Most of this logic is simply removed from ConnectivityService.
The captive portal detection is now done by the NetworkMonitor.
The notification logic is still left in ConnectivityService as
it's used by both the NetworkMonitor and telephony's mobile
provisioning logic.
bug:17324098
Change-Id: Ibd1c42b1a75795f90a6483d3d0a5a14f88b193d8
Now that we support unreachable routes, use those to block
address families on VPNs. This is a much more elegant solution.
Also update LinkProperties when IP addresses are added and
removed, fixing a TODO.
Bug: 17462989
Change-Id: Ib749d84710dca70d672350b9f129bb91419ec77e
The BT and Wifi mechanisms for enabling Tethering did their own
permission checks. This set of changes unifies the check into
a ConnectivityManager function so they can be kept in sync.
bug:17435527
Change-Id: I8c157a5acf56ffbddd349cb6a45160ae7be8541b
Internal logic relies on Arrays.copyOf(), so always give ourselves
valid arrays, using shared empty objects to save overhead.
Bug: 17502649
Change-Id: I5dbb00545bdfe45bbd48144ab505ea08cc92cbcd
Currently just valid/invalid based on NetworkMonitor findings.
Changed NetworkMonitor to start out in default state since starting in Offline causes
a spurious invalid report at creation time.
Added some logging.
bug:17395269
Change-Id: I9ae650b561834d8f8979033744d97df852e76df9
The major change is that consent is now "sticky" and lasts until the
user explicitly disables the VPN connection.
Bug: 17474362
Change-Id: Id4e7807e635bbfc7645741135209d46763e280f9
These APIs were added because we thought we needed them to provide
seamless transition from one server backend to another using local IP
addresses to distinguish between the backends. I.e., connections whose
local IP address was old would be routed to the old backend; connections
whose local IP address was new would be routed to the new backend.
It turns out that's not needed. VpnService already supports seamless
re-establishment, so VPNs just need to call establish() again with a
different IP address. I've verified with a custom VPN app that this
works, and can distinguish traffic based on the old and new addresses.
Nobody is using these APIs at the moment, so we could even consider
removing them altogether, but I prefer just hiding them, just in case.
Bug: 15409819
Change-Id: I30949926a0f859c9d839981ccbc5d8e1e535a3a5
HttpHandler and HttpsHandler classes have a lot of bug fixes baked into
them that the Network.openConnection() API should be using, for example
disabling SPDY support.
bug:17420465
Change-Id: I9f1472753a542d1dd6bffde3a60c37a9145098aa
Future HTTP requests could use an old socket that's bound to a different Network
causing unexpected results. DNS results could also not be appropriate.
bug:17283566
bug:17432215
Change-Id: I88b40b723c7b442000cafe8ce8b9d989d8995991
Network Factories are allowed to go below, but networks need to be
constrained. Allowing the network to go below 0 meant that -1 could
sometimes leak through and foul the logic.
The core of 17361330 will be fixed when we stop sending scores for
listens to NetworkFactories, but it exposed this issue too. Summary:
1 - add a network listener. This isn't a request so it's not sent
to networks.
2 - alter your score (ethernet sets score to -1 when the link goes
down) (16:07:39.782)
3 - a bug in ConnectivityService causes score changes to get sent for
all network requests and network listeners causing NetworkFactories
to no see 2 entities. This bug will be fixed by a pending change
(https://googleplex-android-review.googlesource.com/#/c/540840/).
This causes the ethernet NetworkFactory to see two entities, both
served by networks of score -1. (16:07:39.989)
4 - disconnect Ethernet - this only sends 0 scores for known
requests, not network listeners. Had it been sent for both entities
they both would have evaluated that the networkfactory score (-1)
was lower than the request score (0) and both released their
refcount. (16:08:03.147)
5 - this means the listener is tracked by the EthernetNetworkFactory
with a score of -1 while the factory itself has a score of -1 so the
network release isn't called.
bug:17361330
Change-Id: Ife34ca0f9c233dd3c3df80f6fea580af43afcdeb
Using reflection you could do this and it would crash the system.
Thanks, ServiceFuzzer!
bug:17379629
Change-Id: I8b470bda78a69761ccd92496746f5d295b5d07f2
This constructor does nothing, including doing nothing with its
only argument. This causes it to return a NetworkInfo for
TYPE_MOBILE no matter what was passed in.
Bug: 16610051
Change-Id: I4ccd5ec050f7824fb06496c00fcd7901defeb7bd
If we don't do this, per-network HTTP requests will go over the
wrong network if any previous HTTP request was made by the same
app on another network.
Bug: 17300006
Change-Id: I1854c16dee6adb9e81fb12b097577439d69a644e
This makes tethered clients use the correct DNS servers when
tethering to non-default networks like the DUN APN.
Bug: 16357676
Change-Id: I8933b6de198a92c2aaf0291931ace8966ddba275
This reverts commit e0101cd and removes the related NetworkMonitor code.
The thinking is the broadcasts are not robust enough as they rely on apps
working together and are not sufficiently tested.
bug:17115050
Change-Id: I433032867cc4fea7191a1b13842b16825dc74df4
It seems that SOCKS isn't being handled explicitly by the
PacProxySelector, which results in them just being dropped
from the return list. This will sometimes end up switching
from SOCKS to DIRECT, which could be bad.
Bug: 17104885
Change-Id: Ic8a28230d3ae18c0abb000811a9100787c10c5e0
-The ability to launch VPNs is now sticky; once approved by the user,
further approvals are not needed UNLESS the connection is revoked in
Quick Settings.
-The old persistent notification has been removed in favor of the new
Quick Settings UI.
-The name of the VPN app is now pulled from the label of the VPN
service rather than the app itself, if one is set.
Bug: 12878887
Bug: 16578022
Change-Id: I102a14c05db26ee3aef030cda971e5165f078a91
LinkProperties can represent way more complicated configurations
than what we can actually apply to interfaces. This makes it
error-prone to use it to represent static configuration, both
when trying to apply configuration coming from LinkProperties
and when trying to save configuration from current
LinkProperties.
Instead, move static configuration (IPv4 only, since we don't
support static IPv6 configuration) into a separate
StaticIpConfiguration class.
Bug: 16114392
Bug: 16893413
Change-Id: Ib33f35c004e30b6067bb20235ffa43c247d174df
Add logic to obtain the mtu from the network PCO parameter and set it to kernel
when the mobile data connection is established. When there is no PCO mtu configured
from the network, the mtu size defined in the corresponding APN will be used. In case
no mtu size is defined for an APN used for data connection, the MCC/MNC based MTU
defined in the framework overaly will be applied.
bug:17046179
Change-Id: I6465d4b8f2076aaa380ae3617fb3f24adbe136d4
This addresses a TODO and also makes it possible to create
routes to destinations that are not valid LinkAddresses, such as
multicast addresses.
Bug: 16875580
Change-Id: Id4c77b00dc3064bf27d78cdcbbe035e645748cfe
