The multi-package APKs feature was not validated and is now disabled.
Also the ephemeral work was postponed which requires disabling of the
ephemeral cookie APIs.
bug:28514747
Change-Id: Iab7f11c503a76bcc414408dfebdf200843db814e
cherrypicking to mnc-docs because it's also relevant to the live site as per...
bug: 28431831
We must translate the new page to properly resolve this
Change-Id: Id81461f30f638734470ab9880509e76e4df68ce1
(cherry picked from commit 251a0650a8)
- Previously we did not properly cancel the animations when the user
interacted with the stack while a task was being dismissed. Now, on
touch down, we only cancel the non-dismissing tasks, and in doing so,
set their task override if there is a dismissing task (so that the
stack can scroll correctly), and update the surrounding task views
if the dismiss update comes from the currently dismissing task.
(or in other words, the surrounding task animations will only be a
function of the current task being dismissed)
Bug: 28448670
Change-Id: If7b8ed3f1fe342114ae0f6dd9dbff0d48204f31e
The engine itself knows about it, but that's at one remove from the
code that needs to consume it. Make sure it gets passed up the chain.
Bug 28346706
Change-Id: Ib94c9fbc512d92039bb7db5cd6b0b088a4a66027
AppWindowToken can contain windows from multiple clients (Processes)
If one of the client dies we shouldn't remove all windows in the app
token. We should only remove dea windows.
Bug: 28467642
Change-Id: I8be6a98e0acc79719158567114f4902066069c1b
On Devsite, the redirect parser doesn't understand regular
expressions, which breaks ~ 10% of our existing redirects; fixing
them now. Also fixing a few other redirect-related bugs while I've got
the hood open.
See first comment for doc stage location.
bug: 28447522
bug: 28467616
Change-Id: I5f315fd54cb9353a56674b5aac4dcd53875263c9
- This ensures that we perform the same dismiss animation when
dismissing via the button or via swipe.
Bug: 28443410
Change-Id: I309c38253e9c2dc78a5882dc663eec84a11619dd
Force a setAppVisibility if the previous app was translucent, so that
the resumed app is added to the opening app list. Otherwise opening
app list is empty and the transition can't be set up correctly.
bug: 28463843
Change-Id: I7f8e28d0f804c799a429743df6fc501c06d62f21
Bug: 28519669
This trace tag makes shadows look way more expensive
than they actually are, particularly troublesome
when tracing things with large number of shadows.
Change-Id: Ib6558b1388edd4b006ec15127470cb9ab563f954
There are two parts to VD's native allocation:
1) VD's internal data structure (i.e. groups, paths, etc that make
up of the VD tree). This structure can change, when a VD is used
to load a different drawable resource.
2) Two bitmap caches, not both of which will necessarily be allocated
The size of the bitmap cache depends on canvas matrix and drawable
bounds, and therefore can often change.
We need to count the native allocation from the above against Java heap.
Bug: 26269056
Change-Id: If833aedcf7f3efe00ea73a41ddccb1b48066ffd8
Registering with the maximum potential memory made by visible by an
ImageReader/Writer can cause the VM to try to heavily garbage-collect
the rest of the application. This can have significant impact on camera
applications, since they often have large ImageReader/Writer queues.
Bug: 28454727
Change-Id: I2ba43635f93da66655be024165a15631b3b421d3
- When adding a new user, adds its default whitelist apps.
- When removing an user, clears its revoked default whitelist apps.
BUG: 28451218
Change-Id: Ic30a3d6d30c7575620edd3ad32c3221203626f8d
On non-FBE devices, we offer to migrate CE/DE storage of certain
system apps during boot. However, if this migration happens after
app code has started to run, we end up clobbering open files such
as databases.
The right way to solve this is to apply all migration early during
boot (in the PMS constructor) before any app code can start.
Bug: 28512897, 28525109
Change-Id: I17eb9a09e5862d873b86c05ce35874a34386e433
