Test: Connect android to windows through usbc.
      Enable mtp, then enable the transcoding option.
      Copy H265 video from the camera folder to windows' desktop.
      Transcoded H264 video plays correctly in WMPcorrectly.

Bug: 184117074
Bug: 190422448
Change-Id: I17264f0ee3e742315569d4c50817b1da7f3f2eb4

Some Windows MTP clients ignore the file size sent from the server
while sending MTP objects, and instead rely on the cached MTP metadata
containing the size. This causes file corruption when transcoded files
are transferred from phone to PC because the file is truncated to the
original file size from the cached metadata which is smaller than the
transcoded file size.

To workaround this, MediaProvider added an ugly workaround to the
struct stat st_nlink field to signal when a file requires
transcoding. In such cases, the MTP server can double the metadata
size, the same heuristic used in the MediaProvider to accommodate the
transcoded file size.

Test: atest MtpTests
Bug: 184117074
Bug: 190422448
Change-Id: I0581978fb346e3dc2d7a5204ffd7c601b61ec749

This makes art/build/boot/boot-image-profile.txt +
frameworks/base/boot/boot-image-profile.txt identical to
frameworks/base/config/boot-image-profile.txt, so we can rule out
performance regressions due to profile changes in
https://r.android.com/1896660.

Test: presubmits
Bug: 207590908
Change-Id: I599e8df8818d2d2dc9f471fdc22c9d92719e6cae

Test: N/A
Change-Id: Ia2c1cf25e9750af0374816c06b2b9184f63dc261

Move UnderlyingNetworkRecord to a spearate file, and
move calculatePriorityClass to a new class NetworkPriorityClassifier.

This is a preparation CL to support vcn routeselection with
configurable network priority classes and network metric
monitors.

Bug: 206044122
Test: atest FrameworksVcnTests, CtsVcnTestCases
Change-Id: I88aa9e8dcf01524ddce9f913f40a4c712f8c878d

Bug: 206044122
Test: atest FrameworksVcnTests, CtsVcnTestCases
Change-Id: I6b4888f8252999446a29076937d9695278ff6a66

Devices upgrading to S are not required to update their memtrack HAL
implementation and won't support reporting the total GPU private memory.

In these cases it's likely the HAL reports 0 for the total GPU private
memory, since in S the HAL defition only overloads the already existing
getMemory() API to report this.

Return an error value in these cases to prevent swapping the sum of GL
type memory for all processes with the total GPU private memory when not
supported as doing this effectively leaves the GL memory as part of
lost RAM.

Bug: 203245390
Test: dumpsys meminfo
Change-Id: I478143e435b047bf349bd4ec6808707a5db1505b

Bug: 207464971
Test: w/o LE audio device, and turn on/off airplane mode
Change-Id: I1865c5ff7e8c04a4188dc7c379223e48c8b29ad9

Bug: 206676415
Test: manual
Change-Id: I2af0ed4645fbae621431ac709df3b1e6d540f271

* changes:
  replaced android.se.omapi-V1-ndk_platform library with android.se.omapi-V1-ndk
  VTS tests added for OMAPI Vendor stable interface
  Support for OMAPI Vendor stable interface

Bind to the TvInputService even if TvInputService
calls addHardwareInput(), addHdmiInput(), or
removeHardwareInput().

Bug: 207301256
Test: manual
Change-Id: I91cdf49f9ff1af240b5d87cec02e95ae305cf46f

StorageManagerService and AppFuseBridge are used
NativeDaemonConnectorException to represent app fuse mount
failure. But NativeDaemonConnectorException is part of NSD
module files and it will be moved to the Connectivity module.
Thus, create AppFuseMountException to separate the usage and
make the exception more precisely.

Bug: 206702844
Test: m
Change-Id: I8b0224ce8894bc68b082dcdba054e2ccc22869ef

* changes:
  NFC support in work profile
  NFC support in work profile

Bug: 207311784
Test: scan device normal
Change-Id: I3bb628871b4921f87c92e03386bcb6f66d90ae9f

The code residing in packages/modules/Bluetooth/android/app is
migrated from packages/apps/Bluetooth and requires the same
visibility rules for frameworks source code.

This path currently does not exist, but will be required for
migration.  Soong build rules do not enforce path existence, which
allows us to add the new path now and not during migration.

Added visibility along with migrated code in a local tree and
performed a local build (bramble-userdebug)

BUG: 196026708
Test: TH
Change-Id: I66cac592489fc1689d45a7b6427bc58c40a824e2

clock_nanosleep will return 0 or error no directly.

Bug: 207149564
Test: Build ok
Change-Id: I15866728719d0672a16b4a76a8fd3be4d53d420a

This is a blocking event, needs to be moved off from the main thread.

Also, don't schedule background processing service if the hardware
has no tracing support.

Test: manual
Bug: 79161490
Bug: 206986559
Change-Id: I039211b26f25ead66cf125f152e794b0c08057ee