The HIDL interface now returns a |WifiStatus| instance to indicate the
status of any operation. This is replacing the existing asynchronous
delivery of success or failure (using |FailureReason| instance).
Rename the existing util class to |wifi_status_util| and add a couple of
methods to create a |WifiStatus| instance with empty description.
Bug: 32056230
Test: Compiles
Change-Id: I8488f7cd7d6ad6bd7a0c3c82a7ef83299d877d45
Add missing status returns for a few HIDL methods. This is to keep the
interface consistent.
Bug: 32146455
Test: update_makefile.sh
Change-Id: Ia66fe4d00e884e5ce1e1906db77bb5ffdaebffdd
These stats are used to determine if the device is being woken up
frequently by the wlan chipset. The stats structure describe all the
various resons for which the host was woken up by the wlan chipset
(firmware).
Bug: 32221997
Test: ./hardware/interfaces/update-makefiles.sh
Change-Id: Id52915348fef0283affcd834c6f1db5055e54ace
The packet fate debug mechanism is used to track the state of all the
packets transmitted/received during the association process.
Also,
Add the various debug capabilities in the respective interfaces.
All ring buffer, driver/firmware dump, etc related capabilities are in
IWifiChip object.
The packet fate capability is exposed in IWifiStaIface object.
Bug: 32221997
Test: ./hardware/interfaces/update-makefiles.sh
Change-Id: Ic6bf49d682b70a1cdcd18c69fc25a544921bd548
The debug ring buffers is a purely debug mechanism to let the driver
report debug info like connection events, power events, etc to the
framework.
The framework used to previously dump out the raw byte stream in
the bugreport and the vendors had some tools to parse out the data.
This is now being changed to provide the framework with the
internal ring buffer structs to ease parsing this data in framework
itself. This will eventually be used in the new wifilogd daemon.
Bug: 32221997
Test: ./hardware/interfaces/update-makefiles.sh
Change-Id: I2c90662cfa9d07ae6fc72198a286338dbaacbfc2
Some of these structs are used in the debug framework and hence needs to
live in types.hal. Add a |Sta| prefix to differentiate them from other
types similar to what is done for Nan and Rtt types.
Bug: 31991459
Test: ./hardware/interfaces/update-makefiles.sh
Change-Id: Idc49b28447609c4334b0a086481c404c06a63fea
Changes in the CL:
a) Unions are not allowed in the HIDL generated code that is going to be
used in java.
b) Few RTT constants were defined twice in the legacy rtt.h. One set
as enums, the other set as #define's. Remove the duplicated values in
the HIDL interface.
Bug: 31991076
Bug: 31991232
Test: mmm -j32 hardware/interfaces/wifi/1.0/
Change-Id: I03db70c7b89c0be53a7754ab8b34a19ad2c6e536
Interface is mostly a replication of the wifi_nan.h HAL header file.
1. All the methods are in |IWifiNanIface|.
2. Moved all the callbacks to |IWifiNanIfaceEventCallback.hal|.
2. Moved all the data types to |types.hal|.
3. Changed all the variable size arrays to vecs.
Bug: 31991076
Test: ./hardware/interfaces/update-makefiles.sh
Change-Id: I2af0b0003cf30a0f2bfdcb56b69c492cf831745b
Add HIDL methods for STA iface link layer stats.
While there,
Add the missing gscan API for fetching a list of channels for a given
band.
Bug:31991459
Test: Compiles
Change-Id: I5d448eb823faae4e8f5c25f746cf59e70df454cf
Having all the HIDL methods asynchronous, makes it hard for the
calling code to keep track of the operation status/result.
There are some operations which will generate asynchronous results (like
bgscan), convert all the other methods to synchronous methods.
The |EventCallback| objects will now just broadcast important events
(needed for other clients to listen for state changes). This will no
longer be used to send responses to every command sent to the HAL.
Bug: 32061909
Test: `./hardware/interfaces/update-makefiles.sh`
Change-Id: Id2433f4c8e028268dd027cdeb239ba4082b157b5
Add a status parameter for every HIDL interface object method which can
possibly become invalid. This should help inform the caller that the
object being used is stale/invalid now.
While there,
Rename |CommandFailureReson| to |FailureReasonCode|.
NOTE: |FailureReason| will continue to indicate any errors during the
processing of the command via the corresponding |onFailure| callback.
Bug: 32056230
Test: Compiles
Change-Id: I2ec5af3075221e483579410f344bcedd6bf17a93
Changes in the CL:
1. Add gscan/APF related API's to |IWifiStaIface|.
2. Add a new callback HIDL interface (|IWifiStaIfaceEventCallback)
for all callbacks received from |IWifiStaIface|.
Bug: 31991459
Test: Compiles
Change-Id: Id9f2ded9e20bee393ab53d84efa814d52704cd2c
Changes made to the .hal definition:
- introduce Effect ID returned by the IEffectsFactory that
needs to be passed to IStream.{add|remove}Effect; otherwise
it's impossible to retrieve the underlying HAL effect handle;
- change "bus address" in DeviceAddress to "string" type;
- fix signature of some methods w.r.t. returning Result;
- remove unused "struct AudioPatch".
Bug: 30222631
Test: make
Change-Id: Icb51729ef57bb2a5b0b78609735e7481bc04f95c
Initial commit for the context hub Hal. At this point it is
largely a refactoring of the older HAL with some new abstractions.
Test: Make completes successfully
Change-Id: Ieac4b20403e6e0b3966a4f5570fbbba7151e783b
When combining enum values with bit-wise operators, the result is no
longer of the enum type, so we need to cast it back.
Test: local compile
Change-Id: I3697d37b612f0e29aabb46a6a958ca4ef0fed64f
Test: Tested with stub HAL on angler
Bug: b/32754732
Change-Id: I54a5ac536e1f34cf2498972653aef702cd926b2a
Signed-off-by: Sandeep Patil <sspatil@google.com>
The default Health HAL implementation continues to depend on
libhealthd.<target> BOARD_HAL_STATIC_LIBRARY in order to make
sure healthd continues to work.
New device specific Health HAL implentations don't need to
compile with healthd STATIC_HAL.
Test: Tested healthd with and without the static hal implementation
on Angler.
BUG: b/32724915
Change-Id: I25d439f24a4178666614faf196423ffc8ccafafb
Signed-off-by: Sandeep Patil <sspatil@google.com>
HIDL interface to
a) Initiate ANQP queries & icon queries for hotspot 2.0
b) Add callbacks for the reception of WNM frames (hotspot
remediation, deauth imminent notice).
Bug: 31116047
Test: ./hardware/interfaces/update-makefiles.sh
Change-Id: Ib3353f04b493f50c6415d1f3475efa957d921540