Fix host RX thread being deadlocked. The deadlock occurs because the
RX thread is stuck waiting in conn_tx_alloc with K_FOREVER but if the
connection is disconnected only the RX thread can unblock it in the
handling of the disconnect event.
This commit fixes this deadlock by splitting the processing of the
disconnected event into two parts.
The part needed to unblock the RX is to release resources held by
unack'ed TX packets and mark the connection state as not connected
anymore.
The RX thread waiting for free_tx fifo and the TX thread waiting for
the bt_dev.le.pkts semaphore will both check the connected state after
having acquired them and will abort if disconnected.
The rest of the processing will be handled at normal RX thread
priority like normal.
Move the bt_recv_prio handling to the Bluetooth host when the host
has defined its own RX thread (CONFIG_BT_RECV_IS_RX_THREAD=n).
If the HCI driver has the RX thread (CONFIG_BT_RECV_IS_RX_THREAD=y),
then the responsibility to call bt_recv and bt_recv_prio correctly
falls to the HCI driver.
The helper function bt_hci_evt_is_prio() is replaced with
bt_hci_evt_get_flags() so that the HCI driver can do this correctly.
This decision to replace was made so that existing HCI drivers
maintained out-of-tree will fail at compile time with the new system.
Signed-off-by: Joakim Andersson <joakim.andersson@nordicsemi.no>
Signed-off-by: Carles Cufi <carles.cufi@nordicsemi.no>
Bluetooth: host: Move bt_recv_prio to host when RX thread is defined
Signed-off-by: Joakim Andersson <joakim.andersson@nordicsemi.no>
Use the NRF RNG entropy device as the entropy device for bt_rand and
controller internal functions when LLL is Nordic.
Using an entropy source with a significant increase in stack usage
will invalidate all stack size configurations in the system and lead
to stack overflow issues.
Signed-off-by: Joakim Andersson <joakim.andersson@nordicsemi.no>
For some reason GCC 9.x doesn't seem to be able to realize that set_num
will always be > 0 and so the local variable status will always be
initialized. Use a do {} while instead to keep it happy.
Signed-off-by: Carles Cufi <carles.cufi@nordicsemi.no>
Review rework, added comments, TODOs, FIXMEs and converted
magic number use in advertising set state flags to defines.
Signed-off-by: Vinayak Kariappa Chettimada <vich@nordicsemi.no>
Added comments and code to make note that disabling all
advertising sets is not yet implemented.
Signed-off-by: Vinayak Kariappa Chettimada <vich@nordicsemi.no>
Refactor pdu.h defined in anticipation of reused in
Advertising Extensions implementation. To reuse in
calculating ticks_slots etc.
Signed-off-by: Vinayak Kariappa Chettimada <vich@nordicsemi.no>
Integrate the generated extended adv report PDU list into
HCI layer to generate the HCI LE Extended Advertising Report
event.
Signed-off-by: Vinayak Kariappa Chettimada <vich@nordicsemi.no>
When Privacy is not supported in the build and Advertising
Extensions is enabled, HCI LE Enhanced Connection Complete
event was missing. This is fixed now.
Signed-off-by: Vinayak Kariappa Chettimada <vich@nordicsemi.no>
Added implementation to generate LE Extended Advertising
Report for received legacy PDUs.
Signed-off-by: Vinayak Kariappa Chettimada <vich@nordicsemi.no>
Differentiate call to ll_adv_params_set, whether used by
legacy LE Set Advertising Parameters or LE Set Extended
Advertising Parameters.
Signed-off-by: Vinayak Kariappa Chettimada <vich@nordicsemi.no>
Added debug information print outs in HCI code related to
Advertising Extensions events.
Signed-off-by: Vinayak Kariappa Chettimada <vich@nordicsemi.no>
Split BT_CTLR_CONN_RSSI option into two, the base option enables the
Read RSSI command, while the new BT_CTLR_CONN_RSSI_EVENT enables the
connection RSSI events. There is no handling of RSSI events, only a
BT_INFO log.
Signed-off-by: Joakim Andersson <joakim.andersson@nordicsemi.no>
Set the local RPA field of the enhanced connection complete event only
when the address was generated by the controller. If the host has
set an RPA and this one was used the controller should return all
zeroes.
Signed-off-by: Joakim Andersson <joakim.andersson@nordicsemi.no>
When receiving an unknown command send the unknown command response
as a command status event instead of a command complete event.
A command complete event has no status field, although by convention
all command parameters has a status field as the first parameter the
command status event seems like the more fitting option.
Signed-off-by: Joakim Andersson <joakim.andersson@nordicsemi.no>
The LE Connection Complete HCI event, unlike its Enhanced counterpart,
only uses 0x0 and 0x1 for Peer Address Type. Fix it so that it reflects
the specification correctly.
Signed-off-by: Carles Cufi <carles.cufi@nordicsemi.no>
Deprecate BT_LE_ADV defines in hci.h that are expected to be used by the
application in the scan received callback to identify the advertising
PDU type. These defines are mixing HCI input parameters and advertising
PDU types. Internally it is acceptable to mix these, but at the API we
should to mix in them.
Signed-off-by: Joakim Andersson <joakim.andersson@nordicsemi.no>
Add header definition for bt_read_static_addr function. Declaring it
without a header definition will not give any compilation error when
function definition changes.
Refactor nRF SoC specific code into nRF specific source files and
provide weak definitions when these are not implemented. This will make
it easier to add handlers per vendor.
Signed-off-by: Joakim Andersson <joakim.andersson@nordicsemi.no>
Fix infinite recursion in host-based bt_rand function. This would call
HCI LE Random Number command, which would in turn call bt_rand, causing
an infinite recursion.
bt_rand -> prng_reseed -> BT_HCI_OP_LE_RAND -> le_rand -> bt_rand
To solve this issue the controller should avoid doing calls into the
host, so all calls to bt_rand in the controller should be replaced with
a call to a controller function.
Fixes#22202
Signed-off-by: Joakim Andersson <joakim.andersson@nordicsemi.no>
This commit reverts the change that moved the remote version event from
a priority event to a normal event. This is done because the strategy
for using this event has been changed and will be used with a callback
instead of a semaphore that could be locked from the RX thread.
This commit retains the infrastructure that was added in the controller
so that moving events to priority processing is still possible.
Signed-off-by: Joakim Andersson <joakim.andersson@nordicsemi.no>
Fix multiple issues related to the way the host handles
Identity Information related to privacy
1. If the controller provided a public address the IRK
for this identity would be randomly generated but not
stored persistenly.
2. Fix the handling of the above issue which was fixed
for the random address but would initiate settings save
ID on every boot.
3. Fix the host not using the Vendor Specific HCI commands
related to retrieving the Identity Root (IR) from the
controller and using the key diversified function d1
to generate an IRK as specified in the BT Core spec.
Make sure that a Host generated ID is only saved when it is first
generated.
Signed-off-by: Joakim Andersson <joakim.andersson@nordicsemi.no>
Return the IR defined in FICR as the Identity Root for the static
address through the read static addresses command instead of providing
it through the Read Key Hierarchy Root command.
This is following the recommendations in the Zephyr HCI extension
document in doc/reference/bluetooth/hci.txt
Signed-off-by: Joakim Andersson <joakim.andersson@nordicsemi.no>
This commit targets solving issue #17731 over the LL_SW_SPLIT
arch of the BLE stack in Zephyr. This functionality is exposed
to the user as HCI Zephyr Command extensions
- BT_HCI_OP_VS_WRITE_TX_POWER_LEVEL
- BT_HCI_OP_VS_READ_TX_POWER_LEVEL
which enable Tx power read/write operations within BLE radio events
on a per role/connection basis.
The functionality is enabled upon the Kconfig advanced configuration
triggered by
- BT_CTLR_TX_PWR_DYNAMIC_CONTROL
depending on the enablement of Zephyr HCI vendor-specific command
extensions.
Necessary low-level radio HAL functionality and power definitions
are also supplied to address the high-level functionality of
controlling the Tx power.
Signed-off-by: Andrei Stoica <stoica.razvan.andrei@gmail.com>
Handle invalid ACL flags in HCI transport.
Only Point to Point is supported over HCI in both directions.
Fix flushable start HCI ACL packets not allowed on LE-U connections
from Host to controller.
Signed-off-by: Joakim Andersson <joakim.andersson@nordicsemi.no>
The remote version information event needs to be processed by the
prio_recv_thread() thread in order to unblock the Host RX thread
(effectively hci_driver's recv_thread()) when it blocks waiting for a
response to a remote version information.
Add the same time gate the inclusion of the feature behind a new Kconfig
option: CONFIG_BT_REMOTE_INFO.
Signed-off-by: Carles Cufi <carles.cufi@nordicsemi.no>
In order to be able to distinguish between connection-related events
that are generated by the controller and others genrated by LL control
procedures, introduce a new class for LLCP.
Signed-off-by: Carles Cufi <carles.cufi@nordicsemi.no>
Cache the result of calling hci_get_class() to avoid repeatedly invoking
it on the same data. In order to cache it we take advantage of the fact
that both radio_pdu_node_rx_hdr and node_rx_hdr are not packed
structures and they currently have a spare padding byte (between type
and handle).
Signed-off-by: Carles Cufi <carles.cufi@nordicsemi.no>
Hide the details of obtaining a pointer to the PDU data from a node_rx
structure to simplify the code.
Signed-off-by: Carles Cufi <carles.cufi@nordicsemi.no>
Enable use of settings system in controller and introduce settings for
company_id and subversion_number.
Signed-off-by: Wolfgang Puffitsch <wopu@demant.com>
