Previously, after calling unregisterNetworkOffer(), onNetworkNeeded() /
onNetworkUnneeded callbacks are still being called until CS's handler
thread has finished running. While this behavior could be taken care of
in the network provider implemention (e.g. the ethernet service), it is
much simpler to deal with in NetworkProvider.java.
As NetworkProvider.java wraps the passed callback object inside a proxy
that it forwards to CS, it should be as simple as adding a boolean to
the proxy that is checked inside the executor thread before the
implementation's callback is called. Since for every call to
registerNetworkOffer() with a new callback object a new proxy object is
created, this should be safe. Even if the callback object is reused on
the user side (such as ethernet service does right now), this should be
okay, as the implementation will only receive callbacks from the active
proxy.
Test: atest NetworkProviderTest
Bug: 171872016
Change-Id: Iaf84b5801ad3ee44895e933763927151ea5824dd
Without an IpClient, there is otherwise no way to configure IPv4
addresses on a tap interface in tests.
Test: TH
Change-Id: I6a788687708b2034c464befdfec9267ce3399d70
Add the required lib to the user of net-utils-framework-common-srcs
which has LocationPermissionChecker.
Bug: 158059172
Test: m
Merged-In: Id54de2bd94e8b5821814715683b5b2520f1f026e
Change-Id: Id54de2bd94e8b5821814715683b5b2520f1f026e
(cherry picked from commit d795228affd7e7ad828adb2bf4fbcb9093c3f377)
This commit adds ConnectivityManager#getFirewallChainEnabled to read the
current firewall chain status
Bug: 208371987
Test: m
Change-Id: I1eadb69f953af5d031cd8dabde3e1f098cf0f4df
Both setting the carrier up / down and creating an tun / tap interface
without carrier need TUNSETCARRIER to be present in the kernel.
Bug: 236783925
Test: documentation only change
Change-Id: Ie7f7d2c81c926adbc10b29b4b1748e47edaa4cb7
* changes:
TestNetworkInterface: add support for MacAddress and MTU
TestNetworkService: add support for creating iface without carrier
TestNetworkService: Add support for toggling carrier on tun/tap
TestNetworkService: Fix TestNetworkInterface parcel flags
Patrick Rohrauthored and Gerrit Code Reviewcommited 2022-06-15 20:53:30 +00:00
Both MacAddress and MTU size are often needed in tests and very painful
to get a hold of without system server permissions.
Test: TH
Change-Id: Ie1a91959916ea39277e76aa1b61f38d59a5d2a90
Disabling the carrier of a tun/tap interface before the interface is
brought up is useful in testing.
Test: atest EthernetManagerTest
Change-Id: Ic189b2c0eda76e3f58c3f664e5f5d18c2d7e6e96
Adds support for setting carrier to on/off for a tun/tap interface. It
turns out that TUNSETCARRIER is also usable by tap interfaces, and
TUNSETLINK does not work as expected.
Test: atest EthernetManagerTest
Change-Id: I39d70e0a65a34e7a4c4df70c53e4cc781a24a213
This change allows tests to mock NetworkAgent, and the methods
required for ensuring that the VPN is migrating properly.
Bug: 235853154
Test: Treehugger
Change-Id: I49333c010cc7f3d60ffd39f853777bfc52de3a89
Benedict Wongauthored and Lorenzo Colitticommited 2022-06-15 09:16:50 +00:00
PARCELABLE_WRITE_RETURN_VALUE will close the fd when writeParcelable is
called. This is not always preferable (specifically, if
TestNetworkInterface is passed as an "in" parameter).
writeParcelable(FileDescriptor, ...) should therefore use the flags
passed to writeToParcel which should be correct for all cases.
Test: atest EthernetManagerTest
Change-Id: Ibb5f14028a3d686157504da3657a18ec0bc28110
(This rolls forward part of a previous change, now that jarjar was fixed
to not get very slow when the number of rules increases).
Jarjar rules are hard to keep in sync with code, and hard to maintain
manually as the distinction between what should and should not be
jarjared is not always clear. This results in unsafe binaries that are
manually maintained, and developer frustration when something fails due
to incorrect jarjar rules.
Add utility to autogenerate jarjar rules, which can be run at build time
time (via a genrule) instead. The generator scans pre-jarjar
intermediate artifacts, and outputs jarjar rules for every class to put
it in a package specific to the module. The only exceptions are:
- Classes that are API (module-lib API is the largest API surface of
the module, so module-lib API stubs would typically be used)
- Classes that have unsupportedappusage symbols
- Classes that are excluded manually (for example, because they have
hardcoded external references, like for
ConnectivityServiceInitializer in SystemServer).
Bug: 217129444
Test: atest jarjar-rules-generator-test;
Change-Id: I3493957e39a661b6c2e330944e7c3023b8f3203e
This change allows tests to mock NetworkAgent, and the methods
required for ensuring that the VPN is migrating properly.
Bug: 235853154
Test: Treehugger
Change-Id: If4de0935a3c8627ca12db52ff223834f9f7237e8
LOCKDOWN_VPN was in the FirewallChain IntDef but this was not a right
place because LOCKDOWN_VPN was not a valid value for Connectivity APIs
that take an argument annotated with @FirewallChain(setUidFirewallRule,
setFirewallChainEnabled, replaceFirewallChain).
LOCKDOWN_VPN was in the FirewallChain IntDef because
BpfNetMaps#setUidRule was used to add/remove LOCKDOWN_VPN entries.
This commit adds BpfNetMaps#updateUidLockdownRule and uses this to
add/remove LOCKDOWN_VPN entries instead of BpfNetMaps#setUidRule and
removes LOCKDOWN from FirewallChain.
Bug: 206482423
Test: atest TrafficControllerTest ConnectivityServiceTest
PermissionMonitorTest HostsideVpnTests#testBlockIncomingPacket
Change-Id: Iff9b9792fc0f208f153e10e396c6d5034b412d7c
The depfile concept in Soong and Ninja isn't applicable in Bazel because
Bazel requires listing the deps explicitly. This CL changes the module
type from gensrcs to java_library with proto.type set to "stream".
The end goal is to eventually deprecate depfile in gensrcs to ensure all
gensrcs modules are convertable to Bazel.
Test: CI
Bug: 179452413
Change-Id: I921324e6260b7de6329a1ff445b6326643a9c526
This reverts commit f1ab80ea62.
Reason for revert: DroidMonitor: Potential culprit for Bug 233698500 - verifying through ABTD before revert submission. This is part of the standard investigation process, and does not mean your CL will be reverted.
Change-Id: I7de63e0a3656194b94b29e28862c62fc300dfdfa
This broke my tests in unexpected ways as creating a new
NetworkRequest.Builder() changed the static request object.
Test: TH && atest EthernetManagerTest
Change-Id: Ia6c39145cf92a2795f7fc2f4efbcb46d193f6b72
This reverts commit 53eb35cd82.
Reason for revert: b/233081774, caused long build times in FrameworkNetTests
Change-Id: Ide01c994f694b55fc0e87bb0e81584ac44196e68
Currently, even when VPN Lockdown mode is enabled, incoming packets are
not dropped if VPN is not connected.
This commit fixed this issue.
After this commit, If VPN Lockdown mode is enabled, incoming packets
are dropped regardless of the VPN connectivity.
Bug: 206482423
Test: atest TrafficControllerTest ConnectivityServiceTest PermissionMonitorTest
Change-Id: If52ece613c8aac1073355e43b6fb9cb3fcc87d1d
Jarjar rules are hard to keep in sync with code, and hard to maintain
manually as the distinction between what should and should not be
jarjared is not always clear. This results in unsafe binaries that are
manually maintained, and developer frustration when something fails due
to incorrect jarjar rules.
Autogenerate jarjar rules at build time instead. This is achieved by
introducing a jarjar-rules-generator python-based library, which scans
pre-jarjar intermediate artifacts, and outputs jarjar rules for every
class to put it in a package specific to the module. The only exceptions
are:
- Classes that are API (module-lib API is the largest API surface of
the module)
- Classes that have unsupportedappusage symbols
- Classes that are excluded manually (for example, because they have
hardcoded external references, like for
ConnectivityServiceInitializer in SystemServer).
This change causes all classes in framework-connectivity(-t) and
service-connectivity to be jarjared into android.net.connectivity, but
still avoids jarjaring classes in com.android.server as before, to keep
it small.
For many classes this differs from the original jarjar rule.
Notes on implementation:
- connectivity-jarjar-rules now has a subset
framework-connectivity-jarjar-rules containing only the rules
necessary for framework-connectivity. This is necessary because
framework-connectivity cannot depend on rules generated based on
service-connectivity, as there would be a dependency cycle
(service-connectivity depends on framework-connectivity); Soong even
crashes with a stack overflow.
- framework-wifi.stubs.module_lib is added to
framework-connectivity-pre-jarjar as it is necessary to build it (it
is already in impl_only_libs in the defaults).
It is unclear why framework-connectivity-pre-jarjar could build
before that (possibly because it was only used as "lib" ?)
- Fix package-private visibility; for example NattSocketKeepalive,
TcpSocketKeepalive are not API so should be jarjared, but are used
by ConnectivityManager which is not jarjared, so they are not in the
same package after the change. Package-private members in the
former 2 need to be public to be accessible. Changes in this commit
are all that is needed, as demonstrated by followup commits that move
the classes to a different package without further changes, and that
enforce that no class in an API package gets jarjared.
- framework-connectivity-internal-test-defaults is separated from
framework-connectivity-test-defaults, for unit tests that need to
access internal jarjared classes. Such tests need to use the jarjar
rules themselves too, so this is only appropriate for connectivity
internal unit tests.
Test: atest ConnectivityCoverageTests CtsNetTestCases
Bug: 217129444
Change-Id: Ied17c3955ea2fda130089265d02908937ad8af1e
Multiple enterprise slice can be setup within single user profile based
on different uids. So do not remove profile network preference with same
user profile but with different uids
Bug: 229644102
Test: manual system test and ConnectivityServciceTest
Change-Id: I897b643e01240958fff575de9e15182069efc698
ConnectivityManager have a self reference static instance. This
causes a leak if context is refernce from a static variable. Using
the applicationContext, which will never be freed for the life of the
application, for the sInstance static instance.
Bug: 202978965
Test: atest ConnectivityManagerTest
Change-Id: I87206e1bfbb1f877b5a10f5fdbc25e2f9f11bef4
markchienauthored and Mark Chiencommited 2022-04-29 08:01:23 +00:00
The issue scenario is:
- Telephony registers a network agent and calls markConnected().
- NetworkAgent set mNetworkInfo to CONNECTED and call
queueOrSendNetworkInfo() with mNetworkInfo, but NetworkAgent
don't actually send a message to CS yet because the agent is
not registered.
- Telephony calls unregister because the cellular network is
disconnected.
- NetworkAgent set mNetworkInfo to DISCONNECTED, overwriting the
previous CONNECTED state, and then call queueOrSendNetworkInfo
again with mNetworkInfo. Again this doesn't send any message
because the agent is not connected.
- EVENT_AGENT_CONNECTED arrives. NetworkAgent replies all the
messages in mPreConnectedQueue, but NetworkAgent send two
DISCONNECTED NetworkInfos eventually.
The CONNECTED state should be sent to CS instead of latest state
DISCONNECTED. Thus, make a defensive copy when sending
NetworkInfo change to prevent state overwriting.
Without the NetworkAgent fix, the test will fail with no
onAvailable callback. Because the Network has never been martk as
CONNECTED.
Bug: 228623362
Test: atest FrameworksNetTests CtsNetTestCases
Change-Id: I11681743d3ff87ff9affd0b7e766894dc5111028
Patrick Rohr
authored and
Gerrit Code Review
commited 
Paul Hu
authored and
Chalard Jean
authored and
Treehugger Robot
authored and
Motomu Utsumi
authored and
Jooyung Han
commited 
Remi NGUYEN VAN
authored and
Benedict Wong
authored and
Lorenzo Colitti
commited 
Vinh Tran
commited 
Iván Budnik
commited 
Colin Cross
authored and
Sewook Seo
authored and
Sooraj Sasindran
commited 
Prerana Patil
authored and
Mark Chien
authored and
Taras Antoshchuk
commited 
Xiao Ma
authored and