GitLab

This pulls in the following changes, including Windows build fixes. It
also gets Win32 assembly working.

03a739d Build Win32 with Yasm rather than MASM.
3e700bb Get MASM output working on Win32.
2a0f342 Fix standalone bio_test and v3name_test on Win64.
5f51c25 Add -cipher option to bssl client.
8f1ef1d Fix double-frees on malloc failure in ssl3_get_client_key_exchange.
93d67d3 Refactor ssl3_send_client_key_exchange slightly.
2af684f Add tests for ECDHE_PSK.
491956c Fix ECDHE_PSK key exchange.
48cae08 Add tests for PSK cipher suites.
b06711c Move the X509_NAME typedef into x509.h.
e167976 Make EVP_DigestVerifyFinal return only zero or one.
3cac450 Add SSL_SESSION_to_bytes to replace i2d_SSL_SESSION.
773bb55 Fix build (broken by removal of key_arg from SSL_SESSION parsing).
eb380a4 Fix build on Windows.
aeb8d00 Add less dangerous versions of SRTP functions.
7001a7f Don't bother accepting key_arg when parsing SSL_SESSION.
3b73c18 Fix "integer constant is too large for 'long' type" errors.
7571292 Extended master secret support.
89abaea Reimplement i2d_SSL_SESSION using CBB.
83fd6b6 Reimplement d2i_SSL_SESSION with CBS.
fbe6f49 The empty contents are not a valid ASN.1 INTEGER.
b5b6854 Add CBB_add_asn1_uint64.
9f2c0d7 Remove T** parameter to ssl_bytes_to_cipher_list.
2a39eae Don't condition on another header's #include guard.

BUG=338884, 418820

Review URL: https://codereview.chromium.org/664663004

Cr-Commit-Position: refs/heads/master@{#302016}

This CL builds the ppapi_proxy, ppapi_shared and ppapi_ipc libraries
for nacl_helper_nonsfi.
It is now for just making sure it is built properly. It will be actually used
from nacl_helper_nonsfi in a following CL.

BUG=358465
TEST=Ran trybot. Built nacl_helper_nonsfi on top of this CL, and made sure some browser tests actually work well locally.

Review URL: https://codereview.chromium.org/652393003

Cr-Commit-Position: refs/heads/master@{#302015}

Refactor Cubic congestion control to separate out PRR into a distinct
class.

Merge internal change: 78703144
https://codereview.chromium.org/688593002/

Move receive_window_ from BbrTcpSender and TcpCubicSender to
QuicSentPacketManager, since it's not specific to a congestion
algorithm.

Merge internal change: 78692549
https://codereview.chromium.org/687093002/

Gathering ClientConnectionStats at suitable times (near the end of
request handling and after receiving final ACK).

Exported kNumberOfNacksBeforeRetransmission for gathering
ClientConnectionStats.

Merge internal change: 78688302
https://codereview.chromium.org/687083002/

Deprecate rolled out flag FLAGS_close_quic_connection_unfinished_streams_2

Merge internal change: 78683497
https://codereview.chromium.org/639823010/

Close the QUIC connection if too many sent or received packets are
outstanding.

Protected by FLAGS_quic_too_many_outstanding_packets.

Merge internal change: 78666674
https://codereview.chromium.org/682413002/

Remove QUIC_VERSION_21 from supported versions.

Will remove it entirely later, after QUIC_VERSION_19 is removed.

Merge internal change: 78604601
https://codereview.chromium.org/682383003/

Turning off CID truncation for proxied connections.
Turning on CID truncation by default, since chrome will do it by
default.

Turning off CID truncation for proxied connections for QUIC.

Merge internal change: 78599095
https://codereview.chromium.org/683263003/

Protect UpdateRTT from negative send deltas

Prevent negative send_deltas from being incorporated into the various
RTT calculations.  AFAIK, it's not happening now, but it doesn't hurt to
be vigilant against future bugs.

If we're getting such RTTs, it's due to some pathological error that
we'd like to know about.

Added a unittest for these corner cases.

Low-impact. Don't use negative samples in QUIC rtt calculation.

Merge internal change: 78580114
https://codereview.chromium.org/687883002/

Protect against the divide by zero error in QuicBandwidth::TransferTime.
Currently some send algorithms may return QuicBandwidth::Zero.

Merge internal change: 78578198
https://codereview.chromium.org/684043002/

Add a QUIC connection option, 'NTLP', to disable tail loss probes by
setting max_tail_loss_probes_ to 0, in order to evaluate it's
performance benefit.

Merge internal change: 78465187
https://codereview.chromium.org/682283003/

Adding an option for peers to negotiate the length of the QUIC
connection ID sent to them.     This is a functional no-op until Chrome
starts sending this configuration option, at which point it can still be
turned off via FLAGS_allow_truncated_connection_ids_for_quic

Also fixing quic config to not have protected member variables while I'm
in there, because it was irritating me.

Allowing truncated QUIC connection IDs.   Off by default.

Merge internal change: 78462599
https://codereview.chromium.org/687033002/

R=rch@chromium.org

Review URL: https://codereview.chromium.org/691483003

Cr-Commit-Position: refs/heads/master@{#302014}

The devtools discovery page is rewritten/cleaned up to support two
optional paths for users:
1) Use the remote AppEngine debugger, which connects over WebSockets
   to the device. This triggers a security warning, so a help link
   is added for the user.
2) Provide a link to the Chrome native devtools UI for users.

R=byungchul@chromium.org,lcwu@chromium.org,dgozman@chromium.org
BUG=428471

Review URL: https://codereview.chromium.org/690723005

Cr-Commit-Position: refs/heads/master@{#302013}

When writing tests for cross-process navigation, we need to navigate to different hostnames. Doing this in tests requires lots of boilerplate code, which makes the test code harder to read and unlcear when we have cross-process navigation and when not. Adding this method will allow us to avoid the boilerplate code by putting it in a single method and make tests more readable. As we are ramping up work on out-of-process iframes, such tests will become a lot more common than in the past.

BUG=418236

Review URL: https://codereview.chromium.org/686053005

Cr-Commit-Position: refs/heads/master@{#302012}

1. Change the object ownership to be:
AwContents -> BrowserViewRenderer -> SharedRendererState ->
HardwareRenderer
(The left object owns the right object.)

2. Move DrawGL, InitializeHardwareIfNeeded, ReleaseHardwareIfNeeded from
AwContents to SharedRendererState.

BUG=418346

Review URL: https://codereview.chromium.org/653173004

Cr-Commit-Position: refs/heads/master@{#302011}

Currently, GLRenderer supports tile resources as either texture 2d or
texture rectangle.  Before this patch, it only supported texture 2d
masks.  This fills the gap to handle texture rectangle masks.  In
practice it means that using a zero copy raster worker pool will
now not break masks.  This in turn enables impl-side painting for
layout tests.

Unfortunately, there is a lot of indirection between a test saying it
wants to test a particular target type and what needs to happen.  To
make sure that tests are testing the right thing, the creation of
the raster worker and resource pools is now virtual so that tests
can create them as explicitly as needed.

R=danakj@chromium.org,reveman@chromium.org
BUG=423533

Review URL: https://codereview.chromium.org/668123003

Cr-Commit-Position: refs/heads/master@{#302010}

app_shell was missing blink_resources.pak so the remote debugging code
could not inject InjectedScriptSource.js

This patch does not significantly increase the size of the
app_shell .pak file -- it's still < 6 MB.

BUG=428166
TEST=manual, see bug

Review URL: https://codereview.chromium.org/689833002

Cr-Commit-Position: refs/heads/master@{#302009}

The jank issue has disappeared either because it was fixed or because an unrelated instrumentation is now counting this jank.

BUG=425814
TBR=rdsmith@chromium.org

Review URL: https://codereview.chromium.org/688203003

Cr-Commit-Position: refs/heads/master@{#302008}

This CL introduces a RendererTaskQueueSelector, which will be used by
the upcoming RendererScheduler. A RendererTaskQueueSelector will be queried
by the RenderScheduler's TaskQueueManager whenever it needs to select the
next task to run. The RendererTaskQueueSelector chooses which queue the
TaskQueueManager should service based on both priority and task age.

Queues can be moved between priority levels or disabled at any time,
which will be employed by the RendererScheduler when triggered by external
signals.

Design doc: https://docs.google.com/a/chromium.org/document/d/16f_RIhZa47uEK_OdtTgzWdRU0RFMTQWMpEWyWXIpXUo/edit

BUG=391005

Review URL: https://codereview.chromium.org/657953004

Cr-Commit-Position: refs/heads/master@{#302007}

This CL implements WebDataConsumerHandleImpl, a subclass of
WebDataConsumerHandle. This class will be used by fetch API with Streams.

BUG=418879

Review URL: https://codereview.chromium.org/620303002

Cr-Commit-Position: refs/heads/master@{#302006}

Previous instrumentations showed that URLRequestResourceBundleJob::GetData
alone is responsible for 20.5 janks per hour in IO thread. I need to instrument
the code inside it to find out which part causes jank.

This is a mechanical change that adds instrumentation required to locate the
source of jankiness (i.e. a long-running fragment of code executed as a part of
the task that causes jank) in the code. See the bug for details on what kind of
jank we are after.
A number of similar CLs were landed, and none of them caused issues. They've
helped to find and fix janky code. The code of the instrumentation is highly
optimized and is not expected to affect performance. The code simply creates a
diagnostic task which is identical to ones created by PostTask or IPC message
handlers. The task gets created only in developer build and in Canary channel.

BUG=422489

Review URL: https://codereview.chromium.org/688773002

Cr-Commit-Position: refs/heads/master@{#302005}

Review URL: https://codereview.chromium.org/684713003

Cr-Commit-Position: refs/heads/master@{#302004}

BUG=

Review URL: https://codereview.chromium.org/683333002

Cr-Commit-Position: refs/heads/master@{#302003}

Adding support for authenticated copresence calls. I've added a temporary call to the API to pass in an OAuth token to use with requests. Eventually we need to scope the token better - at least by app.

For privacy, we now register one device ID per account (auth token), including the null (anonymous) account.

Other changes:
- API keys are now scoped by app
- copresence_manager.cc deleted
- Minor performance/stability fixes
- Cleanup in rpc_handler_unittest.cc

BUG=423517

Review URL: https://codereview.chromium.org/671573003

Cr-Commit-Position: refs/heads/master@{#302002}

Belated unit tests to confirm layout, attachment, etc. behaviors when
entering/leaving fullscreen-within-tab mode.

BUG=290403

Review URL: https://codereview.chromium.org/662013003

Cr-Commit-Position: refs/heads/master@{#302001}

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

Review URL: https://codereview.chromium.org/687723002

Cr-Commit-Position: refs/heads/master@{#302000}

Netflix helper plugin has been removed from ChromeOS. See bug for details.

TBR=scottc@chromium.org
BUG=406573

Review URL: https://codereview.chromium.org/686813002

Cr-Commit-Position: refs/heads/master@{#301999}

This patch adds a flag to disable the new zip unpacker. Note, that this flag
is not used yet.

TEST=unit_tests: *AboutFlagsHistogramTest*
BUG=427775

Review URL: https://codereview.chromium.org/683413002

Cr-Commit-Position: refs/heads/master@{#301998}

Always use same browser_options instance. possible_browser may have a stale version because of the cache decorator on browser_finder.FindBrowser.

BUG=428157

Review URL: https://codereview.chromium.org/688163003

Cr-Commit-Position: refs/heads/master@{#301997}

This is guarded by a command-line switch.

TEST=OverrideCreatePluginTest
BUG=364716

Review URL: https://codereview.chromium.org/649873006

Cr-Commit-Position: refs/heads/master@{#301996}

TBR=yurys
BUG=

Review URL: https://codereview.chromium.org/689813002

Cr-Commit-Position: refs/heads/master@{#301995}

wincrypt.h defines macros that conflict with BoringSSL. Introduce a
crypto/wincrypt_shim.h wrapper header which #undefs those macros and
instead #defines replacement ones. All Chromium headers should
use this wrapper header.

BUG=338884

Review URL: https://codereview.chromium.org/686883002

Cr-Commit-Position: refs/heads/master@{#301994}

Minimizes the time of the blank page before before the app UI shows.
Focuses the app rather than opening another window if the app is already
running.

BUG=390086

Review URL: https://codereview.chromium.org/671333002

Cr-Commit-Position: refs/heads/master@{#301993}

This change enables ChannelMojo on Linux.
There could possibly be a perf slowdown, especially
on latency metrics.

If you're looking for the cause of some perf bot redness,
probably this is that one. Please feel free to revert and
let me know which benchmark was hit.

BUG=664413003
R=viettrungluu@chromium.org

Review URL: https://codereview.chromium.org/684323002

Cr-Commit-Position: refs/heads/master@{#301992}

The load_tests() protocol we use for running the
benchmark_smoke_unittests in telemetry_perf_unittests does some
complicated python hacking in order to synthesize python unit
test cases from the benchmark files.

Prior to the fix in this patch, any decorators that were applied
to the original benchmarks (in order to enable or disable tests)
were being ignored. This patch makes sure the decorated
enabled/disabled strings are copied over to the synthesized test
method.

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

Review URL: https://codereview.chromium.org/687893002

Cr-Commit-Position: refs/heads/master@{#301991}

The Google C++ style guide states:

  Explicitly annotate overrides of virtual functions or virtual
  destructors with an override or (less frequently) final specifier.
  Older (pre-C++11) code will use the virtual keyword as an inferior
  alternative annotation. For clarity, use exactly one of override,
  final, or virtual when declaring an override.

To better conform to these guidelines, the following constructs have
been rewritten:

- if a base class has a virtual destructor, then:
    virtual ~Foo();                   ->  ~Foo() override;
- virtual void Foo() override;        ->  void Foo() override;
- virtual void Foo() override final;  ->  void Foo() final;

This patch was automatically generated. The clang plugin can generate
fixit hints, which are suggested edits when it is 100% sure it knows how
to fix a problem. The hints from the clang plugin were applied to the
source tree using the tool in https://codereview.chromium.org/598073004.

BUG=417463
R=jhawkins@chromium.org

Review URL: https://codereview.chromium.org/684613002

Cr-Commit-Position: refs/heads/master@{#301990}

This documents some possibly unexpected behavior, where the pending entry
is cleared if it exists.

BUG=428380
TEST=No behavior change.

Review URL: https://codereview.chromium.org/684173004

Cr-Commit-Position: refs/heads/master@{#301989}

BUG=None
TEST=None

Review URL: https://codereview.chromium.org/688113002

Cr-Commit-Position: refs/heads/master@{#301988}

Use the correct font metrics in base PlatformFontWin if DirectWrite is used in the browser for font metrics.

The PlatformFontWin class uses GDI for creating fonts and their metrics. If DirectWrite is used for
font metrics then the metrics returned by GDI for the fonts and those calculated by skia using DirectWrite
differ leading to text getting clipped etc.

DirectWrite provides an interface IDWriteGdiInterop for interoperability with GDI. Converting the GDI font
to a DirectWrite font face object (IDWriteFontFace) and then converting this back to a GDI font gives us the
correct metrics.

The other option could be to create a PlatformFontSkia class on the same lines as PlatformFontWin and use skia
to get the font metrics. The amount of work needed there is unkown. We could look at this as a possible alternative
in case the above approach does not work.

Changes in this patch are as below:-
1. We use DirectWrite to create a IDWriteFactory interface in content browser. This is passed down to skia as the
   default factory.

2. The factory created in step 1 is also passsed via a static setter function to the PlatformFontWin class. The presence
   of a non NULL IDWriteFactory interface is used by this class to create a DirectWrite metrics compatible GDI font.

3. HarfBuzz text rendering is enabled by default in the browser. The MaybeEnableDirectWriteFontRendering function now
   checks for the disable harfbuzz text rendering switch instead.

BUG=427804,427802

Review URL: https://codereview.chromium.org/692633003

Cr-Commit-Position: refs/heads/master@{#301987}

BUG=401023
TEST=launch chrome with flag --enable-new-avatar-menu, open the new avatar menu, click signin to chrome, in the embedded gaia webview, try cmd+a and cmd+v, both should work.

Review URL: https://codereview.chromium.org/682113002

Cr-Commit-Position: refs/heads/master@{#301986}

This fixes a regression introduced in [ https://codereview.chromium.org/407083002 ].

BUG=423326
TEST=components_unittests
R=estade@chromium.org

Review URL: https://codereview.chromium.org/688543002

Cr-Commit-Position: refs/heads/master@{#301985}

For web to web navigation transition, the data is cleared
in RenderFrameHostManager::ResumeResponseDeferredAtStart().
For web to native app navigation transition, it will never
reach that point because no actual URL request was sent
out. In this case, clank side will call to clear navigation
transition data when the user navigate back from the native
app.

BUG=370696

Review URL: https://codereview.chromium.org/672973003

Cr-Commit-Position: refs/heads/master@{#301984}

While animating the transition, the origin is off in outerspace, as a
result we weren't correctly detecting the layer as non-solid since
the bitmap wasn't in the viewport

Also wrote a test for this.

BUG=420571

Review URL: https://codereview.chromium.org/692473003

Cr-Commit-Position: refs/heads/master@{#301983}

Previously, we allowed socket syscalls which were only needed by the
NaCl gdb stub. Now, we only allow these syscalls when the
--enable-nacl-debug flag is present.

Also restricts cross-process interaction for sched_* syscalls now that
non-crashing SIGSYS handlers are allowed under NaCl.

BUG=270914,413855

Review URL: https://codereview.chromium.org/670603002

Cr-Commit-Position: refs/heads/master@{#301982}

The Google C++ style guide states:

  Explicitly annotate overrides of virtual functions or virtual
  destructors with an override or (less frequently) final specifier.
  Older (pre-C++11) code will use the virtual keyword as an inferior
  alternative annotation. For clarity, use exactly one of override,
  final, or virtual when declaring an override.

To better conform to these guidelines, the following constructs have
been rewritten:

- if a base class has a virtual destructor, then:
    virtual ~Foo();                   ->  ~Foo() override;
- virtual void Foo() override;        ->  void Foo() override;
- virtual void Foo() override final;  ->  void Foo() final;

This patch was automatically generated. The clang plugin can generate
fixit hints, which are suggested edits when it is 100% sure it knows how
to fix a problem. The hints from the clang plugin were applied to the
source tree using the tool in https://codereview.chromium.org/598073004.

BUG=417463
R=nick@chromium.org

Review URL: https://codereview.chromium.org/686563003

Cr-Commit-Position: refs/heads/master@{#301981}

The Google C++ style guide states:

  Explicitly annotate overrides of virtual functions or virtual
  destructors with an override or (less frequently) final specifier.
  Older (pre-C++11) code will use the virtual keyword as an inferior
  alternative annotation. For clarity, use exactly one of override,
  final, or virtual when declaring an override.

To better conform to these guidelines, the following constructs have
been rewritten:

- if a base class has a virtual destructor, then:
    virtual ~Foo();                   ->  ~Foo() override;
- virtual void Foo() override;        ->  void Foo() override;
- virtual void Foo() override final;  ->  void Foo() final;

This patch was automatically generated. The clang plugin can generate
fixit hints, which are suggested edits when it is 100% sure it knows how
to fix a problem. The hints from the clang plugin were applied to the
source tree using the tool in https://codereview.chromium.org/598073004.

BUG=417463
R=jhawkins@chromium.org

Review URL: https://codereview.chromium.org/681893002

Cr-Commit-Position: refs/heads/master@{#301980}

Review URL: https://codereview.chromium.org/684313002

Cr-Commit-Position: refs/heads/master@{#301979}

BUG=426561
TEST=None

Review URL: https://codereview.chromium.org/682183003

Cr-Commit-Position: refs/heads/master@{#301978}

Automatic deployment of the virtual keyboard. For now we hide this under the flag: auto-virtual-keyboard.

This patch makes the controller a input device observer, instead of a shell observer. It records the presence of internal keyboards, external keyboard, and touchscreens. When the flag is enabled, the keyboard will not be shown if there does not exist a touchscreen device. The keyboard will be shown if the user is in maximized mode, or does not have an internal keyboard present.

BUG=373402
TEST=VirtualKeyboardControllerAutoTest.DisabledIfInternalKeyboardPresent, VirtualKeyboardControllerAutoTest.DisabledIfNoTouchScreen

Review URL: https://codereview.chromium.org/613343005

Cr-Commit-Position: refs/heads/master@{#301977}