When Wi-fi connects at L2 layer, the beacons reach and the device
can maintain a connection to the access point, but the application
connectivity can be flaky (due to bigger packet size exchange).
We now use Watchdog to monitor the quality of the last hop on
Wi-Fi using signal strength and ARP connectivity as indicators
to decide if the link is good enough to switch to Wi-Fi as the uplink.
ARP pings are useful for link validation but can still get through
when the application traffic fails to go through and thus not best indicator
real packet loss since they are tiny packets (28 bytes) and have
much low chance of packet corruption than the regular data
packets.
Signal strength and ARP used together ends up working well in tests.
The goal is to switch to Wi-Fi after validating ARP transfer
and RSSI and then switching out of Wi-Fi when we hit a low
signal strength threshold and waiting until the signal strength
improves and validating ARP transfer.
Change-Id: Ica593291ec7772da892f03cf45b649635b730c47
Use of multiple socket connections over wlan0 and p2p and
p2p state machine is now controlled entirely from wifi state machine
Add discovery stop to allow STA scans to proceed after p2p is used
Change-Id: I790c9112d3f475f638f06fc3ae9e191f6d90ef35
Allow multiple instances of WifiMonitor and WifiNative to be
created in preparation for multiple socket connections
Change-Id: I06a227f87f1406dc62f884d56ad639ac623414b6
Add the initial support for supporting seperate socket
connections on the p2p interface.
This is the initial change making primary interface connections
work alongside native support for p2p socket connection
Change-Id: I7619715ca75b10afd29af852279fe713331f75e0
- Move all supplicant commands out of jni code in preparation for multi-socket
changes
- Also, clean up some function names to remove the explicit command
Change-Id: Id609ac7df0792fb85b52554d159f22064227c584
- Change default GO intent to 7
- Handle P2P invitation scenario
- Fix p2p STA connect notification handling to figure device address. Proper fix
for handling both STA connection and disconnection is after we update supplicant
Change-Id: I9ff82f84ad3913905952d1119b7224dff41c24d9
For a connected network, keep status as CURRENT. For a network,
that is disconnected it should be ENABLED. A disabled network
will have the status as DISABLED
Also, add a unit test to ensure there is only one CURRENT network
that is connected
Change-Id: Iaa4a7124a0c372a8f6df3d846ae8c15d9b29cf13
- Add delayed disk write in WifiConfigStore
- Remove synchronization and keep all access to config store
throught the state machine thread
Change-Id: I53768a17895e48da7b99542ac95c6c2fddbcb021
At some point the system UI switched to using 5 rssi levels
while the framework only broadcasts a change for 4 levels.
Hence RSSI change can be slow to be displayed on the UI, fix this
Bug: 5732700
Change-Id: I21700f96ba7a8dd678fb2ccfff8ef59f9b696666
- Add device name on SSID for legacy devices to discover a given
group and join
- Do not remove a group for a device lost event which only indicates
discovery is stopped
Change-Id: Id431c29da371986e694ab230d72d32061b209ebe
= Provision Discovery:Is an optional discovery technique for figuring out
if a peer supports the specific WPS option. For compatibility reasons,
we will now always use it during connection but at the same time support
devices that request to support connection without it
= State machine and UI redesign around the new interaction
= Specify config methods explicitly for compatibility
Bug: 5599949
Change-Id: Ib79435dda472c43cc0746f56bf79ea896b2e33a4
- Use the right interface for untether
- Wait till tethering service confirms that untether() is successful before
shutting down soft Ap. This helps to make sure interface does not go away
before clean up is done.
Bug: 5659166
Change-Id: Iff9d4a67ff40664c2114e766590d6a83b7739738
Causes problems on our wext driver and potentially on cfg
based driver as well.
NLP can trigger very quick acquire and release within seconds
leading to driver switching between start and stop state.
We now keep driver up for atleast couple of minutes after a start
Bug: 5478196
Change-Id: I1b261578252c5fb9a65446241b51e5686d4d9cc3
Now that we also handle supplicant state change to identify that
a network is disconnected (in case CTRL-EVENT-DISCONNECTED goes missing),
it is dangerous to defer all supplicant state change messages
while in DisconnectingState.
It may happen that the CTRL-EVENT-DISCONNECTED goes missing while in
DisconnectingState resulting in a supplicant state change of 0 (disconnect)
getting deferred.
Eventually after a connection completes, the supplicant state change events
get handled and the state machine goes into DisconnectedState.
Fix by having state machine switch out of DisconnectingState once we
do not immediately see a CTRL-EVENT-DISCONNECTED state.
Bug: 5490789
Change-Id: Ia2263795e53c51da0a2bfeefecfeb6256d6c5267
