GitLab

By default, PopupWindows dismiss when the user touches outside the window. The only
way to avert this behavior when using a ListPopupWindow is to rely on a hidden API.

This manifested as the linked bug because each tap dismissed the popup and detached it from the anchor view. C++ didn't realize the popup was dismissed, and tried to update it, but there was no anchor view.

BUG=400601

Review URL: https://codereview.chromium.org/442933002

Cr-Commit-Position: refs/heads/master@{#291463}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291463 0039d316-1c4b-4281-b951-d872f2087c98

Omnibox extensions don't need a default match from SearchProvider, and
indeed they no longer get them following some refactoring I did last week
( https://codereview.chromium.org/476263002/ ).

BUG=406026

Review URL: https://codereview.chromium.org/490383002

Cr-Commit-Position: refs/heads/master@{#291461}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291461 0039d316-1c4b-4281-b951-d872f2087c98

Previously, a rect was used to convey selection bound edge geometry. This
proved confusing and problematic, particularly for non-axis-aligned text.
Instead, start using the bound edge points provided by Blink.

This depends on the Blink change http://codereview.chromium.org/495673003/.

BUG=405666

Review URL: https://codereview.chromium.org/494823002

Cr-Commit-Position: refs/heads/master@{#291460}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291460 0039d316-1c4b-4281-b951-d872f2087c98

BUG=305722
TBR=ben@chromium.org,pauljensen@chromium.org

Review URL: https://codereview.chromium.org/498793002

Cr-Commit-Position: refs/heads/master@{#291459}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291459 0039d316-1c4b-4281-b951-d872f2087c98

BUG=

Review URL: https://codereview.chromium.org/499563004

Cr-Commit-Position: refs/heads/master@{#291458}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291458 0039d316-1c4b-4281-b951-d872f2087c98

Review URL: https://codereview.chromium.org/486193003

Cr-Commit-Position: refs/heads/master@{#291456}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291456 0039d316-1c4b-4281-b951-d872f2087c98

BUG=394303

Review URL: https://codereview.chromium.org/489493002

Cr-Commit-Position: refs/heads/master@{#291455}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291455 0039d316-1c4b-4281-b951-d872f2087c98

Functionality:

The |AppRegistry| has for each running application an |AppActivityRegistry|.

The |AppActivityRegistry| knows all activities associated with the application it represents.
It can furthermore shut the app entirely down upon resource manager request. It will then create
an |AppActivityProxy| for the overview mode which shows a placeholder for an unloaded app. This
placeholder can then ask the |AppActivityRegistry| to restart the application again.

A shutdown request for the application is only performed when all activities were marked for
UNLOAD.

If there were multiple activities upon shutdown for one app, the app has to take care of
re-creating all windows and thus re-creating all activities. Since an activity match cannot
be performed, the |AppActivityProxy| will only be shown once and it will show in the location
of the most recently used activity of that app. If we later on find an app which really uses
multiple windows and it is imperative to keep the history for all of them tact & the app is
recreating them properly, (a lot of if's) we can revisit the single |AppActivityProxy| and
try to address it in a cleaner way, but at this time that seems rather un-useful since it is
not known if required.

BUG=388085
TEST=AppActivityTest.*

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=291221

Review URL: https://codereview.chromium.org/477523002

Cr-Commit-Position: refs/heads/master@{#291454}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291454 0039d316-1c4b-4281-b951-d872f2087c98

Adds new call on cast_sender: SetTargetPlayoutDelay
Caller is responsible for negotiating with the receiver if the extension is actually supported or not.

BUG=405339

Review URL: https://codereview.chromium.org/493823002

Cr-Commit-Position: refs/heads/master@{#291453}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291453 0039d316-1c4b-4281-b951-d872f2087c98

Use double underscore in the preprocessor symbol.
Fixes build breakage caused by r291315.

Review URL: https://codereview.chromium.org/495333003

Cr-Commit-Position: refs/heads/master@{#291452}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291452 0039d316-1c4b-4281-b951-d872f2087c98

All compositors that use SingleThreadProxy are left calling composite
synchronously and now pass a flag to indicate that this is their
intention.  This patch doesn't remove synchronous composite, but now
makes it mutually exclusive with scheduling.

Only cc unittests are exercising this code at this point, although it
also unifies a lot of the code for CompositeImmediately with the
scheduled path.

BUG=329552, 287250

Review URL: https://codereview.chromium.org/134623005

Cr-Commit-Position: refs/heads/master@{#291451}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291451 0039d316-1c4b-4281-b951-d872f2087c98

    
1) HttpServer gets ServerSocket instead of StreamListenSocket.
2) HttpConnection is just a container for socket, websocket, and pending read/write buffers.
3) HttpServer handles data buffering and asynchronous read/write.
4) HttpConnection has limit in data buffering, up to 1Mbytes by default.
5) For devtools, send buffer limit is 100Mbytes.
6) Unittests for buffer handling in HttpConnection.
    
BUG=371906

Review URL: https://codereview.chromium.org/296053012

Cr-Commit-Position: refs/heads/master@{#291447}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291447 0039d316-1c4b-4281-b951-d872f2087c98

This used to be controlled by a field trial created in
https://codereview.chromium.org/184663002

After evaluating this change, we've decided to launch it.
This change turns the flag on by default and removes the field trial code.
It also removes the index_urls_ parameters everywhere because those will
always be true forevermore.

I tested this interactively.
Also, all the unit tests still apparently pass.  And yes some of these do
exercise this feature.

TBR=joaodasilva
(for trivial change to components/policy/core/browser/managed_bookmarks_tracker_unittest.cc )

BUG=157204,378854

Review URL: https://codereview.chromium.org/489373005

Cr-Commit-Position: refs/heads/master@{#291445}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291445 0039d316-1c4b-4281-b951-d872f2087c98

BUG=none
TEST=none

Review URL: https://codereview.chromium.org/477723003

Cr-Commit-Position: refs/heads/master@{#291444}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291444 0039d316-1c4b-4281-b951-d872f2087c98

WebMediaPlayer should own media_log_, fix the leak there.

Review URL: https://codereview.chromium.org/496103002

Cr-Commit-Position: refs/heads/master@{#291443}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291443 0039d316-1c4b-4281-b951-d872f2087c98

Cr-Commit-Position: refs/heads/master@{#291442}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291442 0039d316-1c4b-4281-b951-d872f2087c98

> NaCl: Update revision in DEPS, r13637 -> r13647
> 
> This pulls in the following Native Client changes:
> 
> r13638: (sbc) Update glibc revision in tools/REVISIONS (9b15f6a->150657d)
> r13639: (mseaborn) PNaCl dynamic loading: Add ELF dynamic loader
> r13640: (jvoung) PNaCl: Update LLVM revision in pnacl/COMPONENT_REVISIONS
> r13641: (mseaborn) Use uint32_t for untrusted pointer args: fstat() and some other syscalls
> r13642: (dschuff) Build PNaCl target libraries on mac and run SCons tests on the toolchain bot
> r13643: (petarj) [MIPS] Change output-format to elf32-tradlittlemips-nacl
> r13644: (petarj) PNaCl: Update binutils revision in pnacl/COMPONENT_REVISIONS
> r13645: (mseaborn) Use uint32_t for untrusted pointer args: gettimeofday() syscall
> r13646: (yyanagisawa) Re-enable use of goma on Windows.
> r13647: (jvoung) PNaCl: Update LLVM revision in pnacl/COMPONENT_REVISIONS
> 
> BUG=none
> TEST=browser_tests and nacl_integration
> CQ_EXTRA_TRYBOTS=tryserver.chromium.linux:linux_rel_precise32,linux_arm_cross_compile,linux_nacl_sdk_build
> TBR=jam@chromium.org
> 
> Review URL: https://codereview.chromium.org/474113003

TBR=mseaborn@chromium.org

Review URL: https://codereview.chromium.org/491053007

Cr-Commit-Position: refs/heads/master@{#291441}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291441 0039d316-1c4b-4281-b951-d872f2087c98

BUG=none

Review URL: https://codereview.chromium.org/488393002

Cr-Commit-Position: refs/heads/master@{#291440}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291440 0039d316-1c4b-4281-b951-d872f2087c98

We can hit this, and it seems removing this is fine.

BUG=400968

Review URL: https://codereview.chromium.org/497573004

Cr-Commit-Position: refs/heads/master@{#291439}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291439 0039d316-1c4b-4281-b951-d872f2087c98

Something changed along the way and it seems the coordinate translation
is necessary for mac too.

BUG=406117
Test=Open a <webview> in a chrome app.
Right click on the <webview>, context menu should appear in correct
position, regardless of the position of the <webview> relative to the
embedder.

Review URL: https://codereview.chromium.org/490293003

Cr-Commit-Position: refs/heads/master@{#291438}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291438 0039d316-1c4b-4281-b951-d872f2087c98

This CL fixes a few issues in the cross-process subframe navigation and re-enables the CrossSiteIframe test on most platforms. There are still some issues on ChromeOS and Android, which will be investigated and fixed in follow-up CL.

BUG=399775, 357747

Review URL: https://codereview.chromium.org/479403004

Cr-Commit-Position: refs/heads/master@{#291437}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291437 0039d316-1c4b-4281-b951-d872f2087c98

Couple of tweaks:
. make WatcherThreadManager maintain a queue of requests and process
  them at once.
. Makes State not unnecessarily cancel requests if it was told the
  handle is ready.

BUG=none
TEST=none
R=darin@chromium.org

Review URL: https://codereview.chromium.org/480293004

Cr-Commit-Position: refs/heads/master@{#291436}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291436 0039d316-1c4b-4281-b951-d872f2087c98

Cr-Commit-Position: refs/heads/master@{#291435}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291435 0039d316-1c4b-4281-b951-d872f2087c98

http://build.chromium.org/f/chromium/perf/dashboard/ui/changelog_blink.html?url=/trunk&range=180709:180779&mode=html
TBR=wibling@chromium.org,gavinp@chromium.org
BUG=

Review URL: https://codereview.chromium.org/484033004

Cr-Commit-Position: refs/heads/master@{#291434}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291434 0039d316-1c4b-4281-b951-d872f2087c98

Cr-Commit-Position: refs/heads/master@{#291433}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291433 0039d316-1c4b-4281-b951-d872f2087c98

This pulls in the following Native Client changes:

r13638: (sbc) Update glibc revision in tools/REVISIONS (9b15f6a->150657d)
r13639: (mseaborn) PNaCl dynamic loading: Add ELF dynamic loader
r13640: (jvoung) PNaCl: Update LLVM revision in pnacl/COMPONENT_REVISIONS
r13641: (mseaborn) Use uint32_t for untrusted pointer args: fstat() and some other syscalls
r13642: (dschuff) Build PNaCl target libraries on mac and run SCons tests on the toolchain bot
r13643: (petarj) [MIPS] Change output-format to elf32-tradlittlemips-nacl
r13644: (petarj) PNaCl: Update binutils revision in pnacl/COMPONENT_REVISIONS
r13645: (mseaborn) Use uint32_t for untrusted pointer args: gettimeofday() syscall
r13646: (yyanagisawa) Re-enable use of goma on Windows.
r13647: (jvoung) PNaCl: Update LLVM revision in pnacl/COMPONENT_REVISIONS

BUG=none
TEST=browser_tests and nacl_integration
CQ_EXTRA_TRYBOTS=tryserver.chromium.linux:linux_rel_precise32,linux_arm_cross_compile,linux_nacl_sdk_build
TBR=jam@chromium.org

Review URL: https://codereview.chromium.org/474113003

Cr-Commit-Position: refs/heads/master@{#291432}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291432 0039d316-1c4b-4281-b951-d872f2087c98

Revert of Improve error reporting when indexing tab. (patchset #7 of https://codereview.chromium.org/449753002/)

Reason for revert:
This patch broke chromeos bot in https://code.google.com/p/chromium/issues/detail?id=406010

A reland of this must include fixes in:
https://codereview.chromium.org/484353003/
https://codereview.chromium.org/499453002/

TBR=tonyg@chromium.org,kbr@chromium.org,jennyz@chromium.org,achuith@chromium.org
NOTRY=true
NOTREECHECKS=true
BUG=405955,406010

Original issue's description:
> Improve error reporting when indexing tab.
> 
> If we created 3 new tabs, and 1 of them crashed, then there 
> is no guaranteed that  browser.tabs[1] return the correct tab 
> object. The tabs list can no longer be trusted in this case. Hence if 
> the number of inspectable contexts doesn't match the number 
> of expected tabs (based on the number of 
> tab_list_backend.New & tab_list_backend.Close calls), 
> tab_list_backend's __getitem__ throws a TabCrashException.
> 
> BUG=398467
> 
> Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=290974

Review URL: https://codereview.chromium.org/494753003

Cr-Commit-Position: refs/heads/master@{#291431}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291431 0039d316-1c4b-4281-b951-d872f2087c98

Resubmitting the patch 
https://src.chromium.org/viewvc/chrome?view=rev&revision=291365
which was reverted as it caused a FindBug failure due to not calling 
the Super method in the setUp method in PopupZoomerTest class.

Adding call to to super.setUp() to fix the warning.

BUG=405477, 406272

Review URL: https://codereview.chromium.org/497753003

Cr-Commit-Position: refs/heads/master@{#291430}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291430 0039d316-1c4b-4281-b951-d872f2087c98

BUG=401644

Review URL: https://codereview.chromium.org/496873005

Cr-Commit-Position: refs/heads/master@{#291429}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291429 0039d316-1c4b-4281-b951-d872f2087c98

extensions::InfoMap lives in src/extensions, so run its test suite as part of extensions_unittests.

Two tests are left in Chrome's unit_tests because they depends on the "notifications" permission being available to hosted apps.

BUG=397164
TEST=unit_tests ChromeInfoMapTest, extensions_unittests InfoMapTest

Review URL: https://codereview.chromium.org/473913006

Cr-Commit-Position: refs/heads/master@{#291428}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291428 0039d316-1c4b-4281-b951-d872f2087c98

Remove largest_observed_ from QuicSentPacketManager, since it's also in
QuicUnackedPacketMap.

Merge internal change: 73713269
https://codereview.chromium.org/491663003/


Optimize QUIC's EntropyTracker in QuicReceivedPacketManager by using a
deque.

Merge internal change: 73709652
https://codereview.chromium.org/496133002/


Add logic for a client to enable pacing via config to the QUIC sent
packet manager.

Merge internal change: 73685838
https://codereview.chromium.org/497553003/


Clear the QUIC ping alarm when a connection in closed.

Merge internal change: 73684904
https://codereview.chromium.org/475173005/


Rename QuicConnection::OnPacketSent to OnWriteError, to reflect it's
actual functionality, and to only be called when there is an error.
Clean up some logging when packets are sent.

Merge internal change: 73651039
https://codereview.chromium.org/483093004/


Refactor QuicConnection's async write behavior to record a packet as
sent as soon as the write starts, not when it completes.

In QuicStreamFactoryTest, MockClock's now() is returning 0 and that
was causing a DFATAL in QuicUnackedPacketMap::GetLastPacketSentTime().
Noticed, in QuicConnection unittests, we are advancing clock by 1 second
and made a similar change to QuicStreamFactoryTest's MockClock.

Merge internal change: 73648585
https://codereview.chromium.org/473763004/


QUIC minor clean up to keep in sync with internal sources.

+ Use int64 for MaxBandwidthTimestamp (per rjshade).
+ Use DVLOG(1) instead of DLOG(INFO).
+ Reorder class statements.

This CL has changes that were found while merging (or back porting) the
following chroimum specific changes into the internal source tree.

Merge internal change: 73640117
Merge internal change: 73638556
https://codereview.chromium.org/495233002/


Move least_packet_awaited_by_peer_ from QuicReceivedPacketManager to
QuicSentPacketManager.

Merge internal change: 73630593
https://codereview.chromium.org/493183002/


Remove timestamp_receiver. Not used.

Merge internal change: 73627146
https://codereview.chromium.org/495173002/


Remove unused protected received_packet_manager_ accessor in
quic_connection.

Merge internal change: 73626837
https://codereview.chromium.org/496823002/


Convert QUIC's SentEntropyManager to use a deque instead of a map in
order to improve memory and CPU efficiency.

Merge internal change: 73625323
https://codereview.chromium.org/479543004/


When talking >=QUIC_VERSION_22, regularly send updated bandwidth
estimates to the client (in SCUP messages).

The main work is in QuicConnection, where a new alarm regularly checks
for substantially changed bandwidth estimates and sends them to the
client if so. The majority of the rest of this CL is plumbing, and test
code.

Feel free to suggest better values for:
- Timeout between bandwidth estimate alarms
- Definition of "substantial change" in estimate

The flow is:

- on every congestion event (ACK, packet loss) the sent packet manager
  passes a bandwidth estimate to the sustained bandwidth recorder
- on every congestion window change, the QuicServerSession checks to see
  if the bandwidth has changed sufficiently *and* enough time has passed
  since last time we sent an update to the client
- if so, it populates the CachedNetworkParams proto, and passes this to
  the CryptoStream which sends a SCUP message to the client

Merge internal change: 73579021
https://codereview.chromium.org/490263003/

R=rch@chromium.org

Review URL: https://codereview.chromium.org/497553004

Cr-Commit-Position: refs/heads/master@{#291427}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291427 0039d316-1c4b-4281-b951-d872f2087c98

Revert of Move actions after browser Start in cros_browser_backend.Start to DidStart. (patchset #1 of https://codereview.chromium.org/499453002/)

Reason for revert:
The original patch of this (https://codereview.chromium.org/449753002/) broke chromeos bot in https://code.google.com/p/chromium/issues/detail?id=406010

Original issue's description:
> Move actions after browser Start in cros_browser_backend.Start to DidStart.
> 
> BUG=406010
> NOTRY=true
> NOTREECHECKS=true
> 
> Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=291310

TBR=jennyz@chromium.org,achuith@chromium.org
NOTREECHECKS=true
NOTRY=true
BUG=406010

Review URL: https://codereview.chromium.org/502503002

Cr-Commit-Position: refs/heads/master@{#291426}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291426 0039d316-1c4b-4281-b951-d872f2087c98

Sanity DCHECK for IncognitoModePrefs skipped for guest profile.

BUG=395044

Review URL: https://codereview.chromium.org/496713003

Cr-Commit-Position: refs/heads/master@{#291425}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291425 0039d316-1c4b-4281-b951-d872f2087c98

Revert of Remove the assertion that number of expected tabs must equal to number of live tabs temporarily. (patchset #1 of https://codereview.chromium.org/484353003/)

Reason for revert:
The original patch of this (https://codereview.chromium.org/449753002/) broke chromeos bot in https://code.google.com/p/chromium/issues/detail?id=406010

Original issue's description:
> Remove the assertion that number of expected tabs must equal to number of live tabs temporarily.
> 
> BUG=405955
> NOTRY=true
> 
> Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=291368

TBR=tonyg@chromium.org,dtu@chromium.org
NOTREECHECKS=true
NOTRY=true
BUG=405955

Review URL: https://codereview.chromium.org/495393002

Cr-Commit-Position: refs/heads/master@{#291424}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291424 0039d316-1c4b-4281-b951-d872f2087c98

 * AppListViewDelegate was accessing deleted search_provider.
 * DeviceSocketListener::StopListening can happen after DeviceSocketManager is deleted
 * Explicitly delete FilePathWatcher. This was causing
   recursive callback to FilePathWatcherImpl::Cancel from
   FilePathWatcherImpl::CancelOnMessageLoopThread, which
   caused crash.

Clean ups
 * change OnIO to OnFILE as they run on FILE thread.
 * removed unused singleton related code.

BUG=None
TEST=Run athena_main on desktop and close window.

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=291214

R=flackr@chromium.org, mukai@chromium.org

Review URL: https://codereview.chromium.org/490033003

Cr-Commit-Position: refs/heads/master@{#291423}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291423 0039d316-1c4b-4281-b951-d872f2087c98

This CL forces update of system tray state on new IME State even if
current input method has not changed.

BUG=406186
TEST=manual

Review URL: https://codereview.chromium.org/496993003

Cr-Commit-Position: refs/heads/master@{#291422}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291422 0039d316-1c4b-4281-b951-d872f2087c98

TBR=mtklein@chromium.org,caryclark@google.com
BUG=406425

Review URL: https://codereview.chromium.org/497793002

Cr-Commit-Position: refs/heads/master@{#291421}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291421 0039d316-1c4b-4281-b951-d872f2087c98

There was a downstream unit test package invoking these, not just native
callers.  In order to let downstream autoroll cleanly, bring them back for
now.  These can be deleted later in the same cleanup CL as the others.

NOTRY=true
TBR=tedchoc@chromium.org
BUG=406332

Review URL: https://codereview.chromium.org/477133006

Cr-Commit-Position: refs/heads/master@{#291420}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291420 0039d316-1c4b-4281-b951-d872f2087c98

BUG=402050

Review URL: https://codereview.chromium.org/496853002

Cr-Commit-Position: refs/heads/master@{#291419}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291419 0039d316-1c4b-4281-b951-d872f2087c98

BUG=403810
R=oshima@chromium.org

Review URL: https://codereview.chromium.org/488093004

Cr-Commit-Position: refs/heads/master@{#291417}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291417 0039d316-1c4b-4281-b951-d872f2087c98