https://codereview.chromium.org/740023002

Cr-Commit-Position: refs/heads/master@{#304959}

We don't need this anymore, because the list sub-commands in run_benchmark
will stay where it is.

R=tonyg@chromium.org,dtu@chromium.org
BUG=421276

Review URL: https://codereview.chromium.org/731253003

Cr-Commit-Position: refs/heads/master@{#304958}

R=lcwu@chromium.org,damienv@chromium.org
BUG=408189

Review URL: https://codereview.chromium.org/741823004

Cr-Commit-Position: refs/heads/master@{#304957}

In preparation for lucid sleep, we will be putting all chrome renderers
that do not host GCM extensions into a special freezer cgroup.  All
processes in this cgroup will be frozen when the system suspends and
will not be thawed until the system has fully resumed.

The code to freeze and thaw the cgroup has existed for a while.  This
code actually starts putting processes into the cgroup.

Additionally, to deal with potential races that may occur with other
suspend observers that interact with renderers, the RendererFreezer is
no longer a PowerManagerClient::Observer but is instead a
PowerManagerClient::RenderProcessManagerDelegate.  This guarantees that
the RendererFreezer's suspend related methods will be called only after
all observers have reported ready and before observers are notified that
the suspend has completed.  Any race conditions that still exist now
need to be fixed in the observer that is causing it.

BUG=364339

Review URL: https://codereview.chromium.org/738993002

Cr-Commit-Position: refs/heads/master@{#304956}

R=lcwu@chromium.org,gfhuang@chromium.org
BUG=336640

Review URL: https://codereview.chromium.org/731933005

Cr-Commit-Position: refs/heads/master@{#304955}

This is a minimum change to allow easy cherrypicking to m40.  Much
more code needs to be removed to completely rid the menu of this.

BUG=434608

Review URL: https://codereview.chromium.org/740093002

Cr-Commit-Position: refs/heads/master@{#304954}

BUG=433380

Review URL: https://codereview.chromium.org/734843002

Cr-Commit-Position: refs/heads/master@{#304953}

Revert of Revert of [Base] Use variadic template in callback_list.h (wave 1) (patchset #1 id:1 of https://codereview.chromium.org/663083005/)

Reason for revert:
We now don't support building with MSVC Express any more. Do shout if this change causes problems for your compiler, though.

This is part of http://crbug.com/433164

Original issue's description:
> Revert of [Base] Use variadic template in callback_list.h (wave 1) (patchset #1 id:1 of https://codereview.chromium.org/618573002/)
>
> Reason for revert:
> Breaking compilation on windows with MSVC Express.
>
> Original issue's description:
> > [Base] Use variadic template in callback_list.h
> >
> > Replace pump.py generated base/callback_list.h with variadic template version.
> >
> > BUG=None
> >
> > Committed: https://crrev.com/a28756fae426023c53542f08c1bf80397a28f677
> > Cr-Commit-Position: refs/heads/master@{#298529}
>
> TBR=ajwong@chromium.org,tzik@chromium.org
> NOTREECHECKS=true
> NOTRY=true
> BUG=None
>
> Committed: https://crrev.com/69e698364a0a82a806fc5a4994c5f2040d1ff083
> Cr-Commit-Position: refs/heads/master@{#300487}

TBR=ajwong@chromium.org,tzik@chromium.org,lfg@chromium.org
NOTREECHECKS=true
NOTRY=true
BUG=None

Review URL: https://codereview.chromium.org/744503002

Cr-Commit-Position: refs/heads/master@{#304952}

BUG=434231
TEST=Verified on clapper device.

Review URL: https://codereview.chromium.org/738613003

Cr-Commit-Position: refs/heads/master@{#304951}

This is the second half of the new APIs.

BUG=429052
TEST=bots
R=kbr@chromium.org

Review URL: https://codereview.chromium.org/735173003

Cr-Commit-Position: refs/heads/master@{#304950}

I don't see any page_cycler regression locally so I'll try
to enable this on trunk as the nex step.
This change doesn't turn this on Windows to keep the scope of
possible trouble small.

BUG=377980
R=viettrungluu@chromium.org, darin@chromium.org

Review URL: https://codereview.chromium.org/664413003

Cr-Commit-Position: refs/heads/master@{#304949}

Replace pump.py generated base/callback.h with variadic template version.

BUG=433164

Review URL: https://codereview.chromium.org/610423003

Cr-Commit-Position: refs/heads/master@{#304948}

The warning has been fixed in sqlite upstream 3 years ago, but apparently we
haven't updated sqlite to a version that has this 3 year old commit.

BUG=340757,82385
R=shess@chromium.org

Review URL: https://codereview.chromium.org/742783005

Cr-Commit-Position: refs/heads/master@{#304947}

This includes ExtensionsClient and ExtensionsBrowserClient.

BUG=none
TEST=app_shell_unittests

Review URL: https://codereview.chromium.org/717233003

Cr-Commit-Position: refs/heads/master@{#304946}

Only new change:
fed7b07  Update clang warning flags. by Nico Weber - 42 minutes ago master

BUG=82385

Review URL: https://codereview.chromium.org/738063003

Cr-Commit-Position: refs/heads/master@{#304945}

When Init fails, further Save/Load calls should just fail (and not
crash). This makes the underlying LevelDB properly handle the Init failed case
and adds tests for that.

Adds LevelDB::InitWithOptions to make it easy to have the Init fail in tests.

Review URL: https://codereview.chromium.org/735823004

Cr-Commit-Position: refs/heads/master@{#304944}

Media source codec names will appear in chrome://media-internals
and be recorded by the PipelineStatus histogram.

BUG=431447

Review URL: https://codereview.chromium.org/743483002

Cr-Commit-Position: refs/heads/master@{#304943}

This CL adds browser_tests which run on nacl_helper in Non-SFI mode
to the nacl_helper_nonsfi, too.
As nacl_helper in Non-SFI mode is in prod, the existing tests are
kept as is, and duplicated for the nacl_helper_nonsfi rather than
switching.
Along with the change, PPAPIPrivateNaClPNaClNonSfiTest.FILEIO_Private is fixed. (The fixture was wrong).

BUG=358465
TEST=Ran trybots.
CQ_EXTRA_TRYBOTS=tryserver.chromium.linux:linux_rel_precise32

Review URL: https://codereview.chromium.org/724323002

Cr-Commit-Position: refs/heads/master@{#304942}