Add new SubscriptionPlan API to describe the various types of carrier
data plans. Internally the OS will only use the first plan for
driving policy, but it will blindly plumb through the details for
Settings to display any secondary plans.
As part of getting Settings ready to roll towards SubscriptionPlan,
reduce references to NetworkPolicy internal fields. All usage cycle
details are now obtained from an Iterator which converts to
SubscriptionPlan under the hood.
Replace all data usage cycle calculation with new SubscriptionPlan
implementation, and retrofit large suite of existing tests to
exercise and verify the new logic.
Offer a debugging property that can be used to return "fake" plan
examples for testing.
Bug: 63391323
Test: bit FrameworksServicesTests:com.android.server.NetworkPolicyManagerServiceTest
Exempt-From-Owner-Approval: Bug 63673347
Change-Id: I889c653980eeb7887abdfa4f5b6986f35855ee6d
For a long time we've had a nasty tangled dependency between Wi-Fi
and NPMS, since they both persisted different details for configured
networks. As part of preparing for new carrier data plan APIs, move
the tracking of meteredness over to WifiConfiguration.
This also cleans up how meteredness is communicated through
NetworkAgents to rely completely on NET_CAPABILITY_NOT_METERED by
removing the metered flag on NetworkInfo, which has caused confusion
and staleness.
Migrates any existing user-configured metered networks over to
WifiConfiguration once the device finishes booting.
Remove support for NetworkQuotaInfo, since this information can no
longer be made available to apps. Frustratingly, some apps are
using it, so keep the object around returning stub values, and shame
them in the logs.
Bug: 63391323
Test: builds, boots, Wi-Fi policy is upgraded
Exempt-From-Owner-Approval: Bug 63673347
Change-Id: I64f865ddeb65cfcd330f8d2a847368abdf960a07
Add note for VPN developers that VPN apps started in the background must transition to the foreground in Android O to avoid the system stopping them.
Staged at: go/dac-stage/reference/android/net/VpnService.html
Test: make ds-docs
Bug: 38023983
Change-Id: I33c7ca1717c332ffab495eb51c4c6b9c5c304cef
This patch adds a InitialConfiguration class to IpManager for specifying
IP information in IpManager ProvisioningConfiguration at IpManager
startup.
At the moment this InitialConfiguration is not used, but is validated in
startProvsiioning if ProvisioningConfiguration includes one. It will be
integrated into IpManager IP provisioning logic in follow-up patches.
This patch also includes an example of data driven unit tests using a
table of test case. The highlights of this methodology are:
1) easy extensibility for new test case,
2) rich and informative error messages,
Unfortunately Java support for inlined data structure literals is poor
and some companion static methods for data generation are required for
enabling this methodology.
Bug: 62988545
Test: added new test in FrameworksNetTests,
$ runtest frameworks-net
$ runtest frameworks-wifi
Merged-In: I060b02603af7d73a6407df89344bf0c000574af2
(cherry pick of commit 2757fcf3a1)
Change-Id: I48dbf89232d7758f1b07ed4d76ce93281e5c6b53
Currently, netd is the only source of tethering statistics.
In order to support multiple sources, define a new
ITetheringStatsProvider interface that can be registered with
NetworkManagmentService. Convert the existing code into the
first ITetheringStatsProvider.
Bug: 29337859
Bug: 32163131
Test: builds, boots
Test: tethering stats continue to be collected
Change-Id: Ie1b5a5e47ae4bf5af922365b09fa241e834236e4
This patch adds a InitialConfiguration class to IpManager for specifying
IP information in IpManager ProvisioningConfiguration at IpManager
startup.
At the moment this InitialConfiguration is not used, but is validated in
startProvsiioning if ProvisioningConfiguration includes one. It will be
integrated into IpManager IP provisioning logic in follow-up patches.
This patch also includes an example of data driven unit tests using a
table of test case. The highlights of this methodology are:
1) easy extensibility for new test case,
2) rich and informative error messages,
Unfortunately Java support for inlined data structure literals is poor
and some companion static methods for data generation are required for
enabling this methodology.
Bug: 62988545
Test: added new test in FrameworksNetTests,
$ runtest frameworks-net
$ runtest frameworks-wifi
Change-Id: I060b02603af7d73a6407df89344bf0c000574af2
This change simply adds a new constant, `TRANSPORT_LOWPAN`, for
identifying low-power wireless networks like Thread.
Bug: b/33073713
Test: builds
Change-Id: I50d9b8df1a0270e06c28f001adf6c52a142a85af
Merged-In: Ie4aa77496f8ff466fa1a5fbc556e9c029457a689
(cherry pick from commit 557e0c38b9)
This patch allows to use TYPE_NONE for the legacy network type variable
of NetworkInfo. This usage is "safe" with respect to legacy APIs using
network types as most of them already returns null or do nothing for
TYPE_NONE.
Of the existing APIs in ConnectivityManager that accept a network type
argument, those which were already returning null or doing nothing for
TYPE_NONE are:
getNetworkInfo(int)
getNetworkForType(int)
stopUsingNetworkFeature(int, String)
networkCapabilitiesForType(int)
requestRouteToHostAddress(int, InetAddress)
reportInetCondition(int, int)
isNetworkSupported(int)
getLinkProperties(int)
Only setProvisioningNotificationVisible needs an additional guard
against TYPE_NONE.
Bug: 30088447
Bug: 62844794
Test: runtest frameworks-net
Change-Id: I112596fcd03d3c2cd42a2a84d265adb38e3944bb
This patch allows to use TYPE_NONE for the legacy network type variable
of NetworkInfo. This usage is "safe" with respect to legacy APIs using
network types as most of them already returns null or do nothing for
TYPE_NONE.
Of the existing APIs in ConnectivityManager that accept a network type
argument, those which were already returning null or doing nothing for
TYPE_NONE are:
getNetworkInfo(int)
getNetworkForType(int)
stopUsingNetworkFeature(int, String)
networkCapabilitiesForType(int)
requestRouteToHostAddress(int, InetAddress)
reportInetCondition(int, int)
isNetworkSupported(int)
getLinkProperties(int)
Only setProvisioningNotificationVisible needs an additional guard
against TYPE_NONE.
Bug: 30088447
Bug: 62844794
Test: runtest frameworks-net
Change-Id: I4455f2726d06406047086368628c1f253d854d8d
Apps with a normal UID are typically isolated enough to not require
socket tagging; we're mostly interested in tracking down internal
UIDs that have lots of code sharing the same UID.
Also fix up everyone doing manual string checks of Build.TYPE, since
we now have first-class fields for those.
Bug: 38126076
Test: builds, boots
Change-Id: I3a40348196bd8459289f2b9355d9783a07f1e7dd
This way an app store can shift blame for update-related network
traffic onto the app that is being updated.
Using a well-known tag makes it easy for developers to identify
that they didn't explicitly request the traffic at runtime, similar
to how backup/restore traffic is handled.
Bug: 38282350
Test: builds, boots
Change-Id: I003dd7c9615d4ab318250f1e44fa5d195ac94d23
adds privileged permission for getCaptivePortalServerUrl
adds tether privileged permission for
startTethering,isTetheringSupported
bug:62348162
Test: make and manual testing
Change-Id: I8eb8e3c9dcd7201abe9ea303ee57fe99073d67eb
Tracks the revert of the corresponding conscrypt CL.
This reverts commit 7fdce769c3.
This is a clean revert. The upload hook to fix lint errors was not run
(this CL was uploaded with --no-verify).
Test: Treehugger
Test: make droid cts
Bug: 62424503
Change-Id: I5b8728575c4c02de10fff227ba40ec6ddc7ffa46
This change simply adds a new constant, `TRANSPORT_LOWPAN`, for
identifying low-power wireless networks like Thread.
Bug: b/33073713
Merged-In: Ie4aa77496f8ff466fa1a5fbc556e9c029457a689
(cherry pick from commit 557e0c38b9)
Change-Id: I21f9b41b8b31c63ceeb1bc9c965f6da2614c356a
Test: runtest frameworks-net (not in original commit message)
This patch adds translation from ConnectivityMetricsEvent to
IpConnectivityEvent of recently added fields:
- top-level network id
- top-level ifname
- transports
Also adds inference of link layer from transports or ifname.
At the moment these new fields are not populated in
ConnectivityMetricsEvent. Follow-up patches will fill this gap for
the events of the android.net.metrics package.
Test: new unit tests, $ runtest frameworks-net passes
Bug: 34901696
Merged-In: I563a6a3183470bdfaabb7c781a1beaf6b1058bf0
(partial cherry pick from commit 73fd4d1faf)
Change-Id: I6a00270e73a1bd07f23c367f2394d90a43ced47a
Test: runtest frameworks-net (not in original commit message)
This patch changes the validation of unregisterNetworkCallback in
ConnectivityManager so that the caller can better distinguish the case
of a callback that was never registered from the case of a callback that
has already been unregistered.
Bug: 62497809
Test: runtest frameworks-net passes
Merged-In: I58eda22725dd4e67dc4b64207e38cf482032df10
(cherry picked from commit 605b295011)
Change-Id: I4d2be5d9037cfec03218b702d707f50808ee0694
This patch changes the validation of unregisterNetworkCallback in
ConnectivityManager so that the caller can better distinguish the case
of a callback that was never registered from the case of a callback that
has already been unregistered.
Bug: 62497809
Test: runtest frameworks-net passes
Change-Id: I58eda22725dd4e67dc4b64207e38cf482032df10
This change makes it impossible to return a badged wifi icon, even if
scoring ui is enabled. It also prevents IAE from net badgecurve enums.
Bug: b/35628911
Test: make -j40
Change-Id: I014f27e5c4e42075e7e62da7fffb9ecf38f6eb4c
