We're going to add some non-trivial logic in wifi chip to handle
multiple chip configurations (to support devices with different
capabilities). So, adding a unit test framework in preparation
for testing those changes.
The mocks only mock legacy HAL methods which are in use during chip
configuration & interface creation/deletion.
Bug: 68775880
Test: Compiles
Change-Id: I8575df02f54656f77077d14ac9535ccc974bdff3
Making all of the iface object fields in WifiChip to vectors to support
multiple ifaces of the same type.
Bug: 65671875
Test: Device boots up and connects to wifi networks.
Change-Id: I4c0c927c0269e54210cc1f81203c1797d72e969c
This was previously disallowed because WifiLegacyHal.stop() was non blocking.
WifiLegacyHal.stop() was made blocking using ag/2715673, so reconfiguration
should be supportable now.
Note: This is a partial revert of ag/2125543.
Bug: 65671875
Bug: 70411703
Test: Will send for regression tests.
Test: Reverted the VTS CL: ag/2139830 and ensured that the test works
now. Can't check-in that revert because it would not be backward
compatible.
Change-Id: Ia5f9ccaf5403d171b0def7740d2adece7eb3c157
Use the top level clang-format setup by the HIDL team to format the
entire implementation.
clang-format -i --style file wifi/1.2/default/*
Bug: 32287573
Test: Compiles
Change-Id: I336c21fd9bfdc560117aa7212f92ab5f01df4b8e
Add support for concurrent interfaces in the WifiLegacyHal class:
a) Removed the hardcoded "wlan0" interface handle in WifiLegacyHal.
b) Modified all the interface specific functions to accept the |iface_name|
argument on which the operation needs to be performed.
Each IWifiIface object will hold the name of the underlying network
interface (wlan0, wlan1 or p2p0) which it is representing.
All IWifiChip operations which needs an iface name will continue to use
the default "wlan0".
Bug: 65671875
Test: Device boots up and connects to wifi networks.
Test: Will send for regression tests.
Change-Id: I9bd9c2a9ba33ac1ea5677fc5d7c261d8eac08e1d
Currently the HAL shim uses fake names to ensure that each type of
IWifiIface has a unique name. This is not a true reflection of the
network interfaces exposed by the wifi driver. So, change the HIDL shim
to use the corresponding interfaces names.
IWifiStaIface, IWifiApIface & IWifiNanIface all use the same "wlan0"
network interface.
IWifiP2pIface uses the "p2p0" network interface.
In the future, we'll be extending this to create a second IWifiStaIface
or IWifiApIface using "wlan1" network interface.
IWifiRttController does not need to be associated with an iface object.
So, it will just default to using "wlan0" always.
TODO(b/34702983): Need to deprecate the bound iface from the HIDL interface.
Bug: 65671875
Test: Device boots up and connects to wifi networks.
Test: Will send for regression tests.
Change-Id: I33fef1332f2fe2da3f48ee87ef06660844699253
Removing whenever I see these in code reviews.
Test: none
Merged-In: I4322f533a837d55618ec2ed2125e8966ace9d61d
Change-Id: I4322f533a837d55618ec2ed2125e8966ace9d61d
The stop may fail if the interface down operation takes too long on some
devices. This is not fatal as long as we're able to bring up the
interfae for the next HAL start.
Bug: 68303248
Test: VTS tests
Change-Id: I93943f9d0cfb83cf07253c6f9b3866d6d137028f
Test: I solemnly swear I tested this conflict resolution.
Change-Id: I731eba80656edbf87a5d5ba3af63a2c607ac61a6
Merged-In: Id71e2348eabf5101d1fd286bbc0076154703d159
The NAN-specific VTS tests check whether the device supports NAN.
However, the tests did not abort early enough.
Bug: 68302113
Test: builds
Change-Id: Ie6fb0e2b0bd12840ce6bc52fab1b6d4e8c5a571e
This version of configstore was removed internally.
Test: I solemnly swear I tested this conflict resolution.
Change-Id: I589addff6aec7bb7a8a7938d75c51dcc56116a42
The current timeout of 50ms seems to be insufficient on some devices.
So, changing it to a higher value.
Bug: 66334487
Test: Ran the VTS tests continously in a loop for 30 minutes
without failure. Previously, failure occurence was about once every
5-10 minutes.
`while adb shell /data/VtsHalWifiSupplicantV1_0TargetTest; do :;
done; adb bugreport`
Change-Id: I884de64d748b46b9710a702f3c72cc30dd51627e
In preparation for new feature development - uprev the Wi-Fi Treble HAL
revision from 1.1 to 1.2.
This CL does not contain any functional changes to the HAL itself. Only
updating the revision.
Bug: 65647646
Test: builds and Wi-Fi runs
Change-Id: I8e80e866562a508bf481104a3dbcf2bd8f8015b4
To prevent property name collisions between properties of system and
vendor, 'vendor.' prefix must be added to a vendor HAL service name.
You can see the details in http://go/treble-sysprop-compatibility.
Test: succeeded building and tested on a walleye device
Bug: 36796459
Change-Id: I4e8fbee791ec917a8f627a1366f4d44ec7e6febc
IWifi::stop() is currently non-blocking which makes it hard for the
client to determing when the stop is fully complete. This for example
causes wificond to disable the wlan0 interface while the legacy HAL
stop is being processed. So, add a timed wait to let the legacy HAL
complete processing of the stop before we unblock the IWifi::stop()
HIDL call.
Bug: 64611487
Test: Manual tests by wifi state toggling and verifying the order of
events in logs:
08-15 19:17:53.302 796 796 I android.hardware.wifi@1.0-service:
Stopping legacy HAL
08-15 19:17:53.302 796 796 I WifiHAL : Sent msg on exit sock to
unblock poll()
08-15 19:17:53.302 796 4793 E CLD80211:
/vendor/bin/hw/android.hardware.wifi@1.0-service: Could not find group
host_logs, errno: 0 id: -2
08-15 19:17:53.302 796 4793 E CLD80211:
/vendor/bin/hw/android.hardware.wifi@1.0-service: Could not find group
fw_logs, errno: 0 id: -2
08-15 19:17:53.302 796 4793 E CLD80211:
/vendor/bin/hw/android.hardware.wifi@1.0-service: Could not find group
per_pkt_stats, errno: 0 id: -2
08-15 19:17:53.302 796 4793 E CLD80211:
/vendor/bin/hw/android.hardware.wifi@1.0-service: Could not find group
diag_events, errno: 0 id: -2
08-15 19:17:53.302 796 4793 E CLD80211:
/vendor/bin/hw/android.hardware.wifi@1.0-service: Could not find group
fatal_events, errno: 0 id: -2
08-15 19:17:53.302 796 4793 I CLD80211:
/vendor/bin/hw/android.hardware.wifi@1.0-service: Sent msg on exit sock
to unblock poll()
08-15 19:17:53.302 796 4793 I android.hardware.wifi@1.0-service:
Legacy HAL stop complete callback received
08-15 19:17:53.304 802 838 D CHRE : @ 151.328: [Platform] wifi:
has 0, enabled 0
08-15 19:17:53.321 796 4793 I android.hardware.wifi@1.0-service:
Legacy HAL event loop terminated
08-15 19:17:53.321 796 796 I android.hardware.wifi@1.0-service:
Legacy HAL stop complete
08-15 19:17:53.522 796 796 I android.hardware.wifi@1.0-service: Wifi
HAL stopped
Test: Will send for regression tests.
Change-Id: I394c11724e9459a4b9a6b970e2bcb4e0ad65fefc
IE elements should not be present in the batched scan results, but it
being there doesn't cause any problems. The IE elements will be ignored
when we send it over the HIDL interface, so reduce the log level to
prevent log spamming.
Bug: 64448186
Test: None
Change-Id: Ia0d0312a113c907bb0992bb8f41e734fe9d05114
Clarify meaning of maxNdpSessions capability. Total, not per interface.
Bug: 64685830
Test: builds
Change-Id: I24fc0a97172a4ca161f432b6c5c1dc55a44dfb12
This reverts commit 8ac1971678.
Reason for revert: Didn't remove automotive changes from this CL.
Merged-In: I8608c8f636c35f21e4246a805a9eff6d14124e0a
Change-Id: I1c660cffc8817ad0b33da9f6eceb3d88e7c48416
And use VtsHalTargetTestDefaults.
Bug: 64040096
Test: crop out all non-affected tests from vts-hal-hidl.xml and run
vts-tradefed run commandAndExit vts-hal-hidl --skip-all-system-status-check
--skip-preconditions
2 failures, same as without this change:
GatekeeperHidlTest.DeleteAllUsersTest
GatekeeperHidlTest.DeleteUserTest
Change-Id: I8f6995e9536a9aefe283ee3effec9f5a7f03b620
The tight loop of HAL start/stop exposes a race condition within the HAL.
Adding a fix for preventing this race would need a fix which would be
pretty risky at this point of the release. The issue itself is unlikely
to happen in real use cases because wifi toggles (user initiated or test
scripts initiated) via framework goes through a series of processing
before it invokes the HAL start/stop.
Bug: 64195190
Test: `make vts -j30 BUILD_GOOGLE_VTS=true TARGET_PRODUCT=aosp_arm64 &&
vts-tradefed run commandAndExit vts --skip-all-system-status-check
--primary-abi-only --skip-preconditions --module VtsHalWifiV1_0Target -l
INFO`
Change-Id: I4e4d65f8b6e2e423a3a5f26e5a97a78b7e99c3e8
When IWifi.stop() is invoked back to back (happens in the ConfigureChip
vts test), the HAL would return ERROR_NOT_AVAILABLE if the previous stop
is still being processed. This is not an error that needs to fail the test,
but a legitimate status for stop. We have a retry mechanism to handle
this in both the VTS test and framework for the case where IWifi.start()
is invoked while the previous stop is being processed.
While there, corrected a few log messages emitted by the HAL to debug
such startup/stop issues better.
Bug: 63971806
Test: `vts-tradefed run commandAndExit vts --skip-all-system-status-check
--primary-abi-only --skip-preconditions --module VtsHalWifiV1_0Target -l
INFO`
Change-Id: I5e3470ac97541a6ea10aceec9b737e5d03ed5206
