The test performed a lot of repetitions of the same append call with a buffer that kept being extended which resulted in OOM errors, but now we use a new buffer each time.

Test: atest CharsetUtf8PerfTest
Bug: 235455275
Change-Id: I10427d315a6d0b45394c74d0c8642f0c085ceb07

In that file, we can list methods which we want to compile but that our
boot image profiling implementation did not cover.

Test: m
Bug: 235557326
Change-Id: Id5241dac826b5d4f581573d8b9b556bde2dabdeb

Merge "[Bugfix][RemoteAnimation] Fix IndexOutOfBoundsException in onAnimationFinished of  RemoteAnimationController"

RemoteAnimationController

Sometimes mPendingAnimations.remove will be called more than once in one cycle (line 255 - 270) of 
onAnimationFinished, e.g., mPendingAnimations.remove being called again in 
adapters.mAdapter.mCapturedFinishCallback.onAnimationFinished, resulting in 
IndexOutOfBoundsException.

Bug:237989368

Change-Id: Ica771f1ae4cb4ddddb7684bd964504ca9f31ea12
Test: Monkey test

* changes:
  Update method visibility for testing
  Stop VPN profiles by exiting VpnRunner instead of prepareInternal
  Update the prefix in keystore for app exclusion

Merged-In: I7cffe34a393991c4ef98740f7c3a052c0b6344fb
Merged-In: Ie0b7151fabd09f58d58745a3f9e2e0252e04411a
Change-Id: Id75747f4a2519f9fa1c7850f99f0298a39dbc79e

* changes:
  Make minimum gc sleep time tunnable
  Make smart idle maint update write amount record every period

RPC Parcels need to be interpreted in the context of an RPC session.
Otherwise, they don't make sense.

Bug: N/A
Test: N/A
Change-Id: I7a7b83341372faae357927f1b23be181b2f907af

A kernel binder object for libbinder can either be a binder object
or an FD.

Bug: N/A (from email thread)
Test: N/A
Change-Id: Iaa02595c5859879749f776222eccca52e782cbfb

Bug:238368970

1. IMediaProjectionCallback#onStop to release virtual display and
remove mVirtualDisplayDevices.
2. DisplayManagerService#releaseVirtualDisplayInternal cannot remove display device due to null mVirtualDisplayDevices.

Signed-off-by: linkai <linkai@xiaomi.com>
Change-Id: Idb3f4aabe894b17a9681dae9f1028bd44ac36f1e

This method clears the calling identity of a kernel
binder call. However, there is now a generic API to
disallow the use of kernel binder calling identities.
This is used by RPC binder calls in order to ensure
that code doesn't accidentally assume the default
(<pid>, <uid>) calling identity means that the call
originated from another process.

In the C++ layer, this API is attached to
IPCThreadState. In the future, we could consider
expanding the scope of this API to code and restore
many types of calling IDs, but the current return
type may not have enough space, and I want to
push people away from thread locals (globals)
for now.

Bug: 237245600
Test: N/A
Change-Id: I6e293814769cbd3c41e72afd95385af31ceb099f

As was pointed out, SharedLog in NetworkStackClient (and
ConnectivityModuleConnector) was only used for dump() functionality
which is not called by anyone. This CL removes the unused dump()
functions and removes SharedLog usage in favor of android.util.Log.

Also cleaning up usage of Slog so only logWtf logs to the system log in
both NetworkStackClient and ConnectivityModuleConnector.

Test: m
Bug: 235901424
Change-Id: Ia484ee825cd714fd8eb7704ca1d16acfbe2e7331
Merged-In: Ia484ee825cd714fd8eb7704ca1d16acfbe2e7331

When using audio-coupled-haptic playback, the haptic channels will be
muted if the app doesn't have the vibrate permission. Update the doc to
state that apps should hold vibrate permission to make
audio-coupled-haptic works as expected.

Bug: 236586224
Test: make
Change-Id: I69c1332d39f997a61f3a969d652144d829de9659

Remove VisibleForTesting for onUserStarted, onPackageRemoved and
onPackageAdded because it's not necessary now.

Bug: 230548427
Test: m ; atest FrameworksNetTests
Change-Id: I531feb457bc59dc9969d93311654523f3e6f5270

This test has been timing out due to it's long runtime and causing other
tests to fail. Since it is already the simplest it can be we should just
delete it.

Bug: 237633188
Test: atest ExpensiveObjectsPerfTest
Change-Id: Id334fc9bd650206d0433433ff27abd6fab695d0a

Test: atest {TestName}
Change-Id: Id6340c9cda83bcadfb5b65dbae0226fcb4eb691c

Chroma channels of YUV420 images are subsampled we may need to shrink
the crop region by a whole texel on each side. Since skia still adds
its own 0.5 inset, we apply an additional 0.5 inset. See
GLConsumer::computeTransformMatrix for details.

Bug: 204725913
Test: android.media.cts.DecodeAccuracyTest
Test: android.view.cts.PixelCopyTest
Change-Id: I38bd3d1e5f4c62fff3dd0374c3f2500f1a2d90b9

- FLAG_SYSTEM had a small typo

Test: Documentation only, not needed
Change-Id: Iaa27b7ccac6090658e5f893778997a6ce850b8b0

Change to call the VpnRunner.exit to stop vpn profile
instead of using prepareInternal to leave the package intact.
This aligns with the way that VpnServices are disconnected
so that the package and other related information will not be
reset unexpectedly.

In current design, Vpn will examine if the current prepared
package matches to prevent an existing always-on VPN from being
dethroned by other apps when the VPN always-on is enabled.
However, when the VPN is disabled using stopVpnProfile, the
current package name will be updated to [LEGACY VPN]. The
current package will no longer be the same with the VPN package.
This resulted in the rejection of the VPN app to start the VPN
using VpnManager.startVpnProfile again.

Test: atest FrameworksNetTests
Test: manually test with VPN app to reconnect VPN when always-on
      is enabled.
Bug: 235322391
Change-Id: I83e1e1edf6c3a6653d87216afcd397f296f59cf2

Update the string without the underscore to prevent conflict with
the existing VPN prefix.

Settings app may get exception when the vpn exclusion list was
set since the logic is used in the Settings relying on the
prefix match. The current app exclusion prefix is also started
with "VPN_", so it will be mis-used by Settings as a VPN profile.

Bug: 237345836
Test: atest FrameworksNetTests
Change-Id: I01e773cc15eb1ae5ffaa12aef124bbbf390cc004

Test: not needed as modifying OWNERS file

Change-Id: Id4d57d2067828bee102ee5d5335b15790e454e9f

Upon deserialization, various APIs would check that the passed in class
was assignable from the creator's enclosing class - which was assumed to
always be the Parcelable type.

This assumption is not always true, so updating the check to explicitely
store the parcelable type

Fix: 232589966
Test: m && atest ParcelTest && atest BundleTest
Change-Id: I59b650a854944e9020615a65798c5e54f5540aaa