After upgrading libc++, math.h happens to be included at the point
where this file tries to define NAN. Redefining a macro to a different
value is an error, but #undef'ing a macro is OK even if the macro isn't
actually defined.
Bug: b/175635923
Test: treehugger
(cherry picked from https://googleplex-android-review.googlesource.com/q/commit:dd390faa2fcb49ce0e0ebc6ef180118ec370f957)
Merged-In: Ie8e3e87189bcd7ed8fccccfc0a79e04a5bfda16b
Change-Id: Ie8e3e87189bcd7ed8fccccfc0a79e04a5bfda16b
- Added API to check wifi chip supports AP interface
- Check AP support as a precondition for AP related TCs
Bug: 284925541
Test: atest VtsHalWifiApV1_5TargetTest
Change-Id: I1121d7b787ed972980bab0e05a52f3ca2f3d0a24
the Hostapd AIDL VTS tests.
At the beginning of each test, we now:
1. Disable the Wifi framework, Hostapd, and Vendor HAL.
2. Enable the Hostapd and Vendor HAL.
At the end of each test, we now:
1. Disable the Hostapd and Vendor HAL.
2. Enable the Wifi framework.
Bug: 283402709
Test: atest VtsHalHostapdTargetTest --iterations 10
# Tested on both an Android T and U device.
# All tests passed in both cases.
Change-Id: I8b5e24ad44d13fad09d84f1b212e60ca8df3aa7e
Hostapd AIDL VTS tests.
Main purpose will be to enable/disable the
HALs and Wifi framework before and after
each test.
Bug: 283402709
Test: atest VtsHalHostapdTargetTest
Change-Id: Ib3ceca4ad25f472e1b75fa95661235c880a489fd
the event bitmask.
Comments on ag/24025770 suggest that the bit
shift will lead to undefined behavior for
event codes greater than 31.
Bug: 290318208
Test: atest VtsHalWifiNanIfaceTargetTest
# manually set some of the expected
# events codes to > 31 to verify that
# the bit shift works as expected
Change-Id: I09e11dac8acf90baf047e24cebe1d01970b1dc8a
to better handle overlapping events.
Reimplementation of aosp/2473685 for
the AIDL VTS tests.
Bug: 290318208
Test: atest VtsHalWifiNanIfaceTargetTest
Change-Id: I1a0e6ab8dd529c10fcec8f26424a7181e7d91568
For Panel TV devices, the vendor can implemented TV specific mDNS
offload instead of APF in U. If TV specific mDNS offload is implemented.
MdnsOffloadManagerService will exist in the system ext partition. The
APF vts will be skipped if MdnsOffloadManagerService exist.
Bug: 283038712
Test: atest VtsHalWifiStaIfaceTargetTest
Change-Id: If55ec42507460b9a2c6eee683d85b8109f2af236
This includes:
- Adding a README to help users get started.
- Improving the test script to avoid issue where
device is not ready for the next ADB command
after 'adb root'.
Bug: 271914366
Test: Manual test
1. Reboot device.
2. ./runtests.sh
3. Tests are expected to execute.
Change-Id: I9f452384a7067708f617d1e932f06b2f5c7ab119
in the WifiChip constructor.
gTest suite currently cannot mock the legacy HAL
call for getSupportedIfaceConcurrencyMatrix.
If we set using_dynamic_iface_combination_ to true
in the unit tests, we can avoid making this call.
Bug: 271914366
Test: ./runtests.sh # gTest
Test: atest VtsHalWifiChipTargetTest # VTS
Change-Id: Ic43daab6ff5fdc5f78af614ad775f436fbbe8726
for the Wifi VTS tests.
Main method is getBridgedWifiApIface_1_6, which
allows us to configure the chip based on the
Bridged AP IfaceConcurrencyType, rather than
the AP IfaceType like in getBridgedWifiApIface_1_5.
Implementations were largely copied from
wifi_hidl_test_utils and wifi_hidl_test_utils_1_5.
Bug: 283402709
Test: atest VtsHalHostapdTargetTest
# tested on a Bluejay device that supports
# the AddAccessPointWithDualBandConfig test
Change-Id: I2571876149fc14c8de02e1cec9934dd052cada5c
Merged-In: I2571876149fc14c8de02e1cec9934dd052cada5c
for the Wifi VTS tests.
Main method is getBridgedWifiApIface_1_6, which
allows us to configure the chip based on the
Bridged AP IfaceConcurrencyType, rather than
the AP IfaceType like in getBridgedWifiApIface_1_5.
Implementations were largely copied from
wifi_hidl_test_utils and wifi_hidl_test_utils_1_5.
Bug: 283402709
Test: atest VtsHalHostapdTargetTest
# tested on a Bluejay device that supports
# the AddAccessPointWithDualBandConfig test
Change-Id: I2571876149fc14c8de02e1cec9934dd052cada5c
Merged-In: I2571876149fc14c8de02e1cec9934dd052cada5c
for the Wifi VTS tests.
Main method is getBridgedWifiApIface_1_6, which
allows us to configure the chip based on the
Bridged AP IfaceConcurrencyType, rather than
the AP IfaceType like in getBridgedWifiApIface_1_5.
Implementations were largely copied from
wifi_hidl_test_utils and wifi_hidl_test_utils_1_5.
Bug: 283402709
Test: atest VtsHalHostapdTargetTest
# tested on a Bluejay device that supports
# the AddAccessPointWithDualBandConfig test
Change-Id: I2571876149fc14c8de02e1cec9934dd052cada5c
APF is support is mandatory since VSR 14. Before that, the APF
requirement is optional. Vendors can opted into upgrading their
HAL to AIDL, so the VTS need to check if vendor API level on the device
and only mandate the APF capability check if the vendor API level is
greater or equal to 14.
Bug: 283181837
Test: atest VtsHalWifiStaIfaceTargetTest
Change-Id: I21513187d4ca99d721d0e9dc5e243adad77ab8c6
supplicant_test_utils::initializeService() only stop Supplicant,
don't start it. refer Android T code, It should start Supplicant
when initializeService().
Bug: 279572212
Test: VtsHalWifiSupplicantP2pIfaceTargetTest pass
Change-Id: I9f4b3e21ee0865eb95d71f7f832b2eb405de9607
SetStrictConversativePeerMode HAL API is added in
AIDL version 2. So if the HAL interface version is
less than 2, skip the test.
Bug: 278169264
Bug: 277318919
Test: atest VtsHalWifiSupplicantStaNetworkTargetTest
Change-Id: I7f8ace88b0f37679ecaedd5370b47e220c437d50
VtsHalWifiHostapdV1_2TargetTest.
Test currently sets up a new interface
and uses that interface's name. On Pixel,
the new interface is wlan1, which does not
seem to support the OpenAccessPointWithoutAcs
configuration. However, the default SAP
interface wlan0 does support it.
The method to get the default interface name
is taken from this file before ag/19202002
was merged.
Bug: 270509029
Test: atest VtsHalWifiHostapdV1_2TargetTest
(on a Bramble device running rvc-qpr3-release)
Change-Id: Ic3c516f452dfb54f155246073a9bca20dbecaa0f
Update the aidl API documentation to describe how setDtimMultiplier API
is supposed to work, i.e. we can call this API to set a maximum DTIM
multiplier value to use in driver, any larger value must not be accepted
and meanwhile it allows to set a specific multiplier by OEMs. Try to
make the description clear and avoid the ambiguity.
Bug: 266256943
Test: m
Change-Id: Ic7d2284a25e32d1b0752cf2bc832c069f646fcea
dump method is called.
This should prevent a race condition
between stop() and dump().
stop() already acquires the global lock
when it calls validateAndCall().
Bug: 274641043
Test: adb shell dumpsys wifi # validates dump()
Change-Id: Ie2df0861308954cd2e399c9604d4ae0dcddf95fb
VTS test for IWifiChip#getFeatureSet.
Some implementations of the legacy HAL,
such as Goldfish, do not have any
supported features.
Bug: 273886122
Test: atest VtsHalWifiChipTargetTest
(tested on Cuttlefish)
Change-Id: I3ca22187cab0e77d36a28908e57ff4970535e4ae
