Deprecate the xoroshiro128+ PRNG algorithm in favour of xoshiro128++.
xoshiro128++ is a drop-in replacement which is invisible from the user
perspective.
xoroshiro128+ is unsuitable because it is explicitly a floating-point
PRNG, not a general-purpose PRNG. This means that the lower 4 bits of
the output are actually linear, not random (from the designers,
https://prng.di.unimi.it/). This means 1/8th of the generated data is
not random.
Additionally, xoroshiro128+ is not a 32bit algorithm, it operates on
64bit numbers. For the vast majority of Zephyr devices, this makes the
PRNG slower than it needs to be. The replacement (xoshiro128++) is
32bit, with no loss in state space (still 128 bit).
Signed-off-by: Jordan Yates <jordan.yates@data61.csiro.au>
Adds an implementation of xoshiro128++ as a pseudo random number
generator from https://prng.di.unimi.it/ that operates on 32bit words.
The algorithm postfix signifies the main operation in the generation
function. Therefore xoshiro++ is chosen over xoshiro** as we would
prefer to do 2 additions isntead of 2 multiplications on embedded
hardware. The quality of the generators appears to be the same in all
other respects.
xoshiro+ is not chosen despite being faster as it generates random
floating-point values, not general purpose random values (The lower 4
bits are linear).
Signed-off-by: Jordan Yates <jordan.yates@data61.csiro.au>
ZFD_IOCTL_POLL_OFFLOAD operation needed special handling, as it needed
to modify the fds table for the offloaded implementation, overwriting
websocket file descriptors with the underlying offloaded ones. This is
only needed for the offloaded sockets, as the native implmentation use
POLL_PREPARE/UPDATE operations instead.
Signed-off-by: Robert Lubos <robert.lubos@nordicsemi.no>
The websocket implementation of ioctl wrongly passed websocket context
to the undrelying `ioctl` implementation instead of the context of the
underlying socket.
Additionally, currentl implementation used the vtable of the native
socket implementation unconditionally, making it unusable with an
offloaded underlying socket.
Signed-off-by: Robert Lubos <robert.lubos@nordicsemi.no>
Add support for FORCE_SUSPEND and LOW_POWER actions. Even though these
actions have no clear meaning, they are added for compatibility until
their associated states are discussed. Their usage in new code should be
discouraged until the associated states are clarified.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
If a device does not support, for example, the suspend action the
subsystem should just ignore the device.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Instead of passing target states, use actions for device PM control.
Actions represent better the meaning of the callback argument.
Furthermore, they are more future proof as they can be suitable for
other PM actions that have no direct mapping to a state. If we compare
with Linux, we could have a multi-stage suspend/resume. Such scenario
would not have a good mapping when using target states.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
The verb tense for the suspended state was not consistent with other
states. The likely reason: state was being used as a command/action.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
If the device is already at the given state, do not call the device PM
control function. This makes sure that devices are only called to change
from one state to another.
Even though asynchronous device PM is completely broken, transitional
states are considered too.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Improve the docstrings of the pm_device_state_set function and change to
shorter argument variable names.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Since the state is no longer modified by the device PM callback, just
use the state value.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
The device PM subsystem already holds the device state, so there is no
need to keep duplicates inside the device. The pm_device_state_get has
been refactored to just return the device state. Note that this is still
not safe, but the same applied to the previous implementation. This
problem will be addressed later.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Fix the received Extended Advertising PDU's address
resolution latency due to missing reset of pending IRQ in
the NVIC which caused the wait loop from not exit correctly.
Fixes#30243.
Signed-off-by: Vinayak Kariappa Chettimada <vich@nordicsemi.no>
When generating LE Extended Advertising Report the device
address can be either in primary or auxiliary channel PDU,
do not overwrite resolving list index if present in primary
channel PDU.
Signed-off-by: Vinayak Kariappa Chettimada <vich@nordicsemi.no>
Cleanup standard device request handler.
Pass pointer to setup packet as argument where
it is reasonable.
Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>
Cleanup standard device request handler.
Pass pointer to setup packet as argument where
it is reasonable.
Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>
This class does not handle any vendor request,
therefore just return -ENOTSUP on any vendor request.
Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>
HCI commands are always directed to controller and should be
host-to-device class control requests.
Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>
Return ENOTSUP if request is unsupported and pass on the error
from encapsulated command/response functions.
Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>
Control request handler from CDC ECM implementation does not
care at all about returning errors when a request is not supported.
Return EINVAL/ENOTSUP if request is incorrect/unsupported.
Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>
Set data stage length variable to zero as a precaution
so that no trouble happens if control request handler
does not check the request values sufficiently.
Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>
The USB configuration option is actually a global switch
to enable USB drivers in general, but currently only
the device controller drivers are meant.
USB device controller drivers also have USB_DEVICE_DRIVER option.
Thus the option USB is actually redundant and can be replaced
by the self-explanatory option USB_DEVICE_DRIVER.
The name USB itself is not unique and should not be used as an
configuration option.
With these changes the option USB_DEVICE_DRIVER generally
enables USB device controller drivers. The option USB_DEVICE_STACK
enables USB device support. It is sufficient to enable only option
USB_DEVICE_STACK because it selects USB_DEVICE_DRIVER.
CONFIG_USB Kconfig option is temporary added to subsys/usb/Kconfig.
This is necessary to pass CI and will be removed again
when the USB configuration has been adapted in modules.
Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>
It is mainly for testing purposes where logging is multiple times
reinitialized.
Signed-off-by: Krzysztof Chruscinski <krzysztof.chruscinski@nordicsemi.no>
Fix key distribution not calling pairing complete callback when
distributing encryption information when not bonding.
Encryption information is only distributed in legacy pairing so issue
was only seen during legacy pairing.
Signed-off-by: Joakim Andersson <joakim.andersson@nordicsemi.no>
A mutex is used to syncrhonize the start, stop and service() functions
of the RD client. Previously it could happen that while service() was
working on e.g. bootstrapping, a stop() call by another thread would
close the socket. Then the bootstrapping process would detect it as a
network error, and restart the process.
Fixes#37170.
Signed-off-by: Maik Vermeulen <maik.vermeulen@innotractor.com>
The existing extension tree does not support all the features that are
defined by the specification (e.g. multiple parents).
This patch approaches this problem by defining a circular single-linked
list of extension models. So for a given model, all models that are on
the same list as that model are in some extension relationship with that
model. All models on a list represent a single connected component of an
extension graph but without defining specific relationships between each
pair of models. This list is used to manage a shared subscription list
as per the Mesh Profile Specification:
```4.2.4 Subscription List
Within an element, each model has a separate instance of a Subscription
List, unless the model extends another model on that element. Instances
of models that extend other models (i.e., all models within an extension
relation tree) shall share a single instance of a Subscription List per
element.
```
Signed-off-by: Michał Narajowski <michal.narajowski@codecoup.pl>
The Bluetooth Core spec 5.2 requires to provide CTE type in
periodic advertising report.
The implementation of le_per_adv_sync_report assigned
value zero to be default CTE type. The assigned value
in Bluetooth Core spec 5.2 means "AoA Constant Tone
Extension". In case there is no CTE attached to periodic
advertising PDU, CTE type should contain value 0xFF.
This commit fixes the issue.
Signed-off-by: Piotr Pryga <piotr.pryga@nordicsemi.no>
Fix compilation warning in hci_df_set_cl_cte_rx_enable function.
The struct bt_hci_cp_le_set_cl_cte_sampling_enable *cp variable
was used just to evaluate size in bt_hci_cmd_create call.
The variable is moved to prepare_cte_rx_enable_cmd_params
function and size of the buf is evaluated directly from type.
Code is simpler and prepare_cte_rx_enable_cmd_params does not
require passing additional argument.
Signed-off-by: Piotr Pryga <piotr.pryga@nordicsemi.no>
It is desirable to have the read and write like functions follow the
POSIX read and write parameter types.
Signed-off-by: Abe Kohandel <abe.kohandel@gmail.com>
There were missing calls to periodic advertising functions:
lll_adv_sync_init() and lll_adv_sync_reset() corresponding
functions in lll_adv.c.
Lack of those calls didn't introduce any harm.
I've added them to have complete init and reset patch,
similar to extended advertising implementation.
Signed-off-by: Piotr Pryga <piotr.pryga@nordicsemi.no>
There is missing handling of periodic advertising reset for
HCI_Reset command. That makes impossible to execute e.g.
qualification tests without rebooting of the DUT
for periodic advertising and direction finding.
This commit adds missing implementation.
Signed-off-by: Piotr Pryga <piotr.pryga@nordicsemi.no>
The local function has been missing static specifier.
The commit also removes unused STORAGE_MGMT_ID_ERASE define.
Signed-off-by: Dominik Ermel <dominik.ermel@nordicsemi.no>
In the current code the naming of the
CONFIG_RPMSG_MULTI_INSTANCE_?_IPM_{TX,RX}_NAME symbol is 1-based. While
this is not currently an issue, it could easily become such if the
symbol is programmatically used as part of a preprocessor enumeration
(for example when using DT_INST_FOREACH_STATUS_OKAY(...) & co).
To avoid trouble, just make the index starting from 0 instead than 1.
Signed-off-by: Carlo Caione <ccaione@baylibre.com>
Extract functions which are managing logging sources
and backends into separate file: log_mgmt.
So far those functions were in log_core mixed with functions
specific to log message creation and log processing.
Signed-off-by: Krzysztof Chruscinski <krzysztof.chruscinski@nordicsemi.no>
