Similar to I3876c41e6d0e41d044a5b1d5e57f894c7fb4fb0e where we pause
the watchdog because dexopting can take a while, here moving A/B
artificats can take a long time.
Bug: 134062700
Test: Manually tested by adding an artificial sleep in
OtaDexoptService#main and verified that watchdog was not triggered
Change-Id: I69067fbbb8be4e421918c18b67269044fab51b3e
The previous model could have impact on boot time to read/write from/to
disk, and could potentially fail in some scenarios where the device does
full reboots instead of framework restarts.
The current design most simply avoids crashing in the first 30mins after
a full reboot, and optimistically checks the wall clock to rate-limit
the crashes to every 6h.
Test: manual as below, without IS_DEBUGGABLE condition
Test: Install new NetworkStack, force crash, observe rollback
Test: Set min_uptime_before_crash to 100, force crash, observe crash
Test: min_uptime_before_crash still 100, install new NetworkStack, force
crash: observe rollback (there was already a recent crash)
Test: Set min_crash_interval to 10, force crash: observe crash
Bug: 133725814
Change-Id: I3fd5ba7047d7ac991cb62a7cab16a40f4ee731a3
When the network stack crashes, the system will rebind to it.
Existing references are no longer useful (they just throw
RemoteException) but if the system is still up, then the user
can at least recover the situation by taking actions such as
going into airplane mode, toggling wifi, etc.
This CL stops ConnectivityService from crashing the system when
it cannot talk to NetworkMonitor. This is arguably better than
crashing the system, because crashing the system is disruptive
and carries the serious risk of a bootloop from which it is not
possible to recover.
NetworkStackClient already contains code to crash the system
when the network stack crashes. This change help ensure that
if a crash occurs, it is the result of an explicit decision by
that code instead of an unchecked exception in one of the callers
of the network stack.
Bug: 133725814
Test: builds, boots
Test: atest FrameworksNetTests NetworkStackTests
Change-Id: Ib9a15fececd8579fc5b139fe0341275a45512e0f
Currently the backup of user data is done in the enable rollback stage,
during which there is no guarantee that the package being backed up is
not currently running. Moving the backup to the post install stage will
guarantee that the package is not running.
Test: atest RollbackTest
Test: atest StagedRollbackTest
Bug: 124032231
Change-Id: I4b42a0c5ade1645585d1d6f698637df950d05c72
am: f3eaec257a -s ours
am skip reason: change_id Ide2428dab3fcd6d7cd00aa2a9fd99d6c99b815a4 with SHA1 d5603307e1 is in history
Change-Id: I08e37b4458b3c6242a1cca8aab913b09642a9e53
whenever clear is used as a blend mode to draw bitmaps
Modified SkiaRecordingCanvas to call filterBitmap instead of
filterPaint in order to conduct the necessary compatibility checks
to use dest_out instead of clear whenever a bitmap is drawn
Test: In progress..
Bug: 131689368
Change-Id: I56d015e773b312cac948d850ca2997d8df0cc66d
Merged-In: I56d015e773b312cac948d850ca2997d8df0cc66d
Refactors yesterday caused tests to fail because of dependencies getting
initialized earlier. Refactored so that they are initialized late enough
for the tests to be able to mock them correctly.
Also fixes bug where handles would still show when there was no
Assistant.
Test: atest AssistHandleBehaviorControllerTest
BUG:132424950
Fixes: 134086503
Change-Id: I645a2529bf56b5e996e8250267ead87bed9f4f2b
FIX:132424950
