There were situations where the activity manager ActivityStack was
moved to the front but the corresponding window manager TaskStack
was not. This caused the wrong activity to receive focus which led
to Application Not Responding errors.
One path in particular occurred in startActivityUncheckedLocked()
where curTop.task != intentActivity.task and
sourceStack.topActivity().task != sourceRecord.task. In this case
targetStack.moveTaskToFrontLocked() was never called.
This fix forces all calls to ActivityStack.moveToFront() to make
a call to WindowManagerService.moveTaskToTop() and eliminates
redundant calls to moveTaskToTop().
Fixes bug 17721767.
Change-Id: Ibf01389810dd36724eaec5a4a07560144b2f4cef
Larger OTAs (750 MB tested) are taking 3-4 minutes
to write, due to the O_SYNC needed to ensure that
the data is actually committed to disk prior to
reboot.
Bug: 18750317
Change-Id: Idfab3ffd0276df4548f69a09c72ad6f4a344b6e6
On boot-up we restore all persistent tasks to an activity stack.
This can cause issues with the activity stack supervisor when it
tries to make restored tasks with activities visiable when they
shouldn't be. Which ends up messing up the order of the recents
list. Now we don't restore persistent tasks to an activity stack
on boot-up. Instead we add the task to the stack when it is needed.
Also, fixes issue with not been able to launch task records with
activity records that were restored from an other device.
Bug: 18692762
Change-Id: Iad0e6635f8c5d1dab4d341feb3e7b06291a94739
This CL passes a port ID when enabling/disabling ARC in case
there are multiple HDMI ports that support the feature.
Bug: 18781204
Change-Id: I632518132bf07c8ae6f0ff5135429ca719b596b2
The man bent over his hourglass,
A programmer of sorts. The day was green.
They said, "You have a blue hourglass,
You do not fire alarms when they're asked."
The man replied, "Alarms as they're asked
are changed within the blue hourglass."
And they said then, "But fire, you must
Alarms beyond us, yet themselves,
Alarms within the blue hourglass
That trigger exactly when they're asked."
---
Fix the delivery-fuzzing semantics that had been introduced in
81f9882b5aadd6a2289c9f521a06a7af5f35ebf0. That patch turned out
to be incomplete; in particular, alarms scheduled later might require
the validity of an already-scheduled kernel alarm even if they did
not affect the head alarm batch directly, and this was not being
addressed. For now, roll back the fuzzed delivery logic entirely.
(This is not a full revert because that patch also caused exact alarms
to be considered standalone for batching purposes, and we need to
preserve that new policy.)
Bug 18726690
Bug 18765436
This is a 'git revert' of 81f9882b5aadd6a2289c9f521a06a7af5f35ebf0
*except* that this CL preserves the "exact alarms are treated as
standalone" portion of the original patch.
(Cherrypick of 5c3e277fb42bd799287936c5aee0d30fbcc7e65c from its
original branch.)
Change-Id: Ib9c3200f7e6bc6fa0c9928ee9db4cfd87f039353
Now that the delay between connectivity changes and CONNECTIVITY_ACTION
has been removed (ag/599650) races between CONNECTIVITY_ACTION and
the setting of the default network become more evident.
In http://crbug.com/441818 Chrome is calling getaddrinfo()
immediately after a device goes from no connectivity to cellular
connectivity, and Chrome is erroneously getting back EAI_NODATA
because netd hasn't yet set the default network for DNS resolutions.
bug:18757162
Change-Id: Ib607dcb3697403272a8c838713a9cb602e9c6820
