GitLab

This fixes the component=shared_library build on Chrome OS. Previously,
the TPM token initialization would crash the browser because libchaps.so
loads a 2nd, incompatible copy of the protobuf-lite library into the
browser.

The crash happens because by default the global scope takes precedence,
so symbols in the bundled copy shadow symbols in the system copy. This
can lead to accessing the wrong object or calling the wrong function.
The usual symptom is a crash during static initialization of libchaps.so.

RTL_DEEPBIND rearranges the scope for libchaps.so so that the system
library takes precedence instead. The scope for chrome is unaffected.

Some other possibilities were considered:

  - Unbundling the library. This would cause us to lose our local
    modifications to libprotobuf on Chrome OS.

  - Fixing the soname so that the bundled copy is used by chaps. This
    doesn't even link because the bundled code is not ABI compatible
    with upstream.

  - Statically linking libchaps.so. This is tricky because we need
    position-independent static libs for its deps, and we don't support
    building those currently. It will also cause bloat.

This hack is minimally invasive and allows use of shared library builds
on Chrome OS devices.

BUG=175508
TEST=cros chrome-sdk --board=link
  GYP_DEFINES="$GYP_DEFINES component=shared_library" gclient runhooks
  ninja -C out_link/Release chrome chrome-sandbox
  deploy_chrome --board link --build-dir out_link/Release --to $IP
  Log into the system. No crash.
  TPM-backed user certs show up in chrome://certificate-manager.

Review URL: https://codereview.chromium.org/1111373003

Cr-Commit-Position: refs/heads/master@{#329743}

Whitespace in the form of tabs and line feeds and the like can screw
things up.

This fixes the fresh.amazon.com bug.  (I tested it interatively.)

By the way, there is no need to do a migration pass.
The data will be correctly loaded when read from the
database.  (I checked.)  While it's true the database
will be out of date until written again, given that
the actual data used is correct, I don't think that's
an issue.

TBR=rdevlin.cronin,isherman,lazyboy,pvalenzuela,kmadhusu,avi
rdevlin.cronin - this trivial template URL data API change required some changes in chrome/browser/extensions
isherman - this trivial template URL data API change required some changes in chrome/browser/importer
lazyboy - this trivial template URL data API change required some changes in chrome/browser/renderer_context_menu
pvalenzuela - this trivial template URL data API change required some changes in chrome/browser/sync/test/integration
kmadhusu - this trivial template URL data API change required some changes in chrome/browser/ui/search/
avi - this trivial template URL data API change required some changes in chrome/browser/ui/cocoa/browser/

BUG=485357

Review URL: https://codereview.chromium.org/1135163002

Cr-Commit-Position: refs/heads/master@{#329742}

Reason for revert:
broke linux gn build

../../third_party/angle/src/libANGLE/Display.cpp:113:24:error: no type named 'DisplayGLX' in namespace 'rx'
        impl = new rx::DisplayGLX();
                   ~~~~^
1 error generated.
ninja: build stopped: subcommand failed.

http://build.chromium.org/p/chromium.linux/builders/Linux%20GN%20Clobber/builds/190/steps/compile/logs/stdio

Original issue's description:
> Roll ANGLE.
>
> https://chromium.googlesource.com/angle/angle/+log/b11e24..71632d
>
> BUG=484963
> BUG=483252
>
> Committed: https://crrev.com/6e8e29564f8590bff2215668d959b17075b96240
> Cr-Commit-Position: refs/heads/master@{#329727}

TBR=jmadill@chromium.org,geofflang@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=484963

Review URL: https://codereview.chromium.org/1128913010

Cr-Commit-Position: refs/heads/master@{#329741}

Four new tests for SyncCustomizationFragment:

- SCF state is correct starting from sync on and transitioning to off.
- SCF state is correct starting from sync off and transitioning to on.
- SCF's data type controls respond correctly to the sync everything switch.
- SCF opening and closing does not start sync.

In order to prevent flakiness, this CL makes some additional changes.

- Remove MockAccountManager.postAsyncAccountChangedEvent().
  Justification: we have no control about when Android performs
  the callback from it. The callback coming in at weird points
  during other test cases was making tests flaky. Testing mechanisms
  should be predictable.
- SCF now caches the backend initialized state in order to filter
  syncStateChanged() calls down to only changes to that state.

BUG=480604

Review URL: https://codereview.chromium.org/1118833002

Cr-Commit-Position: refs/heads/master@{#329740}

This is necessary because we want to remove the 'blink_tests' target
from Blink (third_party/WebKit).

BUG=483171
TEST=build/gyp_chromium
R=dpranke@chromium.org

Review URL: https://codereview.chromium.org/1129653004

Cr-Commit-Position: refs/heads/master@{#329739}

BUG=484541
TEST=none
R=tengs@chromium.org

Review URL: https://codereview.chromium.org/1144463002

Cr-Commit-Position: refs/heads/master@{#329738}

[Smart Lock] Brading string (Google Smart Lock) in save password infobar should be link to help center article.

Google Smart Lock in save password infobar should be a link to help
center article. Link is open in new tab, infobar is not dismissed when
link is clicked.

BUG=454815,486739

Review URL: https://codereview.chromium.org/1128073003

Cr-Commit-Position: refs/heads/master@{#329737}

Revert of Pass Surface ID namespace to renderer compositors. (patchset #4 id:60001 of https://codereview.chromium.org/1129883003/)

Reason for revert:
Speculative revert to address widespread blink layout test failures on mac following a chromium revision in range 329707-329709.

https://build.chromium.org/p/chromium.webkit/builders/WebKit%20Mac10.9%20%28retina%29/builds/655

No stacks but sample output:

STDERR: Assertion failed: (impl_.get() != nullptr), function operator->, file ../../base/memory/scoped_ptr.h, line 386.

We'll try reproducing locally and watch the bots. We can reland if this does not green up the webkit waterfall.

Original issue's description:
> Pass Surface ID namespace to renderer compositors.
>
> In order to enable renderer processes to create SurfaceSequences, they
> have to receive Surface ID namespaces from the browser process so that
> they can avoid collisions.
>
> BUG=478802
>
> Committed: https://crrev.com/c7a5afd66499fc8442e1b3146f55b227638934d8
> Cr-Commit-Position: refs/heads/master@{#329708}

TBR=jbauman@chromium.org,nasko@chromium.org,kenrb@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=478802

Review URL: https://codereview.chromium.org/1131673009

Cr-Commit-Position: refs/heads/master@{#329736}

Also exit early in the function when possible.

Review URL: https://codereview.chromium.org/1130153005

Cr-Commit-Position: refs/heads/master@{#329735}

Summary of changes available at:
https://chromium.googlesource.com/chromium/blink/+log/7a62fa1..b333144

TBR=alancutter@chromium.org,jsbell@chromium.org

Review URL: https://codereview.chromium.org/1135423003

Cr-Commit-Position: refs/heads/master@{#329734}

BUG=485381

Review URL: https://codereview.chromium.org/1138013005

Cr-Commit-Position: refs/heads/master@{#329733}

This is a no-op change. BoringSSL deprecated EVP_PKEY_dup in favor of a new
EVP_PKEY_up_ref. This makes it more obvious there isn't actually a copy and is
more consistent with other ref-counted types.

See https://boringssl.googlesource.com/boringssl/+/517da2f1efc91b179dfb1898f826b18a6a38f547%5E%21/#F3

BUG=none

Review URL: https://codereview.chromium.org/1129083007

Cr-Commit-Position: refs/heads/master@{#329732}

The header include guards should match the file path from root.

These entries were found using a modified version of Eric's
fix-include-guards.py script found attached in
https://code.google.com/p/chromium/issues/detail?id=435361#c7.

BUG=435361
TEST=base_unittests
R=danakj@chromium.org

Review URL: https://codereview.chromium.org/1137323003

Cr-Commit-Position: refs/heads/master@{#329731}

This allows us to add global parameters that
control the behavior of the default process heap. In this
case we need to be less agressive about freeing memory.

BUG=434397

Review URL: https://codereview.chromium.org/1133313003

Cr-Commit-Position: refs/heads/master@{#329730}

Adding link

NOTRY=true

Review URL: https://codereview.chromium.org/1131773005

Cr-Commit-Position: refs/heads/master@{#329729}

Revert of content/common: Remove use of MessageLoopProxy and deprecated MessageLoop APIs (patchset #9 id:160001 of https://codereview.chromium.org/1134113002/)

Reason for revert:
Broke the build for CHROMEOS && USE_V4L2_CODEC

Original issue's description:
> content/common: Remove use of MessageLoopProxy and deprecated MessageLoop APIs
>
> This patch was mostly autogenerated with https://codereview.chromium.org/1010073002/.
>
> BUG=465354
>
> Committed: https://crrev.com/adc9b4ed3d797c2990d682a3784d2e892198ce00
> Cr-Commit-Position: refs/heads/master@{#329696}

TBR=sievers@chromium.org,skyostil@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=465354

Review URL: https://codereview.chromium.org/1135943005

Cr-Commit-Position: refs/heads/master@{#329728}

https://chromium.googlesource.com/angle/angle/+log/b11e24..71632d

BUG=484963
BUG=483252

Review URL: https://codereview.chromium.org/1141753002

Cr-Commit-Position: refs/heads/master@{#329727}

The UwS interstitial launch went well, so there's no longer any reason
for a Finch gate. This kills the Finch-related logic.

BUG=451123

Review URL: https://codereview.chromium.org/1134603004

Cr-Commit-Position: refs/heads/master@{#329726}

Fixes 6 out of 8 slow StrictMode violations.

BUG=473357

Review URL: https://codereview.chromium.org/1129283004

Cr-Commit-Position: refs/heads/master@{#329725}

This CL moves |GetSecurityLevelForWebContents| and the |SecurityLevel|
enum into a class of their own: |ConnectionSecurityHelper| in
//chrome/browser/ssl (since it is applying Chrome-specific security
policy to determine the security level of a page).

Later, we will use the new |ConnectionSecurityHelper| class to compute a
security level for a page to send to the DevTools Security
panel. Specifically, we need to be able to calculate the security level
for a given page, not just the WebContents for the active tab.

BUG=484392

Committed: https://crrev.com/ca7c0a3374c21ab10c821f4f3a890afc79182acf
Cr-Commit-Position: refs/heads/master@{#329591}

Review URL: https://codereview.chromium.org/1123943002

Cr-Commit-Position: refs/heads/master@{#329724}

The pre-existing |CertificateErrorReporter| class (in
//chrome/browser/net) is now only in charge of sending reports over the
network. A new class (|CertificateErrorReport| in //chrome/browser/ssl)
is in charge of building and serializing the reports.

The motivation for this change is to allow reports to include
interstitial-specific information (such as the type of interstitial that
was shown, whether the user clicked through, etc.). So as to avoid
introducing interstitial knowledge into //c/b/net, all the report
building and serializing knowledge (including the report protobuf) has
been moved into //c/b/ssl. |SSLBlockingPage| now sends a serialized report
through |ChromeContentBrowserClient| to the SafeBrowsing UIManager to be
sent over the network.

|ChromeFraudulentCertificateReporter| (responsible for reporting
Google-property pinning violations) has also been moved into //c/b/ssl
so that it can use the new |CertificateErrorReport| class to build
reports before sending them with a |CertificateErrorReporter|.

BUG=462713,461588

Review URL: https://codereview.chromium.org/1117173004

Cr-Commit-Position: refs/heads/master@{#329723}

Right now we have no easy visibility on time taken for individual tests, since everything runs under one massive "Sharded Perf Tests" step. We're trying to balance the tests across devices to improve cycle time, so we need a way to see how long each test is actually taking.

BUG=466101

Review URL: https://codereview.chromium.org/1140783002

Cr-Commit-Position: refs/heads/master@{#329722}

There are ongoing improvements to smart selection that likely will not
land before M44 branch. Disable it for now, avoiding the need to
cherry-pick any eventual updates.

BUG=451255

Review URL: https://codereview.chromium.org/1135313003

Cr-Commit-Position: refs/heads/master@{#329721}

Implements WebUI for WKWebView by using URLFetcher to manually request
and retrieve WebUI resources and subresources, and load the page via
loadHTMLString. Sets favicons for history page using JavaScript.

BUG=450362

Review URL: https://codereview.chromium.org/1137143004

Cr-Commit-Position: refs/heads/master@{#329720}

Rotations and flips were listed in the enum, but not being allowed by
OverlayCandidate::GetOverlayTransform.

BUG=

Review URL: https://codereview.chromium.org/1135813005

Cr-Commit-Position: refs/heads/master@{#329719}

StartSrpcServices will require using the service_runtime pointer,
to set up the srpc_client pointer. So we cannot destroy the
service_runtime at the same time.

E.g., if I add a sleep() to induce the race, I get:

../../native_client/src/include/nacl_scoped_ptr.h:96: C *nacl::scoped_ptr<plugin::SelLdrLauncherChrome>::operator->() const [C = plugin::SelLdrLauncherChrome]: Assertion `ptr_ != __null' failed.
Received signal 6
#0 0x7f687f4aef7e base::debug::StackTrace::StackTrace()
#1 0x7f687f4aeabf base::debug::(anonymous namespace)::StackDumpSignalHandler()
#2 0x7f6879dee340 <unknown>
#3 0x7f6876504cc9 gsignal
#4 0x7f68765080d8 abort
#5 0x7f68764fdb86 <unknown>
#6 0x7f68764fdc32 __assert_fail
#7 0x7f688e2887f2 nacl::scoped_ptr<>::operator->()
#8 0x7f688e288487 plugin::ServiceRuntime::SetupAppChannel()
#9 0x7f688e28cb76 plugin::NaClSubprocess::StartSrpcServices()
#10 0x7f688e281810 plugin::PnaclTranslateThread::DoLink()
#11 0x7f688e27fe7d plugin::PnaclTranslateThread::DoLinkThread()
#12 0x7f6879de6182 start_thread
#13 0x7f68765c847d clone

Should have checked this in previous refactoring:
https://codereview.chromium.org/1128943003

BUG=473474

Review URL: https://codereview.chromium.org/1128373005

Cr-Commit-Position: refs/heads/master@{#329718}

Otherwise we lose track of display layout. In particular this is a
problem when dealing with software mirror mode since the state goes out
of sync and the second monitor will have the same origin as the primary
even though it's supposed to be right-of primary.

Follow-up to https://codereview.chromium.org/1129923004/ to fix software
mirror mode.

BUG=484294
TEST=Test software mirror mode and make sure it works

Review URL: https://codereview.chromium.org/1140773003

Cr-Commit-Position: refs/heads/master@{#329717}

JSON string literals are hard to work with because of all the escaping
necessary. The DictionaryBuilder class makes it easy to construct a
base::Value with code.

Review URL: https://codereview.chromium.org/1138013006

Cr-Commit-Position: refs/heads/master@{#329716}

Revert of Fix WebViewPlugin::scheduleAnimation crash (patchset #2 id:20001 of https://codereview.chromium.org/1137663006/)

Reason for revert:
This is converting the original crash to another crash. Will need another null check.

Original issue's description:
> Fix WebViewPlugin::scheduleAnimation crash
>
> The crash was probably due to accessing a dangling pointer to the plugin
> container during a small time frame between PepperWebPluginImpl::destroy()
> and the destructor being called. (Speculated from source since no reliable
> repro is found.)
>
> This CL clears eveything in the destroy() function as if the destructor has
> been called, only delaying memory release.
>
> R=tommycli
> BUG=483068
>
> Committed: https://crrev.com/6a9b5b10ff3515adb47718fc23ebe039a9e3b9ff
> Cr-Commit-Position: refs/heads/master@{#329309}

TBR=tommycli@chromium.org,raymes@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=487607,483068

Review URL: https://codereview.chromium.org/1127293004

Cr-Commit-Position: refs/heads/master@{#329715}

Summary of changes available at:
https://chromium.googlesource.com/chromium/blink/+log/ce97eb7..7a62fa1

TBR=alancutter@chromium.org,jsbell@chromium.org

Review URL: https://codereview.chromium.org/1140083003

Cr-Commit-Position: refs/heads/master@{#329714}

BUG=477632

Review URL: https://codereview.chromium.org/1143443002

Cr-Commit-Position: refs/heads/master@{#329713}

BUG=412250

Review URL: https://codereview.chromium.org/1136283007

Cr-Commit-Position: refs/heads/master@{#329712}

Without this, the downscoped token was being revoked, but the call to get a
fresh one was returning the cached non-downscoped token.

Also adds an assert to catch regressions.

BUG=b/20662147

Review URL: https://codereview.chromium.org/1123183005

Cr-Commit-Position: refs/heads/master@{#329711}

Summary of changes available at:
https://chromium.googlesource.com/skia/+log/38a017b..1ba8cc9

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

Review URL: https://codereview.chromium.org/1137353002

Cr-Commit-Position: refs/heads/master@{#329710}

Reduce GPU watchdog's polling interval from 2 seconds to 5 seconds.
Also remove unnecessary CheckArmed() call from DidProcessTask(). Although
CheckArmed() checks first whether it needs to post the task, WillProcessTask
and DidProcessTask is called so quickly that armed() still returns true, so we
ended up posting the task twice most of the time.

BUG=483622

Review URL: https://codereview.chromium.org/1133703002

Cr-Commit-Position: refs/heads/master@{#329709}

In order to enable renderer processes to create SurfaceSequences, they
have to receive Surface ID namespaces from the browser process so that
they can avoid collisions.

BUG=478802

Review URL: https://codereview.chromium.org/1129883003

Cr-Commit-Position: refs/heads/master@{#329708}

Previously SSLServerSocketOpenSSL was ignoring disabled_cipher_suites
list and require_forward_secrecy flag from SSLConfig. Fixed
SSLServerSocketOpenSSL to trim the list of cipher suites used in BoringSSL.

BUG=481163

Committed: https://crrev.com/d0eae58087e6f45088d6ef349d9ebaa2da450ea1
Cr-Commit-Position: refs/heads/master@{#329528}

Review URL: https://codereview.chromium.org/1138813003

Cr-Commit-Position: refs/heads/master@{#329707}

This moves the resource provider into the main core service process,
too, eliminating one of the processes.

BUG=477435

Review URL: https://codereview.chromium.org/1137223002

Cr-Commit-Position: refs/heads/master@{#329706}

Drawing LCD text into non-opaque layers is never correct - we should
update DrawProperties.can_use_lcd_text during animations also.

BUG=486892
R=danakj@chromium.org,enne@chromium.org

Review URL: https://codereview.chromium.org/1132253003

Cr-Commit-Position: refs/heads/master@{#329705}

Only update icon if the new one is not null on x11, no-op for null icon (follow the same way as Windows platform does in HWNDMessageHandler::SetWindowIcons)

BUG=478714.

Review URL: https://codereview.chromium.org/1138343003

Cr-Commit-Position: refs/heads/master@{#329704}