- Use getMobileBlabla() instead of getTotalBlabla()
- Have getTotalBlabla() reflect the fact that it includes loopback.
- Adjust checks to take headers, acks, and packet counts into account.
Change-Id: I9e1dfbbbdb9a6f932a14a2e3baee7a34469b9008
Checks java and native calls.
Adds some performance tests, but need to look at how that should be used.
Change-Id: I158164829da13302d9532275cef3482c4736168e
WifiManagerTest was waiting for SUPPLICANT_STATE_CHANGED_ACTION
after enabling or disabling WiFi. Fix the code to check if the
WiFi state is already the desired state, and if not, to wait for
the WIFI_STATE_CHANGED_ACTION broadcast intent.
Bug: 7082455
Change-Id: Id1c2242c32311084f5587ea5403f6b227d1b8b04
Ensure the packet count test does not fail due to a single attempt.
Our goal is to just ensure the packet count API exists.
Bug: 7001746
Change-Id: I8c6604528946166969126cd5b085024f81790a77
The new WiFi watchdog requires kernel/driver to export some packet loss
counters. This CTS tests whether those counters are correctly exported.
Change-Id: Ic764eff64ff2ef354b46f17e46eb74b14f191e4c
wifiConfigurations returned by getConfiguredNetworks() can be null.
It is only incidental that this test passed fine in the past.
Bug: 6822950
Change-Id: I570ab6e7fd6cc5eb150914a5d7ca527ccc2f16cd
Poll a bit longer waiting for getNetworkId and getWifiState to return
the proper values. The bug reports seem to indicate that the WiFi
is shutting down properly.
Add a new static check method to PollingCheck that I think is simpler
than the existing way of creating a new PollingCheck, overriding
check, and then making sure to call run. This new variation also allows
you to specify a message rather than just "unexpected timeout."
Bug 6443337
Change-Id: I9f7c942f6e26b957bb717b58b1ab984acc556bf7
Check if type of network is supported on the devices.
The device supporting TYPE_WIFI returns failureCode(-1).
But, not supported device returns wifiOnlyStartFailureCode or
wifiOnlyStopFailureCode.
Change-Id: I070bb19f6740367c476e78b884c2a36d111ff3ee
Make sure to wait first and then check the expected state rather than
check the expected state first and then waiting.
Bug 6293413
Change-Id: Ibcb1127b935708c51fafdc8624cb3a7eefc01bda
com.android.cts.stub is growing very large. Split apart some of
the popular components into separate static libraries. This
should allow packages to depend on the components they need
rather than all of cts.stub.
Current code at the moment doesn't have to be changed, because
I have CtsTestStubs depending on these new shared libraries.
However, change the net package to depend on the ctstestserver
static library as a proof of concept rather than including
its source directly.
Change-Id: I32c54eab3ddfb1d4391d6ffc347fbc9cb2fe97f9
