The two major changes here are:
- Move lingering out of NetworkMonitor. The fact that lingering
is currently its own state in NetworkMonitor complicates the
logic there: while a network is lingering it cannot be in any
other state, we have to take care not to leave LingeringState
for the wrong reason, etc.
- Instead of keeping a single per-network boolean to indicate
whether a network is lingered or not, keep a linger timer for
every request. This allows us to fix various corner-case bugs
in lingering.
The changes in behaviour compared to the current code can be seen
in the unit test changes. Specifically:
1. Bug fix: when a network is lingered, and a request is added
and removed to it, the existing code tears the network down
immediately. The new code just sends another CALLBACK_LOSING
and resumes lingering with the original timeout.
2. Bug fix: if cell is unvalidated and wifi comes up and
validates before cell does (as might happen on boot), the
existing code immediately tears down cell. The new code
lingers cell, which is correct because unvalidated cell was
the default network, so an app might have been using it.
3. Correctness improvement: always send CALLBACK_AVAILABLE for
the new network before sending CALLBACK_LOSING. This was not
really an issue in practice, because the usual flow is:
- Network A is the default.
- Network B connects, CALLBACK_AVAILABLE.
- Network B validates, CALLBACK_LOSING.
Bug: 23113288
Change-Id: I2f1e779ff6eb869e62921a95aa9d356f380cf30a
As explained in the TODO, the loop serves no purpose since only
one network can be satisfying a given request at a time.
Instead of looping, look up the nai in the mNetworkForRequestId
array that exists for this purpose.
Keep the loop around with an Slog.wtf statement on it so we can
see if we ever hit it, and add a TODO to delete it if we don't.
Bug: 23113288
Change-Id: I173de4bd45c5a4169b7a062a981f2ecccaa44143
http://ag/1194313 broke unregisterNetworkCallback because the
system does not parcel the type of the request back to the app.
So when the app calls unregisterNetworkCallback, the
NetworkRequest that's passed in does not have a type and thus
doesn't match the request in mNetworkRequests.
Fix this by parceling over the type as well.
This was not caught by the unit test because the unit test all
runs in the same process with no parceling.
Bug: 23113288
Change-Id: I58b2ed651b9bf5cbdcca5b25c3ca24db53cffdf1
Ensure every public method is annotated with why it's public.
This can be either an @Override or @VisibleForTesting annotation
or a comment explaining why it's public.
Bug: 29927488
Change-Id: I3582aef7997dc0d723718ca5e3dd115647d22979
This allows us to keep track of how many live requests a network
is satisfying without having to count them every time.
Bug: 23113288
Change-Id: Ic4756676491e09071dbf80b7c48da3be028d68eb
This will allow us to simplify code that deals with
NetworkRequests outside ConnectivityService.
Bug: 23113288
Change-Id: I9b3a859d0c68cad73d7f6baa4b584d13ffd2ae36
1. Support multiple callbacks in TestNetworkCallback. This is
necessary to test situations where multiple callbacks are
generated by the same event (e.g., CALLBACK_LOSING on cell
with CALLBACK_AVAILABLE on wifi when wifi connects), which is
necessary to test callback order. So far this has not been
covered because all callback testing was using per-network
callbacks.
2. Add a benchmark test for registering NetworkRequests and for
sending onAvailable and onLosing callbacks.
Bug: 23113288
Change-Id: Ib69373358ad766ab1bd989e864a5a51ef762c73c
This patch removes static methods for logging IP connectivity events
defined in android.net.metrics and replaces them with a single log()
instance method defined on IpConnectivityLog. Event constructors are
now public also. Every classes logging such events now create an
instance of IpConnectivityLog for logging event objects directly
instantiated with new.
Removing static dependencies allow straightforward testing of logging.
This patch also removes the base IpConnectivityEvent class which is not
needed any more.
Bug: 29035129
Change-Id: I3de700f93f46deaa48a759f938f7d00e1d8bff98
This patch adds synchronization inside LegacyTypeTracker so that
getNetworkForType() can safely run concurrently with remove().
Without synchronization if remove() removes the last network for a
given type while getNetworkForType() runs for the same type, it is
possible that getNetworkForType tries to access the head of an empty
list, resulting in a runtime exception.
This issue was found by zoran.jovanovic@sonymobile.com who proposed a
fix in AOSP (Change-Id: Ia963662edb9d643790e8d9439e4dbdcac4c2187b).
This patch differs from the fix proposed by the bug reporter and tries
instead to do the minimum amount of locking to make getNetworkForType
safe.
Bug: 29030387
Change-Id: I915aac527fc8828b32bf35fee870add2dfb11d8d
It's with the rest of the logic now and allows checking whether the
lockdown state matches, too, which led to a lot of misunderstandings.
Fix: 29199431
Change-Id: I94a2c38c4837f9c33b5b9c2becb52eeb7e2a2534
All users should be made aware a captive portal is in place and be
given the opportunity to sign into the network. Without this fix
other users are not notified and given a chance to sign-in.
Change-Id: I1bf823d5f6a36f391dca4be5f6a584e8562a72a7
Fixes: 23079964
* changes:
Give WakeupMessage the ability to transport an object as well.
Don't treat the lingerExpired broadcast specially.
Add a test for mobile data always on.
Add a FakeSettingsProvider and use it in ConnectivityServiceTest.
NetworkMonitor no longer uses the broadcast for lingering, it
uses WakeupMessage instead.
Bug: 23113288
Change-Id: Idb0c55fc68cb8f45b3213c7134213904f227852e
This class makes it easier to test code that uses Settings:
1. Real device or emulator settings don't affect the code under
test; all settings always start off empty.
2. It's possible to change settings from the test without
affecting system settings.
3. No changes are needed to the code under test. The changes to
the tests are simple: just add a fake ContentResolver to
whatever mock Context is already used by the test, and make
that ContentResolver use the fake provider.
Bug: 23113288
Change-Id: I5e7e5a87571444ae49ccf551705620675a36cd17
Fix 2 problems of always-on vpn after always-on package is removed
1. Prevent network being locked down (blocking all network traffic)
Otherwise, user has no way to download the vpn app from Play Store,
and never be able to gain control of the network again.
2. Allow user to connect other vpn app.
Implementation
1. Switch off always-on mode if the package gets removed.
2. Restart always-on mode if the package gets replaced/upgraded.
Bug: 29050764
Change-Id: Id3e389ae0b11c6002a5167919292d9634c2014cb
