GitLab

jam says this is obsolete/unused

Also breaks chrome->appcache_interfaces.h dependency

BUG=338338

Review URL: https://codereview.chromium.org/327823002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276389 0039d316-1c4b-4281-b951-d872f2087c98

BUG=379986

Review URL: https://codereview.chromium.org/323163003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276388 0039d316-1c4b-4281-b951-d872f2087c98

Revert of RenderTextHarfBuzz: Decide run direction by BiDi embedding level (https://codereview.chromium.org/326123003/)

Reason for revert:
Memory leak: http://build.chromium.org/p/chromium.memory/builders/Linux%20ASan%20LSan%20Tests%20%282%29/builds/3570/steps/ui_unittests/logs/HarfBuzz_RunDirection

Original issue's description:
> RenderTextHarfBuzz: Decide run direction by BiDi embedding level
> 
> BUG=382178
> NOTRY=true
> R=msw
> 
> Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=276379

TBR=msw@chromium.org
NOTREECHECKS=true
NOTRY=true
BUG=382178

Review URL: https://codereview.chromium.org/324413002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276387 0039d316-1c4b-4281-b951-d872f2087c98

ActionRunner instead.

Rename parameter to ExecuteJavaScript (both in action_runner.py and
web_contents.py) from expr/expression to statement.

BUG=361809

Review URL: https://codereview.chromium.org/328663007

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276386 0039d316-1c4b-4281-b951-d872f2087c98

The Launch test failed on mac 10.7 tests:
http://build.chromium.org/p/chromium.mac/builders/Mac%2010.7%20Tests%20%28dbg%29%283%29/builds/24632.

> [Mac] Add interactive App Shim test.
> 
> This test creates shims in the user data dir and actually starts them up and
> expects them to connect to the IPC socket.
> 
> BUG=168080
> 
> Review URL: https://codereview.chromium.org/316493002

TBR=jackhou@chromium.org

Review URL: https://codereview.chromium.org/325123004

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276385 0039d316-1c4b-4281-b951-d872f2087c98

Upon registration of new components, we reset the timer, as if the
component were registered upfront.

BUG=379880

Review URL: https://codereview.chromium.org/325073002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276384 0039d316-1c4b-4281-b951-d872f2087c98

This extra call to communicate was added by mistake. Remove it and add a
test.

Review URL: https://codereview.chromium.org/328553007

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276383 0039d316-1c4b-4281-b951-d872f2087c98

Resource added in 
https://codereview.chromium.org/322903004/

BUG=363644

Review URL: https://codereview.chromium.org/326813002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276382 0039d316-1c4b-4281-b951-d872f2087c98

Turns out the number of contexts in the gaia page can be more than 15, so we need this.

Also fix a bug in cros_test_case where browser_options were not being properly set by the test.

BUG=361862
TEST=testIframe
NOTRY=True

Review URL: https://codereview.chromium.org/214963004

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276380 0039d316-1c4b-4281-b951-d872f2087c98

BUG=382178
NOTRY=true
R=msw

Review URL: https://codereview.chromium.org/326123003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276379 0039d316-1c4b-4281-b951-d872f2087c98

The correct place for this adjustment seems to be in BeginFrame (or maybe
AdjustedBeginImplFrameDeadline?) but definately not in the
SyntheticBeginFrameSource.

Currently the code does;
  base::TimeDelta parent_draw_time =
      output_surface_->capabilities().adjust_deadline_for_parent
          ? BeginFrameArgs::DefaultDeadlineAdjustment()
          : base::TimeDelta();
  client_->SetEstimatedParentDrawTime(parent_draw_time);

BUG=345459

Review URL: https://codereview.chromium.org/328603002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276378 0039d316-1c4b-4281-b951-d872f2087c98

For the apk specific java files, we now run proguard on the jar instead of the classes.

BUG=375324,359249

Review URL: https://codereview.chromium.org/322443005

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276377 0039d316-1c4b-4281-b951-d872f2087c98

LegacyFrameProcessor reported decode error if ChunkDemuxerStream
(really, if SourceBufferStream)::Append() returned false. This could be
for a variety of reasons, mostly sanity checking the timestamps.
The new FrameProcessor ignored the ChunkDemuxerStream::Append() result
and assumed success, hiding potential issues. This change makes
FrameProcessor report decode error if CDS::Append() fails.

R=acolwell@chromium.org
TEST=no media_unittest or MSE layout test regression locally on linux

Review URL: https://codereview.chromium.org/324243002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276375 0039d316-1c4b-4281-b951-d872f2087c98

BUG=None 
CID=115981
TBR=battre

Review URL: https://codereview.chromium.org/310463009

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276374 0039d316-1c4b-4281-b951-d872f2087c98

Make constrained window use embedder's native window
(through its WebContents) instead of using webview's.

After print preview got enabled by default, the code
trying to bring up print preview dialog (a constrained
window) started crashing the browser.

BUG=286487
Test=In chromeos device/LinuxAura/Mac: load the sample
browser app
https://github.com/GoogleChrome/chrome-app-samples/tree/master/webview-samples/browser
Navigate to a page that calls window.print()
http://www.w3schools.com/jsref/tryit.asp?filename=tryjsref_print
Click "Print this page" to trigger window.print() inside
webview.
It should
1) not crash
2) correctly show the preview and print the contents of
the webview, not the entire app. This
means it shouldn't print the chrome (address bar/buttons etc.)
of the browser sample.

Review URL: https://codereview.chromium.org/321193003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276372 0039d316-1c4b-4281-b951-d872f2087c98

Now we're switching LoadNexeAndStart from synchronous implementation to
asynchronous one. Under PPAPI implementation, the async callbacks are
handled on main thread.
To keep the similicity of the thread model, this CL introduces the limitation
that Plugin::LoadNexeAndStart must be called on the main thread.

TEST=Ran browser_tests --gtest_filter=NaCl* locally and trybots.
BUG=333950
R=teravest@chromium.org

Review URL: https://codereview.chromium.org/321053004

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276371 0039d316-1c4b-4281-b951-d872f2087c98

In order to use the chrome.bluetoothSocket and chrome.bluetoothLowEnergy
APIs you will need to add "socket": true or "low_energy": true
respectively to the "bluetooth" manifest section.

BUG=376607,365839
TEST=tests updated

Review URL: https://codereview.chromium.org/325893002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276370 0039d316-1c4b-4281-b951-d872f2087c98

BUG=380987

Review URL: https://codereview.chromium.org/311263016

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276369 0039d316-1c4b-4281-b951-d872f2087c98

This test creates shims in the user data dir and actually starts them up and
expects them to connect to the IPC socket.

BUG=168080

Review URL: https://codereview.chromium.org/316493002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276368 0039d316-1c4b-4281-b951-d872f2087c98

By moving the options to a list, we can add more fields in the future without
breaking backward compatibility.

This CL is the first part of the cleanup. Arguments for callbacks will be
refactored separately.

TEST=unit_tests, browser_tests: *FileSystemProvider*
BUG=373165

Review URL: https://codereview.chromium.org/329483002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276367 0039d316-1c4b-4281-b951-d872f2087c98

This way, the LKGR builder and the developers who want to try ASan (or just Clang) on Windows won't get a broken Clang revision.

The FYI bots got LLVM_WIN_REVISION=HEAD in r276301 so won't be affected.

BUG=82385,345874
NOTRY=true

Review URL: https://codereview.chromium.org/320383004

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276366 0039d316-1c4b-4281-b951-d872f2087c98

Should provide creation context and isolate for WebArrayConverter::toV8Value and WebArrayConverter::createFromV8Value.

BUG=369451

Review URL: https://codereview.chromium.org/301883002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276361 0039d316-1c4b-4281-b951-d872f2087c98

The _Module object is used by ATL COM classes to lock the module, like so

_pAtlModule->Lock();

For example when CreateInstance is called:

BrowserAccessibility* BrowserAccessibility::Create() {
  CComObject<BrowserAccessibilityWin>* instance;
  HRESULT hr = CComObject<BrowserAccessibilityWin>::CreateInstance(&instance);

We need to get rid of WTL so we can get rid of ATL since it
causes many problems. See bug for details.

BUG=5027
TEST=chrome runs. Accessibility works.

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=275987

Review URL: https://codereview.chromium.org/323723002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276360 0039d316-1c4b-4281-b951-d872f2087c98

The current code is calling updateGearMenu_() before changing the
currentVolumeInfo_ that is referred by the updating function.

BUG=none
TEST=Open a zip file. Close by eject icon. Check no error in background page console.

Review URL: https://codereview.chromium.org/328003004

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276359 0039d316-1c4b-4281-b951-d872f2087c98

Allocating these can be expensive (especially on OS X) so it's best not to block the IO thread while doing that.

BUG=

Review URL: https://codereview.chromium.org/317973004

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276358 0039d316-1c4b-4281-b951-d872f2087c98

Adds Rappor metrics for Default Search Engine, Default Startup Tabs, and
New Tab Pages.

BUG=381325

Review URL: https://codereview.chromium.org/311133006

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276357 0039d316-1c4b-4281-b951-d872f2087c98

BUG=368874,368910
TEST=gpu waterfall
TBR=bajones@chromium.org

Review URL: https://codereview.chromium.org/325063002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276356 0039d316-1c4b-4281-b951-d872f2087c98

When we invoke +[ExtensionPopupController showURL:...],
we close the old Popup using [gPopup close],but we don't
release it here. So host_ is only released 
when -[ExtensionPopupController windowWillClose] is invoked.
At this time, the new popup might be ready. And when we use
chrome.extension.getViews or some other apis we might get
the dying window.
So move shutdown of host_ from -[ExtensionPopupController windowWillClose]
into -[ExtensionPopupController close].

BUG=376511

Review URL: https://codereview.chromium.org/321093002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276355 0039d316-1c4b-4281-b951-d872f2087c98

BUG=372212
R=mukai@chromium.org
TBR=sky@chromium.org

Review URL: https://codereview.chromium.org/326073005

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276354 0039d316-1c4b-4281-b951-d872f2087c98

This fixes a bug where links on the page underneath the bubble couldn't
be clicked.  Now, when the permission bubble first appears, the bubble
will not immediately become the key window.  However, if the user clicks
within the bubble, it will be come key window, requiring a click
within the content area to have the content area become key window again.

BUG=381287

Review URL: https://codereview.chromium.org/324223002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276353 0039d316-1c4b-4281-b951-d872f2087c98

BUG=381889

Review URL: https://codereview.chromium.org/326673004

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276352 0039d316-1c4b-4281-b951-d872f2087c98

The bug says to "require" that people put the permissions in their page sets,
but from talking to people it sounded like an optional parameter was better.
If this should be more restrictive, let me know.

R=dtu@chromium.org
BUG=356956

Review URL: https://codereview.chromium.org/324503002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276350 0039d316-1c4b-4281-b951-d872f2087c98

BUG=352447
TBR=sky

Review URL: https://codereview.chromium.org/199153003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276348 0039d316-1c4b-4281-b951-d872f2087c98

R=jbauman@chromium.org

Review URL: https://codereview.chromium.org/327103002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276347 0039d316-1c4b-4281-b951-d872f2087c98

The BookmarkClient can now provide additional nodes to the root of the
BookmarkModel, which are optional. Currently none of those nodes should sync;
this change makes that intention explicit in the bookmarks sync glue code.

BUG=49598

Review URL: https://codereview.chromium.org/319133007

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276346 0039d316-1c4b-4281-b951-d872f2087c98

BUG=none

Review URL: https://codereview.chromium.org/327943004

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276345 0039d316-1c4b-4281-b951-d872f2087c98

Previously decoded frames were returned using the DecodeCB callback,
which is called in response to Decode(). This didn't work very well,
particularly for GPU decoder, because decode buffers do not always map
to decoded frames 1 to 1.

BUG=338529

Review URL: https://codereview.chromium.org/297553002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276344 0039d316-1c4b-4281-b951-d872f2087c98

BUG=None
R=rdevlin.cronin@chromium.org, sky@chromium.org
TBR=sky

Review URL: https://codereview.chromium.org/312943002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276343 0039d316-1c4b-4281-b951-d872f2087c98

If the user opens many tabs and navigates to a NativePage in each tab,
the NativePages may use enough memory to cause a Java out-of-memory error.
This adds a method Tab.freezeNativePage() which destroys the current
native page and replaces it with a FrozenNativePage, which allows querying
getTitle(), getUrl(), etc, but can't be displayed.

BUG=359573

Review URL: https://codereview.chromium.org/322253002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276342 0039d316-1c4b-4281-b951-d872f2087c98

Textfields require RenderTextHarfBuzz to work properly on the Mac because
RenderTextMac's implementation is not complete.
But RenderTextMac is used by the Cocoa UI to perform string sizing and elliding,
HarfBuzz will not give the same answers.
So for now we turn on HarfBuzz conditionally on OS_MACOSX and TOOLKIT_VIEWS.
TextFieldModel and StyledLabel unit tests now pass without failures.

BUG=378134

Review URL: https://codereview.chromium.org/326093004

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276341 0039d316-1c4b-4281-b951-d872f2087c98