After device bootup, the signal strength triangle is empty although there is
both voice and data connection. This is due to when TelephonyRegistry doing call
back listen register, some APP use mDefaultSubID. However, when the reigister
happen, the mDefaultSubId does not exist. Althouhgh it can be update later, if
the update event comes too late (especially after the steady state), no
ServiceStatechange event can be received anymore. Thus the service is always not
available and the signal stength trangle has no chance to be updated anymore.
This is a risk condition.
Bug:17472622
Change-Id: I772d29e6dd9da212b78fd0d4438b8273e3318dba
This turns out to happen a lot in normal usage, but we're logging
copiously about it each time. It's reporting "oh hey we're already
in this state" -- and so the log is not useful anyway.
Logspam is spam.
Bug 15169507
Change-Id: Ie2d01ff1b0b3600dd9c15ccf83d60875558f1dc2
...due to dying renderer process
Don't kill processes if they are bound to a service but not impacting
oom adjustment.
Change-Id: I1cc44e633feaeaad6e996b79a6cfd7b386c04095
Since we're reusing the cast tile for screen sharing we need to make sure there's only ever one
mirroring presentation going on at a time.
Bug: 9905068
Change-Id: I249932d0f8853880dd453eb57c2a258e6ae952b0
The spooler communicates with he remote PDF renderer asynchronously.
When print is confirmed we close the renderer, destroy it, and unbind
from its service. If we unbind from the service after the print activiy
is finished we get a crash. The bug was that we did not wait until we
disconnect from the remote renderer before finishing the print activity.
bug:17583115
Change-Id: I55b0135f9c5658b3a4fda2901b8b3bdef044e211
If RemoteViews get a null package in the constructor we are
using the context in which they are applied to obtain resources
during inflation. In such a case if we call the getPackage()
method we were getting null, i.e. the package passed in the
contructor, but now we are returning the package if the app
that created the remote views.
bug:17513823
Change-Id: I0c4a8953c8320469a9160dbaae46ce31465b313b
This allows us to ensure windows are fully drawn before unblocking
screen on while dozing.
Bug: 17516245
Change-Id: Ibe63c212b8db855ce26a34a8169f33764b266ee6
This change fixes several issues in how the display's power state
was coordinated to ensure that the display is not turned on until
the pulse is ready to be shown. It also greatly simplifies the
wakelock logic and makes it completely deterministic instead of
relying on timed delays.
Bug: 17516245
Change-Id: If2b96845a010f2d05c610970257d898f32ccacb7
The screen on blocker is used to keep the screen blank while the
system is drawing new content to prevent the user from seeing a
flash of stale content while the screen is being turned on.
This patch ensures that the screen on blocker functionality
is also applied while dozing.
Bug: 17516245
Change-Id: I77c2d0f2b99476a59ad212099f44c63aa2ef9c34
Very useful for testing persisting/restoring, to make sure
that all pending changes have been written.
Change-Id: I0e3b7cd3af8afb0b6e751e086081566ab00b76c9
Allow edgeeffect colors to be changed independently of the primary
color for the current theme.
Bug 16512225
Change-Id: Ibc13ad755f489fe1f6d16af0c61a1d12dd1a61f7
