Replaces deprecated constants with newly added constants in the actual
implementation.
Bug: 27932489
Change-Id: Id54236a05127cd7ce3bf0668c002635fb86489a9
Adds android.permission.UPDATE_LOCK_TASK_PACKAGES, available only to
the system and setup wizard, providing access to
updateLockTaskPackages() without registering with the device policy
manager as a device owner.
Bug: 20016740
Change-Id: I6eceb5cece9407c84eabf3a818bb908f95989b91
Adds PROTECTION_LEVEL_SETUP, a privileged permission for use only by
the setup wizard.
Bug: 20016740
Change-Id: Ib95e349c54d5d12465bf43162975dfb628ef2434
Previously we would be writing out the Manifest.java file as we
processed the AndroidManifest.xml. This would lead to empty
Manifest classes if there were no permissions or permission groups
defined in the AndroidManifest.xml.
This would pose problems for processes that checked for public classes
and considered them part of the API (support lib).
Now we collect the structure of the Java class in memory before deciding
if a file should be created.
Change-Id: I6b909f28d74356414c6ef5ad005180d6ea5e44ca
Bug: 27719830
To turn on warnings, apps have to add to their Activity.onCreate() method
the following code.
StrictMode.setVmPolicy(new StrictMode.VmPolicy.Builder()
.detectLeakedClosableObjects()
.penaltyLog()
.build());
For Slang generated ScriptC derived classes, we assume their
constructors won't throw exceptions after calling the ScriptC
constructor. In addition, ScriptIntrinsic derived classes do not seem
to throw exceptions in their constructors either. Therefore, we can
leave the guard.open() call in the Script constructor. This may be
only an approximation, but allows us to add CloseGuard for script
objects without making changes to slang.
Change-Id: I77ed45239a60b85af5c811dee6c124fb53da9060
(cherry picked from commit eb4dd08ec132f83745b8b28fa7da58eb4478b5b9)
Generally we return info for the latest installed package; which could
either be a built-in [i.e. on the /system partition] package or a user
updated package. In certain circumstances, we want to be able to get
the version on the /system partition regardless of whether or not the
user has updated it. We do this by passing MATCH_FACTORY_ONLY to
getPackageInfo().
Bug: 27469181
Change-Id: I8dd1d110e2d72e5c6f024812d0b5d15d8b217347
scrap
documentInfo.PageCount might be set to "unknown" == -1. Hence we don't
know where the last page to scrap ends. Hence defer this decision all
way until PdfManipulationService.removePages as then we know the actual
pages of the document.
Change-Id: I063c3cd084d65a3ac5c60c3d85cec3a346be2680
Fixed: 27948632
This reverts commit 35e2ea02035e8b7eb7de5d7a1d51bf030dcb5b76.
This patch was based on two different wrong assumptions.
Bug 27822069
Change-Id: I20b1017f91f3fce3c23dd8446459d6f3e3150a48
There's no point in mounting it, since it's living on top of a tmpfs
that will be discarded in a few moments once the user authenticates.
Bug: 27932087
Change-Id: I2df2da0136eac6cc27aa402fb9b00cfd23d562a8
We set deferRelaunchUntilPaused when a relaunch is delayed due to the app
in pausing state, but this flag is not cleared after it's relaunched upon
paused. After that every time the app is paused, it relaunches again.
Change-Id: Ib98956e0760b3c4fe108d93f5c7843e2cbd313de
related-to: b/27834014
Removing entries from a Bundle may change its has-FDs state, so now
induce a recheck after removals. In support of that, make the
recheck set the known state in all cases, not just in the has-FDs
case; and note that filterValues() inherently produces a result that
is known to have no FDs.
Bug 26676151
Change-Id: Ida07b77ad2046aeafb98e0b164cd571ce45d064f
