The venue friendly name comes from the network, and only plain strings
should be used in practice. writeCharSequence is also not usable in
CaptivePortalData, as it must build using module_api.
Use a String as the underlying type in CaptivePortalData, keeping
CharSequence in the API for consistency with UI-related elements.
This brings CaptivePortalData in line with downstream branches.
Bug: 183174863
Test: atest FrameworksNetTests
Merged-In: Ifa3ea9aae0cabf5757791c4ae13e6f551759ed97
Change-Id: Ic662d3b546d52f825f9c3a24c5e0d4ba03818ab0
The callers cannot call the hidden APIs after ConnectivityManager
became a part of mainline module, so expose them for callers.
Also change the value of ACTION_PROMPT_PARTIAL_CONNECTIVITY,
ACTION_PROMPT_LOST_VALIDATION and ACTION_PROMPT_UNVALIDATED because
of API lint errors.
Bug: 172183305
Test: Check private DNS settings is normal, and test NO_INTERNET
notification can be shown normally.
Change-Id: I715c766ad8e5eb54f4dc67239c1dbca7239506fc
Merged-In: I715c766ad8e5eb54f4dc67239c1dbca7239506fc
These metrics are deprecated so remove them from jobscheduler.
Also remove dumpDebug method from Network, NetworkRequest and
NetworkCapabilities because there's no caller anymore.
This change also for connectivity mainline module. These
three classes are inculded in framework-connectivity so
external module cannot have dependencies on its hidden API.
With this change, the dependencies can be removed.
(cherry-picked from ag/13959431)
Bug: 178777253
Test: FrameworksNetTests
JobStoreTest
adb shell dumpsys jobscheduler --proto
CtsIncidentHostTestCases:JobSchedulerIncidentTest
Merged-In: Ie0c540303ba06b8fba029d2b98ae753afb08c963
Change-Id: Ie0c540303ba06b8fba029d2b98ae753afb08c963
The value should be assigned as a long to do the bit calculation
as the mNetworkCapabilities is intended to be a long. Otherwise,
the value will be temporary assigned into an integer then
assigned to the target long. When the bit shift calculation
is out of the integer scope, the calculation will overflow and
result in unexpected bebavior.
Without assigning to a long, ConnectivityServiceTest will get
Out-Of-Memory in StringBuilder while generating toString() in
NetworkCapabilities after updating tests to verify
NET_CAPABILITY_VSIM and NET_CAPABILITY_BIP.
Bug: 130869457
Test: atest FrameworksNetTests
Change-Id: I4d34c1215c7efb6dc352c314107792e3fa512ad7
Added two new per-app OEM network preferences to be used only with
testing.
Although there is a tradeoff adding these preferences to
non-test code, the benefit is that it will allow testing of the OEM
network preferences app via CTS as using the shell with elevated
permissions is not an option due to the OEM network preferences API
potentially being abused. On consideration, a small amount of test code
was deemed an acceptable tradeoff if it enabled CTS testing.
Bug: 176496580
Bug: 176494815
Test: atest FrameworksNetTests
atest FrameworksNetIntegrationTests
atest CtsNetTestCasesLatestSdk
Change-Id: Ie896a7b40945a46a352f37d2c00671da8a8f4489
Most DNS fails are caused by network issues, like connection timed out.
Lower log severity from error to warning to conform to go/greenlog.
Test: 1. connect to a WiFi AP without WAN
2. atest CtsNetTestCases:android.net.cts.DnsResolverTest
3. adb logcat | grep resNetworkResult:android
Bug: 181269159
Change-Id: Ifeaf250c91a4b9123d200bd62f085c11009f2491
This exception log is usually printed when the device connects to
a wifi. The cause is that when the device connects to a wifi,
wifi will send CONNECTING state first, then CONNECTED state.
And in CONNECTING state, wifi will send the LinkProperties to
ConnectivityService and ConnectivityService will try to update
that LinkProperties(including interface) to netd. But the netid
hasn't set to netd yet before wifi CONNECTED, so this exception
log will be printed.
There is no really issue in this case, so downgrade the log level
might be a better thing for all Android developer.
Bug: 181269159
Test: m
Change-Id: I61c09352225a610fcd6d3dd33b782ee114c35fc5
This is a small cleanup of a function called by DevicePolicyManager
coming from a conflicting change. This no longer makes a concrete
difference but is still a good change.
Test: m services.devicepolicy
Bug: 172183305
Change-Id: I7ee907314ddb253eb4e97d177f0ea0ab3b58cf03
This was broken by aosp/1657763, only on eng builds because
of a wtf. The root cause of the issue is that the test is
acquiring MAINLINE_NETWORK_STACK after it registers a callback
but before in unregisters it. After aosp/1657763, the objects
keeping track of callbacks are separate for processes with and
without MAINLINE_NETWORK_STACK, which means the removal is
looked up in a different counter.
Solve this by having the NRI remember the relevant counter
keeping track of it.
Bug: none
Test: testCaptivePortalApp
Change-Id: I566a28522fa68be9cc2bdb4624b588b4f01d3b7e
