- It appears that there is contention between startActivityAsUser() and
removeTask() (called on two separate threads) which can cause jank when
a user removes all the tasks from their recents list. This CL ensures
that startActivityAsUser() is always run first so it is not blocked
by the other call (which should be able to run in the background
uninterrupted).
Bug: 22760556
Change-Id: I7564a2f0e43414686419d3657379bbd0ca6b4152
BUG: 23086704
Cherry-picked from https://android-review.googlesource.com/#/c/162280/
When the screen goes off or dreaming start, an alarm will be
scheduled and idle state will be true when the alarm expired.
If the screen goes on or dreaming stop happens before
the alarm expired, the alarm isn't cancelled and idle state is
set to be true when the device is in SCREEN_ON or DREADING_STOPPED
state. There is also a case that Idle alarm triggered when
the screen on or dreaming stop just start to be processed.
ACTION_TRIGGER_IDLE will set mIdle to true during screen on
or dreaming stop.
In this patch, the alarm will be cancelled when the screen goes
on or dreaming stop and screen-on flag will be set. So the idle
state can only be set when screen is off or dreaming started.
Change-Id: Ic21a2394418ca55513ab932b3bfad1126b8769c1
This shouldn't happen, since there are many places where
invalid icons should already have been either fixed (in the
case where there's an .icon but no .mSmallIcon) or rejected
(if they're both null or invalid). But if a notification
makes it all the way to SystemUI without a valid icon, let's
crash the sender.
Bug: 23011305
Change-Id: Ifaebec57d59baa1defb4520178b5815d47ed5712
This tightens the guarantee that detached stack won't be used. We also
add logging to detecting a situation where a stack not belonging to a
display is being moved on that display.
Bug: 22191609
Change-Id: Ia674bb5960018104a56c5138775ab5216906675b
We weren't closing the ZipFiles created in WebViewFactory to check
inside APKs - use try-with-resources to get them closed automatically.
Bug: 23072621
Change-Id: I11c6b77e960a7d240d19d22240cac177b6ba27b2
Initialize current network name to correct value from the
SubscriptionInfo until we get a broadcast about its current state.
Bug: 22212693
Change-Id: I17fa4378cc7a540c81268f8c4d5aa6a505f3ee40
The main looper needs to run freely for a moment after disabling
wifi in order for various signals (content observers, broadcast) to
propagate to all the listeners that need to take action for the
wifi stack to shut all the way down. This patch breaks up the
disable-and-rewrite-config sequence of wifi AP restore in to two
distinct operations separated by a moment so as not to block those
necessary messages.
Bug 22979342
Change-Id: I271766cad0e454669a194652fb67f835bb022cd1
...to persist after reboot
We were writing a corrupt settings file, so would always reset back
to the default app ops state after boot...!
Also add new appops service commands to manually write and read
its settings, since that is very useful for testing.
Change-Id: Ia510507764738fd82e45ec0be6db840c6ea30c28
We now have a new whitelist you can put apps in, which
opts them out of the old battery saver mode and new app idle,
but doesn't keep them from going in to doze. This is for a few
special cases that we had previously whitelisted for battery saver,
and inherited to the new modes... ultimately we should figure out
how to get these apps out of the whitelist completely, but this
will help for now.
Apps in this new whitelist are not shown in the UI, because they
are still significantly restricted by not being able to operate
normally in doze. This also means they are still visible in the
list of all apps for the user to be able to put them on/off the
complete whitelist if that is what they really want.
In the course of doing this, I needed to clean up code in the
network policy manager to better separate management of the
two firewall rules that now have different whitelists applied
to them. This also hopefully just generally simplifies and cleans
up that code. Hopefully!
Change-Id: I92e15f2f85899571dd8b049b5e3eb1354f55f353
Merge the CHANGE_NETWORK_STATE permission with WRITE_SETTINGS.
AndroidManifest.xml:
Raised the protection level of CHANGE_NETWORK_STATE permission from
normal to signature|appops and pre23|preinstall for compatibility
provider/Settings:
Wrote new helper methods to check if app is allowed to change network
state.
ConnectivityManager.java & ConnectivityService.java:
Replace enforcement checks for CHANGE_NETWORK_STATE with
checkAndNoteChangeNetworkStateOperations instead.
Change-Id: If8c2dd3c76a5324ca43f1d90fa17973216c2bcc5
When creating a work profile, system apps are uninstalled and then
sometimes reinstalled.
In the process, they lose their intent verification status.
BUG:22943461
Change-Id:I5b008c6de2125f190063b08908076a649067c60d
Internal volume was not available during PackageManagerService creation,
which resulted in a zombie user's folder not being cleaned after a reboot.
Add the internal volume earlier in the boot cycle so that it can be accessed
for user cleanup.
Bug: 22483086
Change-Id: I8f3ffbb25f3902d00a96d1ee2d7a79373c5e35b7
We've seen case of it taking longer than 1500ms for the wifi system to
actually shut down after the triggering settings element is written,
so extend the wait time a bit. We've seen it take more than 1500ms
but not more than 2500ms, so that's the new heuristic.
This will of course all become happily obsolete once we start
applying restored AP definitions programmatically rather than by
filesystem-level operations.
Bug 22979342
Change-Id: I6acf1baac23d4100124093128b82abf242b11a0e
We have been seeing crashes on the Theme editor related to an assertion
on getDelegate when the object does not exist
(http://b.android.com/181951).
When debugging, the crash was happening without seeing a previous call
to removeJavaReference. This seems to completely remove that crash.
Since the DelegateManager can be called at least from two threads
(main thread and the finalizer), it should be thread safe.
Bug: http://b.android.com/181951
Change-Id: I1b28f863ff198f8592f170a98f9de391b2ac3ea2
