GitLab

Style guidelines encourage using either BindOnce() or BindRepeating(),
not Bind(). This CL converts some Bind() calls to their respective
type.

Change-Id: I7e816bfd102f9e28989cff40003fc0f7192458d8
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1546887
Commit-Queue: Kevin Bailey <krb@chromium.org>
Reviewed-by: Tommy Li <tommycli@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646592}

Bug: 944115
Change-Id: I42573db7c4e99f21c9ef208aeb31e0a10b9f10f3
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1548319
Auto-Submit: Tibor Goldschwendt <tiborg@chromium.org>
Commit-Queue: Andrew Grieve <agrieve@chromium.org>
Reviewed-by: Andrew Grieve <agrieve@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646591}

This CL :
1 - Disables the download auto-resumptions in native feature
2 - Adds a flag for testing purposes

Bug: 948299
Change-Id: I4a04eb17e61e40e9a4d471e1bfaccd77c0029270
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1547763Reviewed-by: Min Qin <qinmin@chromium.org>
Commit-Queue: Shakti Sahu <shaktisahu@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646590}

Bug: chromium:698902
Change-Id: Ic2963f13439db4176e9650e6f111e42dbcc753d1
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1548377Reviewed-by: Kenneth Russell <kbr@chromium.org>
Commit-Queue: Rakib Hasan <rmhasan@google.com>
Cr-Commit-Position: refs/heads/master@{#646589}

We want to be able to use smaller VMs for these tester-only machines,
and currently they're defaulting to requiring 8 cores.

This will be followed by a CL that removes this requirement for the
same VMs on the chromium.gpu waterfall.

Bug: 881625
Bug: 948220
No-Try: True
Change-Id: Iaf5400bfe7319be37a045b0cc781864eb6755f42
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1548545
Commit-Queue: Kenneth Russell <kbr@chromium.org>
Commit-Queue: Jao-ke Chin-Lee <jchinlee@chromium.org>
Auto-Submit: Kenneth Russell <kbr@chromium.org>
Reviewed-by: Jao-ke Chin-Lee <jchinlee@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646588}

The corresponding histograms were removed, but we forgot to remove
this bit of code.  Remove it.

Bug: 946853, 774526
Change-Id: Idfaf4f44083e97bb0082ada7f215ff3efb79a642
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1542671Reviewed-by: Hongchan Choi <hongchan@chromium.org>
Commit-Queue: Raymond Toy <rtoy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646587}

Add a new field in scripts to allow setting values to a form field with
keyboard presses simulation. This allows scripter to try different way if
some fields are not working with setting value directly.

Change-Id: I87818605fb156470606eb06bf3091da125f79c24
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1542053Reviewed-by: Mathias Carlen <mcarlen@chromium.org>
Reviewed-by: Clemens Arbesser <arbesser@google.com>
Commit-Queue: Hui Wu <wuandy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646586}

Bug: chromium:698902
Change-Id: I861eaa6596d2d9240f16514b2b6fe2287d0fd2ca
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1548132Reviewed-by: Kenneth Russell <kbr@chromium.org>
Commit-Queue: Rakib Hasan <rmhasan@google.com>
Cr-Commit-Position: refs/heads/master@{#646585}

String.append does an alloc and free on each append.

BUG=268281

Change-Id: Ie0cfdfd1741624afcbb82c0269acb772d2653411
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1545765Reviewed-by: Kentaro Hara <haraken@chromium.org>
Commit-Queue: Dave Tapuska <dtapuska@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646584}

For some provider, a media status changed event was triggered by a
intermediate (probably caused by provider overriding the playback state)
where audio focus state is suspended while playback state is playing.

We considered this as invalid state and skip updating.

Bug: b:128037880
Change-Id: I9654eb4ac7b78c5c63ecba1dfefc9033ef9797a3
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1546254
Commit-Queue: Li Lin <llin@chromium.org>
Reviewed-by: Xiaohui Chen <xiaohuic@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646583}

TBR=hbos@chromium.org

Bug: chromium:948055
Change-Id: I93faa0fc0b164a56fd34862c1c213c2a4305b474
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1547708Reviewed-by: Harald Alvestrand <hta@chromium.org>
Commit-Queue: Harald Alvestrand <hta@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646582}

Previously symbolizer.py was using a for loop to read the symbolizer
output. Iterator implementation in File class adds buffering. As result
output was often shown on the screen with delay. Replaced for loop with
a while loop that invokes readline(), which avoids the buffering issue.

Bug: 943745
Change-Id: Ie37c2a001052aa38506a14e48035c382bda9e268
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1544621
Commit-Queue: Sergey Ulanov <sergeyu@chromium.org>
Commit-Queue: Kevin Marshall <kmarshall@chromium.org>
Reviewed-by: Kevin Marshall <kmarshall@chromium.org>
Auto-Submit: Sergey Ulanov <sergeyu@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646581}

Bug: 945906
Change-Id: I4b085fa0a7f62b7b331ae99a48fb0a3dd5d46422
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1547533Reviewed-by: Theresa <twellington@chromium.org>
Commit-Queue: Sinan Sahin <sinansahin@google.com>
Cr-Commit-Position: refs/heads/master@{#646580}

Parameterized tests need to be careful not to allocate data before
WTF is initialized. Some NG Layout tests tried to work around this
by initializing partitions before accessing the WTF::operator<< which
calls ToString() on objects. Just replace the operator<< with a simple
test case string to avoid the complicated logic of trying to get
partitions and WTF initialized.

PaymentRequestDetails was marked as a fast malloc case and it really
should,'t be. It was working because of the partition initialization by
the layout tests. But when the test case is initialized by gtest it is
done before blink is initialized.

Change-Id: I4377aeb7b207392b794a4406089643cd5d2bbea1
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1547346Reviewed-by: Kentaro Hara <haraken@chromium.org>
Commit-Queue: Dave Tapuska <dtapuska@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646579}

All uses of these functions have been converted to NumberToString.

Bug: 929827
Change-Id: I60365fdf05f4b151dbf7230464d3ad98778273af
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1545881
Auto-Submit: Raul Tambre <raul@tambre.ee>
Commit-Queue: Daniel Cheng <dcheng@chromium.org>
Reviewed-by: Daniel Cheng <dcheng@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646578}

Currently, skeleton generation of engaged sites is done by the navigation
throttle on every navigation. To make this more effective, this CL precomputes
skeletons in the lookalike service instead when it fetches the list of engaged
sites.

Change-Id: I252c8ebdcc21f7a91b7b40832fe664153ec10f3f
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1546711
Commit-Queue: Mustafa Emre Acer <meacer@chromium.org>
Reviewed-by: Joe DeBlasio <jdeblasio@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646577}

There has been some flakiness in the UIA event-dump tests during recent
try-jobs.  It looks like most of the failures are due to spurious focus
events on the document object.

There was already a default filter rule to ignore these events for MSAA
tests; this change adds a rule to ignore the corresponding UIA event.

In addition, a number of tests using "continuations" receive spurious
focus events (on elements other than the document) at the start of a
new section.  This change filters out the focus events, which aren't
germane to the tests anyway.

Bug: 928949
Change-Id: If01a76aa36efe52fb6dd8f1c683c077d469efef6
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1545081Reviewed-by: Dominic Mazzoni <dmazzoni@chromium.org>
Commit-Queue: Ian Prest <iapres@microsoft.com>
Cr-Commit-Position: refs/heads/master@{#646576}

This can lead to two peers deadlocking, including between processes: If
peer A is sending lots of messages to peer B, and B's kernel message
queue attached to the receive port is full, A would block during the send.
If at the same time B were sending lots of messages to A, and A's kernel
message queue became full, B would also block. Neither A nor B would be
able to make process because the IO threads that would receive messages
are blocked in a send. (This assumes that only one thread is sending and
receiving for each peer).

Instead, MACH_SEND_MSG with 0 timeout. If the send times out, indicating
that the receiver's kernel message queue is full, set a flag to start
queueing future messages and arrange to re-send the failed one.

Bug: 932175
Change-Id: I8d2047485bffcabfecd2009b609d494f3c4b6551
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1544310Reviewed-by: Ken Rockot <rockot@google.com>
Commit-Queue: Robert Sesek <rsesek@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646575}

This histogram is intended to report network only ad bytes for
the pageload. However, change
https://chromium-review.googlesource.com/c/chromium/src/+/1478091,
introduced a bug that caused it to record network and cache bytes.
This change fixes the error and versions the histogram. It also
adds a test to verify that only network bytes are recorded.

Change-Id: Ia97c0d4107562bfe7714abf14e9940acaa9645d3
Bug: 948130
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1544824Reviewed-by: Charlie Harrison <csharrison@chromium.org>
Commit-Queue: John Delaney <johnidel@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646574}

Bug: 947738
Bug: 947834
Bug: 948171
Bug: 947032
Change-Id: Ic41418ef66e76375171ba3e36c660cada83ee0bd
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1546858
Commit-Queue: Łukasz Anforowicz <lukasza@chromium.org>
Commit-Queue: Karan Bhatia <karandeepb@chromium.org>
Auto-Submit: Łukasz Anforowicz <lukasza@chromium.org>
Reviewed-by: Karan Bhatia <karandeepb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646573}

BUG=554299

Change-Id: Ic6d1418743812a444b80e101c2c73347e1f00f06
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1546159Reviewed-by: Bo <boliu@chromium.org>
Reviewed-by: Ken Rockot <rockot@google.com>
Reviewed-by: Peter Williamson <petewil@chromium.org>
Reviewed-by: David Trainor <dtrainor@chromium.org>
Commit-Queue: Avi Drissman <avi@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646572}

Bug: 746182
Change-Id: Iaa0a2e5f7c3950f6fe2d1e1a1e14dd2d501311cd
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1541584Reviewed-by: Frank Liberato <liberato@chromium.org>
Commit-Queue: Mounir Lamouri <mlamouri@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646571}

Bug: chromium:698902
Change-Id: I961197fde9627078743fb3b3dc76ff4a97bcae53
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1545228Reviewed-by: Kenneth Russell <kbr@chromium.org>
Commit-Queue: Rakib Hasan <rmhasan@google.com>
Cr-Commit-Position: refs/heads/master@{#646570}

Remove unused includes and add missing ones.

BUG=chromium:948309
TEST=code compiles

Change-Id: Idc271c561e0569ab4d87dd5b657bac1c5b460d49
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1547798Reviewed-by: Tommy Li <tommycli@chromium.org>
Commit-Queue: Maksim Ivanov <emaxx@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646569}

It's been unused since codereview.chromium.org/2518493002 .

BUG=chromium:948309

Change-Id: I8e11af84062d232a43f70af508debde98d935fb7
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1547707Reviewed-by: Tommy Li <tommycli@chromium.org>
Commit-Queue: Maksim Ivanov <emaxx@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646568}

R=abdulsyed@chromium.org

Change-Id: Ia5895d5929e338ae165eb2905923c5f890d7d4fc
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1548354Reviewed-by: Mustafa Emre Acer <meacer@chromium.org>
Commit-Queue: Krishna Govind <govind@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646567}

Also add some sRGB pixel color space tests.

sRGB colorspace + float16 pixelFormat is unsupported with accelerated
canvases, keep that failing for now:
https://bugs.chromium.org/p/chromium/issues/detail?id=948141

Bug: 946514
Change-Id: I8ab85d57edc1edcce9a08e889382e7e357f4f10f
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1544232
Commit-Queue: Aaron Krajeski <aaronhk@chromium.org>
Reviewed-by: Antoine Labour <piman@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646566}

This CL adds a new observer method for that.
We don't really need the OnManagerGoingDown() method since the coordinator
can live longer

BUG=942770

Change-Id: Ieaf3e13facc52022f2ec35c8cdf0f3a7c0170fd5
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1546259
Commit-Queue: Min Qin <qinmin@chromium.org>
Reviewed-by: Shakti Sahu <shaktisahu@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646565}

Mostly similar to what we've done for DownloadInSandbox with the following
differences:
- For ad, we collect 2 UseCounters - both with & without gesture, as the usage
is quite low so it's possible we block them all in the future.
- Use base::FeatureList rather than RuntimeEnabledFeature to avoid plumbing
extra information through layers. (For sandbox intervention
RuntimeEnabledFeatures is used for WPT coverage before the feature is default
to enabled)
- This CL is missing the part where the frame navigating is an ad, as we don't
know the ad status of remote frame. Subsequent CL will address this issue which
will probably query this info in the browser process.

Bug: 929911
Change-Id: Ib8885ef5b3e349cef3250d354561c16fdc28aab7
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1532996
Commit-Queue: Yao Xiao <yaoxia@chromium.org>
Reviewed-by: Jochen Eisinger <jochen@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646564}

- PrinterQueue is an input text field and should be part of
  printerInfoChanged_.

Bug: 896637
Test: BrowserTest/gtest_filter=CrSettingsPrinting*
Change-Id: Ie0c02e06d80f1c8f232406c9245093c2cbf6d9a7
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1535192Reviewed-by: Steven Bennetts <stevenjb@chromium.org>
Reviewed-by: Bailey Berro <baileyberro@chromium.org>
Commit-Queue: jimmy gong <jimmyxgong@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646563}

This CL adds support for requesting the result of a JPEG decode in an
arbitrary FOURCC. This will allow us to eventually get the result in
RGBX format. It also allows us to refactor how we handle the case in
which regardless of whether the input JPEG is YUV420 or YUV422, we want
I420 output: this is okay with Intel but not with AMD Stoney Ridge
(i.e., grunt). There was already code to handle this Intel/AMD
discrepancy. However, this relied on the grunt driver not reporting I420
as a supported image format which seems incorrect because it can convert
to IYUV (synonym of I420) in vaGetImage() (see
https://gitlab.freedesktop.org/mesa/mesa/commit/b0a44f10).

This CL centralizes the determination of a suitable output FOURCC in the
new VaapiWrapper::GetJpegDecodeSuitableImageFourCC() static method. The
caller can express a preferred FOURCC. If the conversion is not
supported, this method returns a fallback FOURCC that is supported.

Unit tests are added to make sure that all the combinations of supported
subsamplings and supported image formats are successful: for now, we can
check the decode result against software decoding as long as the output
is either I420 or YUY2/YUYV.

Test: JDA tests pass on grunt, nocturne, and samus.
Bug: 868400
Change-Id: I1a8dee1599469544bbaf6fd40bcd5662f7637fdc
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1494803
Commit-Queue: Andres Calderon Jaramillo <andrescj@chromium.org>
Reviewed-by: Miguel Casas <mcasas@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646562}

This CL creates helper classes for creating node properties with
corresponding observer notifications, and it refactors all existing
node properties to use the helper classes. This helps remove
redundant code and makes the class declarations easier to read.

BUG=

Change-Id: I440f3c0d40ca70f537525cfa5ec04ae38f1a5f5f
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1546874
Commit-Queue: Chris Hamilton <chrisha@chromium.org>
Reviewed-by: Sigurður Ásgeirsson <siggi@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646561}

And remove a couple old suites from the fyi bot.

Bug: 946685
Change-Id: I621e6482c39b9af47173db644a5a1866741fa7eb
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1548054Reviewed-by: John Budorick <jbudorick@chromium.org>
Commit-Queue: Ben Pastene <bpastene@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646560}

This CL adds search functionality to the activity log stream. It works
slightly differently than searching in the history tab.

Stream search is done on the client side and is a case-insensitive,
partial match on an activity's name (API call or content script name),
activity type or page url. Activities not matched with the search term
(pre-existing and new activities) will appear/reappear when the search
term is modified to include them.

Screenshot: https://imgur.com/a/o00bDWu

Bug: 932768
Change-Id: Ib535d6aa231b9c6789eb701be22ca6d9b7200fca
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1542716Reviewed-by: Demetrios Papadopoulos <dpapad@chromium.org>
Reviewed-by: Esmael El-Moslimany <aee@chromium.org>
Commit-Queue: Kelvin Jiang <kelvinjiang@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646559}

The comments previously said that |kMsgError| was a non-terminal state
that could be recovered from by sending a valid message. However, that
wasn't really true: the private |state_| member needs to be forced to
|kReady| in order to recover. |FidoDevice| itself did that to handle
CTAP2 GetInfo requests resulting in certain errors, but nothing else
can.

However, other code tested only |kDeviceError| and ignored |kMsgError|.

So |kMsgError| should be seen as a hack to work around devices
potentially rejecting GetInfo requests. In all other ways, |kMsgError|
and |kDeviceError| are essentially the same state.

Also fix a crash when a BLE device returns an empty error.

Change-Id: I7d9a8eb70287db87eb7e7892d8b752c1311e3122
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1546720Reviewed-by: Martin Kreichgauer <martinkr@google.com>
Commit-Queue: Martin Kreichgauer <martinkr@google.com>
Auto-Submit: Adam Langley <agl@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646558}

This is a reland of f7bfe7d0

This version updates gpu/vulkan/generate_bindings.py instead of
changing vulkan_function_pointers.[h|cc] directly.

Original change's description:
> Add gpu::SemaphoreHandle and use it in VulkanImplmentation
>
> Previously VulkanImplementation defined functions for semaphore
> import/export from/to file descriptors. This approach doesn't work on
> platforms that don't support FDs for semaphores, specifically on
> Fuchsia and Windows.
>
> This CL:
> 1. Adds cross-platform gpu::SemaphoreHandle that allows to
> represent semaphore handle on any platforms.
> 2. Replaces ImportSemaphoreFdKHR() with ImportSemaphoreHandle().
> 3. Replaces GetSemaphoreFdKHR() with GetSemaphoreHandle()
>
> This allows to remove dependency on ScopedFD in
> gpu::ExternalVkImageFactory, which will make it easier to reuse that
> class on Fuchsia and Windows.
>
> Bug: 934526
> Change-Id: I16b1a9f3a606683b76a73c4e68a33b3294f25a90
> Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1540724
> Commit-Queue: Sergey Ulanov <sergeyu@chromium.org>
> Commit-Queue: Bo <boliu@chromium.org>
> Auto-Submit: Sergey Ulanov <sergeyu@chromium.org>
> Reviewed-by: Bo <boliu@chromium.org>
> Reviewed-by: Antoine Labour <piman@chromium.org>
> Reviewed-by: Michael Spang <spang@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#646424}

TBR=piman@chromium.org (for reland CL).

Bug: 934526
Change-Id: If0a2aaf30558a2801dc8719f80d7e914246274d1
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1548353
Commit-Queue: Sergey Ulanov <sergeyu@chromium.org>
Reviewed-by: Peng Huang <penghuang@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646557}

This removes the context argument from the FontSizePrefs instantiation
call.

Bug: 938537
Change-Id: I8029392c1b81921089997ac5b8379ae98e1f684b
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1546940
Auto-Submit: Mehran Mahmoudi <mahmoudi@chromium.org>
Commit-Queue: Michael Thiessen <mthiesse@chromium.org>
Reviewed-by: Michael Thiessen <mthiesse@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646556}

- We now instead just limit their logon hours so that their tile
still appears but they can't login.
- This reduces code complexity since we do not need anonymous
reauth credentials to show users that have been denied access.
- This also fixes weird cases where the "other user" tile would
appear along with the normal users on the system.

Bug: 935695
Change-Id: Ie826cbe3b7cd67982f0a260fe7db446470d57a17
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1526000
Commit-Queue: Tien Mai <tienmai@chromium.org>
Reviewed-by: Roger Tawa <rogerta@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646555}

Bug: chromium:698902
Change-Id: Ie552751c4e2232f6332884f012ea3b71e98d476b
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1546787Reviewed-by: Kenneth Russell <kbr@chromium.org>
Commit-Queue: Rakib Hasan <rmhasan@google.com>
Cr-Commit-Position: refs/heads/master@{#646554}

When one drag indicator expands to become a preview area, the other
indicator contracts to 20 dips. The present CL makes it 3% screen width
as specified here:
https://mccanny.users.x20web.corp.google.com/www/splitscreen-motion/index.html#indicator-contract

Test: manual
Bug: 948221
Change-Id: I54a56ef237f6845b966a92117cca73f9b5f53331
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1547946Reviewed-by: Xiaoqian Dai <xdai@chromium.org>
Commit-Queue: Avery Musbach <amusbach@chromium.org>
Cr-Commit-Position: refs/heads/master@{#646553}