Bug: 242170869
Test: m
Change-Id: Ie7e39c8db95d66033f77881dc076099664e0bb2d

Bug: 282715698
Test: presubmit
(cherry picked from https://googleplex-android-review.googlesource.com/q/commit:1814dffd1c6f9edc0115ac73a434263321d77db9)
Merged-In: Ifd0de818ee99dfd6e44750f9bff4a6a7ef1e6dd2
Change-Id: Ifd0de818ee99dfd6e44750f9bff4a6a7ef1e6dd2

This partially reverts commit cd1e70b8.

Breakages are fixed in the current stable Clang.

Test: presubmit
Bug: 169004486
Change-Id: I545553eccf9ed83864158087c5746923c3d452d5

Apple Wallet's `.pkpass` and `.pkpasses` file are considered as `application/octet-stream` in Android.

Adding the mapping in android.mime.types so those files could be resolved correctly.

Change-Id: Ic3315bc24441b134bd594237a35cacef77ec9082

Bug: 310547345
Change-Id: I3f702495ec9877a3c2bbac0b4470baf512da26ed
Test: n/a

If the client process was killed by lmkd or other,
and its pid rescued by any process. Then it may be
killed by mistaken.

Bug: 312576121

Change-Id: I577372953c5f9090d11dbf9ca5bd67e1f5dd9408
Signed-off-by: liulvping <liulvping@xiaomi.com>

Bug: 309596860
Test: NA
Change-Id: I35f75f5d26e252ed2d4e3724a1d8a67690259fd7

When a process is cached, broadcasts to the process could be
delayed but we do not want to block delivery to other receivers
that are in a lower priority tranche than the deferred receivers.
We already take care of this, but only if all receivers in a
process are deferred until active. Since it's possible we defer
delivery by a few minutes (not defer until active) to cached
processes, we should skip beyond receivers in the cached process.

Bug: 298300047
Bug: 305877431
Test: atest services/tests/mockingservicestests/src/com/android/server/am/BroadcastQueueTest.java
Test: atest services/tests/mockingservicestests/src/com/android/server/am/BroadcastQueueModernImplTest.java
Test: atest CtsBroadcastTestCases
Change-Id: Iaf310716433824d2136b8385f7642ffdd49431eb
Merged-In: Iaf310716433824d2136b8385f7642ffdd49431eb

Fix areBundleEqual in Call.

Currently the sentence in areBundleEqual arranged in a wrong order.
Fixed this by moving the check sentence before the actual query
sentence.

Bug: b/312604549
Test: cts call tests
Change-Id: Id8caa5719b4df724cbe03edd75b99a2b2e3882b5

Change-Id: I62eb0570ea17e087996687b991598bc2359ec95c

This is needed because there is potentially a race condition between when NativeTombstoneManager tries to delete leftover .tmp files and when those files are added to dropbox.

Direct cherrypick from internal master

Test: atest ErrorsTest
Bug: 284900963
Change-Id: I07936c51646915cb8760f45c4bb920f5fb335025
Merged-In: I07936c51646915cb8760f45c4bb920f5fb335025

* changes:
  Revert "Move the ParcelFileDescritor initializations in both BootReceiver and"
  Move the ParcelFileDescritor initializations in both BootReceiver and NativeTombstoneManager into the try-catch block so that they would get closed automatically.

Cerry-picked from internal master.

This reverts commit 95b28e90.

Reason for revert: NativeTombstoneManager change breaks atest CtsGwpAsanTestCases:android.cts.gwp_asan.EnabledTest#testAppExitInfo

Test: n/a reverts breakage

Change-Id: I5dd690b939a977a9a873b2cae9d2b1c6739a1601
Merged-In: I5dd690b939a977a9a873b2cae9d2b1c6739a1601

NativeTombstoneManager into the try-catch block so
that they would get closed automatically.

Test: atest ErrorsTest
Bug: 269182937
Change-Id: I659f927ad65b73c81e7881c9a09a6e705e47b98e
Merged-In: I659f927ad65b73c81e7881c9a09a6e705e47b98e

In the legacy queue, it is posssible that the broadcast
timeout is elapsed before the onApplicationAttachedLocked()
callback is received, which will result in the broadcast
receiver getting discarded.

Bug: 289108400
Test: atest services/tests/mockingservicestests/src/com/android/server/am/BroadcastQueueTest.java
(cherry picked from https://googleplex-android-review.googlesource.com/q/commit:dcd5c460abe8aaa1e89d5204f3d34343becc8e30)
Merged-In: Ic3456f0711d09a5d1e5549a326ad056bde939546
Change-Id: Ic3456f0711d09a5d1e5549a326ad056bde939546

The special casing for core uids was added to ensure that we are not
adding the normal broadcast delay to these uids but if a process
belonging to a core uid is in a freezable state, we should defer all
the deferrable broadcasts.

Bug: 289397383
Test: atest services/tests/mockingservicestests/src/com/android/server/am/BroadcastQueueTest.java
Test: atest services/tests/mockingservicestests/src/com/android/server/am/BroadcastQueueModernImplTest.java
(cherry picked from https://googleplex-android-review.googlesource.com/q/commit:9a7407bc013a8a468d554813c400f640f6a2ad5c)
Merged-In: I47ce6ea68e2716a320198c008752cfaa7dedbb81
Change-Id: I47ce6ea68e2716a320198c008752cfaa7dedbb81

Bug: 304478692
Test: TH
Change-Id: I7e190b7d14e5818fb68af1753bf149366d01e61d

Bug: N/A
Test: N/A
Change-Id: Ibfa2efa85f8b69b5210b3fe5b068f7c224dbaf6b

onUidStateChanged callbacks are invoked on android.ui thread, so
to avoid blocking it for too long, offload the work to handle
the uid state changes to a different thread.

Bug: 289157124
Test: atest services/tests/mockingservicestests/src/com/android/server/am/BroadcastQueueTest.java
Test: atest services/tests/mockingservicestests/src/com/android/server/am/BroadcastQueueModernImplTest.java
(cherry picked from https://googleplex-android-review.googlesource.com/q/commit:c41bfb089c5357d956a0d8e06a4358037ee811a7)
Merged-In: I005ee6a7486ec1cc6dfe0c7ad4b488309dbda270
Change-Id: I005ee6a7486ec1cc6dfe0c7ad4b488309dbda270

We would be moving these files to a new module.
Separating them to a new filegroup, which would be later moved to the
new module behind guarded by build flag when ready.

Bug: b/289203818
Test: m
Change-Id: I5275cf6d416fb74384eb2f0a66d3bcba50dd3fc4