Introduce a new concept of "B" services. All running services are
classified as either A or B. B services are later in the LRU list.
Their oom_adj is after the home app. This allows us to better pick
services to kill based on how long they have running, and should
reduce the amount that we end up killing the home app.
This temporarly turns on a debug log when the oom_adj of a process
is changed. Sorry, I know it is noisy. This is needed to try to
track down why some processes are being killed.
Also add a flag to the SyncManager's service binding to allow the
syncing process to be more aggressively killed if it has done UI.
This is to address cases we have seen where sync is causing an 80MB
gmail process to be kept around, preventing other process from running.
Now what will happen is that the syncing process will aggressively be
killed by the system, and can then be restarted in a much lighter-weight
state.
Do a little tweak in the power manager to allow us to still do smooth
brightness changes even when the fancy TV off animation is in use.
And get rid of a debug log in the window manager that was accidentally
left in.
Change-Id: I64a8eeaaa1f096bab29c665fbff804c7f1d029e2
The basic problem was that at some points during setup wizard, this would
happen:
1. The app's process is killed.
2. The app's process is restarted, but not to actually resume the
setup wizard activity, just to put it in the stopped state.
When doing this, the saved state is cleared but the app will
never provide a new one.
3. The app's process is killed again. At this point, because the
saved state is cleared, the activity is completely removed.
4. Eventually the entire activity stack becomes empty, and a new
setup wizard activity needs to be created as the home app.
There is a combination of bad stuff going on here.
First, why is the process being killed? At this point the setup
wizard is the home app, so it shouldn't be killed. There were two
reasons why this was happening:
- CryptKeeper still was not completely cleanly going away. To fix
this, I removed the check in the activity manager to not allow
an activity to finish if it is the only activity on the stack and
maybe-kindof looks like the home app. This really wasn't necessary
(we always take care of starting a new home activity if we find the
stack is empty), and outright dangerous with all of these things
purporting to be home but not.
- There was an issue in computing the oom_adj where the home app
would not be marked as "not hidden", and if we had to re-compute
its oom adj in the current sequence would then give it an
adjustment as a background process... and with all the processes
we spin through during boot, it quickly got down to background
#16 and killed.
Second, what is going on with the state? This is easier, the code
in the activity manager to create a new activity but put it in the
stopped state was still clearing the saved state. The saved state
should only be cleared when going in to the resumed state. When
going in to the stopped state, we can just keep holding the same
saved state.
Change-Id: I7d21cdcfa082d98ca70c79d9923e29605ee4353e
...dev.bootcomplete flags is set before boot animation is out
Also:
- Fix crash in recent apps if the intent for an old app didn't
happen to have the new task flag set.
- Fix issue where a crash in system UI would cause the crash
dialog to be displayed below it, effectively locking the UI. Now
the crash dialog for persistent processes is shown above everything
else.
Change-Id: I0312001a92beeae5f644c7c3e5c5e19f6716df36
...the "Complete action using" dialog
When an application goes idle, it sends back to the activity manager
the configuration it last used, to make sure the two don't get out
of sync. Fix a bunch of edge cases here in dealing with that, and
be sure to also send the current configuration when launching an
activity so the client is always up-to-date when launching.
Also a small fix to not show the upgrading dialog during first boot.
Change-Id: I14ed366a87cd689d1c78787369e052422290ac6f
Replace TrafficStats calls by reading values from xt_qtaguid kernel
module. To keep BatteryStatsImpl changes lightweight, cache recently
parsed stats. Tracks mobile ifaces from ConnectivityService.
Refactor xt_qtaguid parsing into factory outside of NMS. Add stats
grouping based on UID, and total based on limiting filters like iface
prefix and UID.
Bug: 4902271
Change-Id: I533f116c434b77f93355bf95b839e7478528505b
Some services have info that is not readily available any other way. Add
a permission check to those services.
Bug: 5389201
Change-Id: I5a9724ec89e1c11ef7813eeb52a6a05b3ea92af5
...Should Skip Unsecure Lockscreen (ICS)
Also while I am in there, clean up logging of intent objects to include
even less sensitive information, while showing the true Intent in dump
output (since apps can't get to that).
Change-Id: I35fed714645b21e4304ba38a11ebb9c4c963538e
If the app had activities still finishing, when we checked whether it was
now stopped we would get told no. Also some other improvements:
- Schedule an idle as part of the force stop, to get any finishing
activities out of the stack soon rather than waiting for some activity
to idle.
- Don't filter out stopped system apps. This is dangerous because
system apps may have no way for the user to explicitly launch them,
so they could get put into a stopped state for which there is no way
to get them out. Also if the user really wants a system app to not
run, the new disabling mechanism is more appropriate.
Change-Id: I34003f21dac29e2ca0f66a23b88c710de41bab99
Provider was not being removed from the class map because it was using
the wrong key. D'oh.
Also a little cleanup.
Change-Id: I318e8b1a265318ac1474e0a7f14f27f89f357505
When force stopping an app, when removing its activities also finish any
activities from other apps above it in the task. This avoids some situations
where the task gets into a bad state where its root becomes a different app.
Change-Id: I79e5cd520ae321bec80adefd2ccc2b0370ace372
The resolver activity was hiding the following activity from recents.
Also some other fixes: a little better memory use debugging, removed
some unneeded code from window manager, moved some system activities
into their own process, added some more running process information for
manage apps.
Change-Id: I66687d16989ff965d524b92dc360f37c19199717
Persistent process can no longer use hardware acclerated drawing
when running on a low-memory device.
Change-Id: I3110335617af1c98fcede9bf41f4a1d0c20d0e87
Change-Id: Iff2cfec5280a314989d915aa830c16124f921611
5214105: taking a screenshot while "Android is upgrading..." crashes device
5109947: Race condition between retrieving a content provider and updating its oom adj
...(when turning display on after recently turning it off)
Also clean up when we decide to turn the screen on to improve that
transition. There are still problems here with turning it on
before the wallpaper gets dispayed.
Change-Id: I2bc56c12e5ad75a1ce5a0546f43a845bf0823e66
This introduces a new facility for code during the boot process
to display messages to the user through a progress dialog. This
is only for use when performing longer-than-usual post-upgrade
operations such as running dexopt on applications or upgrading
databases.
Change-Id: I0e78439ccec3850fb67872c22f235bf12a158dae
Now the dumpsys battery output will show up in bugreports again.
Change-Id: Id36e87d27e9d3c06dcc17671c81aa1d3fe260d1e
Signed-off-by: Mike Lockwood <lockwood@android.com>
- When shutting down, if the screen goes to sleep there is code
that tries to do a notifyAll without holding the lock:
java.lang.IllegalMonitorStateException: object not locked by thread before notifyAll()
at java.lang.Object.notifyAll(Native Method)
at com.android.server.am.ActivityStack.checkReadyForSleepLocked(ActivityStack.java:776)
at com.android.server.am.ActivityStack$1.handleMessage(ActivityStack.java:282)
at android.os.Handler.dispatchMessage(Handler.java:99)
at android.os.Looper.loop(Looper.java:137)
at com.android.server.ServerThread.run(SystemServer.java:603)
- If an invalid Uri object is sent to the system process it can crash because
the Uri class throws an assertion while unmarshalling. Change this to an
IllegalArgumentException so it gets sent back to the caller:
java.lang.AssertionError
at android.net.Uri$PathPart.readFrom(Uri.java:2224)
at android.net.Uri$HierarchicalUri.readFrom(Uri.java:1106)
at android.net.Uri$1.createFromParcel(Uri.java:1689)
at android.net.Uri$1.createFromParcel(Uri.java:1681)
at android.content.IContentService$Stub.onTransact(IContentService.java:53)
at android.content.ContentService.onTransact(ContentService.java:120)
at android.os.Binder.execTransact(Binder.java:338)
at dalvik.system.NativeStart.run(Native Method)
- StrictMode can try to access the first index in the stack crawl of a stack crawl
array of length 0. Not sure why this happens, but make the code more robust:
java.lang.ArrayIndexOutOfBoundsException: length=0; index=0
at android.app.ApplicationErrorReport$CrashInfo.<init>(ApplicationErrorReport.java:341)
at android.os.StrictMode$ViolationInfo.<init>(StrictMode.java:1978)
at android.os.StrictMode$AndroidBlockGuardPolicy.startHandlingViolationException(StrictMode.java:1097)
at android.os.StrictMode$AndroidBlockGuardPolicy.onReadFromDisk(StrictMode.java:1068)
at libcore.io.BlockGuardOs.read(BlockGuardOs.java:137)
at libcore.io.IoBridge.read(IoBridge.java:426)
at java.io.FileInputStream.read(FileInputStream.java:179)
at java.io.InputStream.read(InputStream.java:148)
at com.android.internal.os.ProcessStats.readFile(ProcessStats.java:804)
at com.android.internal.os.ProcessStats.getCpuSpeedTimes(ProcessStats.java:564)
at com.android.internal.os.ProcessStats.getLastCpuSpeedTimes(ProcessStats.java:545)
at com.android.server.am.ActivityManagerService.updateCpuStatsNow(ActivityManagerService.java:1470)
at com.android.server.am.ActivityManagerService.batteryNeedsCpuUpdate(ActivityManagerService.java:1522)
at com.android.internal.os.BatteryStatsImpl$MyHandler.handleMessage(BatteryStatsImpl.java:110)
at android.os.Handler.dispatchMessage(Handler.java:99)
at android.os.Looper.loop(Looper.java:137)
at com.android.server.am.ActivityManagerService$AThread.run(ActivityManagerService.java:1302)
(Also fix this code to not cause strict mode to trigger at all, because there is
no need, because this is just reading stuff from /proc.)
- The system seems to crash during boot if it thinks it needs to rotate
the screen, when it is trying to take the freeze snapshot way too early.
There is no need to freeze the screen during boot or if the screen is off:
java.lang.NullPointerException
at android.view.Surface.init(Native Method)
at android.view.Surface.<init>(Surface.java:256)
at com.android.server.wm.ScreenRotationAnimation.<init>(ScreenRotationAnimation.java:91)
at com.android.server.wm.WindowManagerService.startFreezingDisplayLocked(WindowManagerService.java:8758)
at com.android.server.wm.WindowManagerService.startAppFreezingScreenLocked(WindowManagerService.java:3971)
at com.android.server.wm.WindowManagerService.startAppFreezingScreen(WindowManagerService.java:4003)
at com.android.server.am.ActivityRecord.startFreezingScreenLocked(ActivityRecord.java:515)
at com.android.server.am.ActivityStack.ensureActivityConfigurationLocked(ActivityStack.java:3997)
at com.android.server.am.ActivityManagerService.updateConfigurationLocked(ActivityManagerService.java:12535)
at com.android.server.am.ActivityManagerService.updateConfiguration(ActivityManagerService.java:12439)
at com.android.server.wm.WindowManagerService.systemReady(WindowManagerService.java:6161)
at com.android.server.ServerThread.run(SystemServer.java:521)
Change-Id: I85062bb5f6b0909a0f52feedaa75e7611d9d7fbd
In UsageStatsService, separate last-resume times from the rest of
the statistics, and serialize them to an XML file daily.
This way, ApplicationsProvider will still be able to acces this data,
even thoguh other statistics are flushed to disk and discarded each day.
Bug: 5108745
Change-Id: Id3df3c98243ba02cde16b31e5e29bd9ff3602108
The activity manager now take care of plugging the correct settings
into the OOM killer in the kernel. This is a lot cleaner because
it is really central to how the activity manager works, and nobody
else cares about them.
Taking advantage of this, the activity manager computes what it
thinks are appropriate OOM levels based on the RAM and display
size of the device.
Also a small optization to the package manager to keep a binding
to the package install helper for a bit after done using it, to
avoid thrashing on it.
And some new APIs that are now needed by Settings.
Change-Id: I2b2d379194445d8305bde331c19bde91c8f24751
- Improve how we handle processes that have shown UI, to take care
of more cases where we want to push them into the background LRU
list.
- New trim memory level for when an application that has done UI
is no longer visible to the user.
- Add APIs to get new trim memory callback.
- Add a host of new bind flags to tweak how the system will adjust
the OOM level of the target process.
Change-Id: I23ba354112f411a9f8773a67426b4dff85fa2439
...apk reinstall; affects user privacy
Disconnecting a ServiceConnection after an app is torn down could
impact the bookkeeping of the same service if it has been started
for the app.
Also address issue #5073927: GSF process can't be killed
A new flag allows the systems location manager service to tell
the activity manager to not pull bound services up forever into
the visible adj level.
Change-Id: I2557eca0e4bd48f3b10007c40ec878e769fd96a8
A later CL will introduce an API for querying whether a given package
runs in a persistent process; UIs such as Settings will be able to use
that to determine whether to disable the 'force stop' action.
Change-Id: Iab47c2300fdce285da7d83e02263c9a5f69edd70
Also improve how we handle services, keeping track of whether they showed
UI and if so putting them immediately on the LRU list.
Change-Id: I816834668722fc67071863acdb4a7f427a982a08
To profile the looper, run the following command:
adb shell am profile looper start <process> <file>
adb shell am profile looper stop <process>
Change-Id: I781f156e473d7bdbb6d13aaffeeaae88bc01a69f
This should fix a leak of process death recipients in the activity manager.
Also add debugging of content observers to try to track down what looks
like a leak of them in the content service.
Change-Id: Id6823679493ef0cde5307bb66490ebe31b878556
We now collect more detailed information splitting the maps into
additional useful categories.
Fixed some bugs in account, such as not correctly handling all of
the current dalvik allocations.
The activity manager now prints a final summary of all pss organized
by the apps and the categories.
Change-Id: Iafc5f27c998095812b1483c6803b8e0f0587aeae
Now classify background processes into a set of bins of how much
memory they should try to clear. The last bin also involves
destroying all activities in that process.
Removed the old code for the simulator that is no longer needed
(yay). The debugging features it had are now integrated into the
regular oom adj code.
Small fixes to load average service.
Change-Id: Ic8df401714b188c73b50dbc8f8e6345b58f1f3a0
