Include the host timespec only if we are building with
the host library for the POSIX architecture.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
Cache operations that act on unaligned buffers (with respect to cache
line) or buffers that are not multiple of the cache line size are at
least dangerous with an high risk of data and memory corruption.
While we do not enforce a contract on the buffer characteristics between
users and APIs, we can at least add a note in the documentation
specifying an undefined behaviour when the passed buffer is unaligned or
wrongly sized.
Signed-off-by: Carlo Caione <ccaione@baylibre.com>
Provides a way to use pinctrl to allow internal loopback
on a peripheral pin for testing purposes.
This is done by using output-enable on a input pin and
input-enable on a output pin.
Signed-off-by: Lucas Tamborrino <lucas.tamborrino@espressif.com>
This adds support for 32 double-precision registers in the context
switching of aarch32 architecture.
Signed-off-by: Huifeng Zhang <Huifeng.Zhang@arm.com>
This remanes send_buffer to tx_buffer to be
consistent with Rx buffer naming and with
mbox naming.
Signed-off-by: Kamil Gawor <Kamil.Gawor@nordicsemi.no>
This improves packet reception and fix an issue
where packets bigger than internal Rx buffer
were silently dropped. In current solution
local data coping for reception is not needed.
User gets direct pointer to shared memory which
allow to efficient receive as much data as was sent.
Signed-off-by: Kamil Gawor <Kamil.Gawor@nordicsemi.no>
This clarifies the spsc_buf_claim function documentation.
It should be easier now to understand what this function
does and what are returned memory properties.
Signed-off-by: Kamil Gawor <Kamil.Gawor@nordicsemi.no>
Adds an optional Kconfig that adds mutex locks to image management
group functions, this prevents collision between multiple threads
and/or transports.
Signed-off-by: Jamie McCrae <jamie.mccrae@nordicsemi.no>
The `net_if` layer should not be forcing `NET_IF_LOWER_UP` to be set on
all interfaces at runtime. Because `net_init` runs relatively late by
default, this is overriding any control the driver itself may have
performed using `net_if_carrier_on/off`.
Initialise this bit at compile-time instead. As the structure is already
being initialised, this doesn't increase any footprints.
Signed-off-by: Jordan Yates <jordan.yates@data61.csiro.au>
Add Nuvoton numaker series clock controller support, including:
1. Do system clock initialization in z_arm_platform_init().
2. Support peripheral clock control API equivalent to BSP
CLK_EnableModuleClock()/CLK_SetModuleClock().
Signed-off-by: cyliang tw <cyliang@nuvoton.com>
When writing string data to resources which are string types,
we should count in the terminating character into the data length.
Corner cases exist where LwM2M resource type is opaque but
lwm2m_get_string() or lwm2m_set_string() are used to read/write
the data. We must ensure string termination on those case, but
terminating character must not be stored in the engine buffer
or counted in the data length as this might be considered
as part of the binary data.
Fixes#59196
Signed-off-by: Seppo Takalo <seppo.takalo@nordicsemi.no>
Currently NET_ETH_MAX_HDR_SIZE is computed as the size of the
structure struct net_eth_hdr, but this does not take into account
the additional 4 bytes on the ethernet frame header in the presence
of VLAN, fix the issue using the size of struct net_eth_vlan_hdr
when CONFIG_NET_VLAN is enabled.
Signed-off-by: Luca Fancellu <luca.fancellu@arm.com>
Device dependencies are not always required, so make them optional via
CONFIG_DEVICE_DEPS. When enabled, the gen_device_deps script will run so
that dependencies are collected and part of the final image. Related
APIs will be also made available. Since device dependencies are used in
just a few places (power domains), disable the feature by default. When
not enabled, a second linking pass will not be required.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Rename the Kconfig option to be in line with recent renamings in device
handles/dependencies.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Rename the device dependencies array variable to use the "deps" name, in
line with latest renamings in device.h
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Rename struct device `handles` member to `deps`, in line with previous
renamings in the device API.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Rename multiple internal device macros to use the DEVICE_DEPS naming, so
that it is clear they belong to the device dependencies APIs.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
These macros are used internally by the device dependencies functions.
There's no need to expose them publicly, so prefix them with Z_ and add
them under INTERNAL_HIDDEN docs section.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Use the "device_deps" naming scheme to emphasize we are storing device
dependencies. The fact we are using device handles to store them is an
implementation detail.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Rename the gen_handles script and all of its references/associated files
to gen_device_deps. The new new makes things more clear, because the
script just take care of generating, for each device, an array of device
dependencies. While device handles are used internally to store this
information, it is in reality an implementation detail.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Note that only the the hardware round robin port arbitration capability
is being used.
Signed-off-by: Tomasz Bursztyka <tomasz.bursztyka@linux.intel.com>
The initial implementation was broken during improvements.
There was incorrect assumption that all pages are unmapped at
initials state. In reality at the beginning whole memory is
powered on, so we should mark all pages as mapped. Later in
initialization code unused pages are unmapped and if after this
some banks become empty (all pages unmapped), the power is
switched off.
Signed-off-by: Jaroslaw Stelter <Jaroslaw.Stelter@intel.com>
If the embedded application has a main(),
Zephyr requires it to have a signature with
C linkage (no C++ name mangling).
Otherwise Zephyr's init will not call it,
but will call the default weak stub.
But, when building with clang/llvm, when we
build calling the compiler with --frestanding
(for ex if CONFIG_MINIMAL_LIBC is set), the
compiler will mangle the main() symbol name.
This is not incorrect behavior from the compiler,
as, in principle, in a freestanding environment
main() does not have a special meaning.
gcc is still not mangling it when called with
--frestanding.
To avoid this issue, we define explicitly
the main linkage as extern C,
in a header the application will always include.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
This PR adds ability to build mesh with tf-m psa
for platforms those support tf-m.
Signed-off-by: Aleksandr Khromykh <aleksandr.khromykh@nordicsemi.no>
Added an MFD driver for the nPM1300.
This driver has register access helper functions that can be used
by subsystems. This will avoid each subsystem having to duplicate
the register access code.
Signed-off-by: Andy Sinclair <andy.sinclair@nordicsemi.no>
This commit adds the definition of sig_atomic_t to the posix/signal.h
file. The definition was copied from the file
modules/lib/picolibc/newlib/libc/include/signal.h. The problem of the
missing definition arose when I tried to compile a lua module for the
3.3.0 version of zephyr. Adding this definition solves the problem.
Signed-off-by: Christian Tenllado <ctenllado@gmail.com>
The function `z_nrf_rtc_timer_exact_set` is added to allow
setting compare channel without possible creeping of cc val.
Signed-off-by: Andrzej Kuroś <andrzej.kuros@nordicsemi.no>
The ACK procedure had the following issues:
- MAC commands were not acknowledged.
- When the package is a broadcast package the package must not be
acknowledged.
Signed-off-by: Florian Grandel <fgrandel@code-for-humans.de>
Acknowledgment is mandatory if legitimately requested by the package's
"ACK requested" flag. The L2 layer will have to ensure that compliant
ACK packages will always be sent out automatically as required by the
standard.
For IEEE 802.15.4 compliance, the NET_L2_IEEE802154_ACK_REPLY option is
therefore being deprecated.
Signed-off-by: Florian Grandel <fgrandel@code-for-humans.de>
The existing calls to ieee802154_radio_send() and soft MAC ACK handling
were inconsistent and/or not properly integrated with more recent
radio driver capabilities as CSMA/CA and ACK in hardware.
Signed-off-by: Florian Grandel <fgrandel@code-for-humans.de>
The method ieee802154_radio_handle_ack() does not belong to the
PHY/radio layer but to the L2 layer. It is a callback called from the
radio layer into the L2 layer and to be implemented by all L2 stacks.
This is the same pattern as is used for ieee802154_init(). The
'_radio_' infix in this function is therefore confusing and
conceptually wrong.
This change fixes the naming inconsistency and extensively documents
its rationale.
It is assumed that the change can be made without prior deprecation of the
existing method as in the rare cases where users have implemented custom
radio drivers these will break in obvious ways and can easily be fixed.
Nevertheless such a rename would not be justified on its own if it were
not for an important conceptual reason:
The renamed function represents a generic "inversion-of-control" pattern
which will become important in the TSCH context: It allows for clean
separation of concerns between the PHY/radio driver layer and the
MAC/L2 layer even in situations where the radio driver needs to be
involved for performance or deterministic timing reasons. This
"inversion-of-control" pattern can be applied to negotiate timing
sensitive reception and transmission windows, it let's the L2 layer
deterministically timestamp information elements just-in-time with
internal radio timer counter values, etc.
Signed-off-by: Florian Grandel <fgrandel@code-for-humans.de>
The CSMA/CA algorithm had multiple issues:
- Timing of backoff periods depends on the PHY's symbol rate and other
PHY-specific settings. We introduce a preliminary solution that works
with current drivers. A fully standard-compliant long-term solution
has already been conceptualized but requires further pre-conditions,
see #50336 (issuecomment-1251122582).
- We enforce the condition defined in the standard that macMinBe must be
less than or equal macMaxBe.
- According to the standard a CSMA/CA failure should lead to immediate
abortion of the transmission attempt, no matter how many
retransmissions have been configured.
- The number of retransmissions was off by one. It is now used as
defined in the standard algorithm.
- Retransmissions are only allowed when acknowledgement is requested in
the packet.
- prepare_for_ack() has side effects and must be called before each
retransmission.
We also replace variables by constants where possible.
The function was renamed as it represents unslotted CSMA/CA and does not
support other CSMA/CA modes. These may be introduced in the future.
Signed-off-by: Florian Grandel <fgrandel@code-for-humans.de>
This patch adds the ability to change virtual i2c eeprom target address
at runtime using a single function.
Added CONFIG_I2C_EEPROM_TARGET_RUNTIME_ADDR as an optional Kconfig.
Signed-off-by: Sahaj Sarup <sahaj.sarup@linaro.org>
This commit adds the following features related to virtual addresses
support:
- Allows to store Label UUIDs which virtual addresses collide;
- Allows to decrypt messages encrypted with a virtual address with
collision;
- Allows to publish a message to a specific Label UUID to avoid virtual
addresses collision by adding a pointer to Label UUID to
struct bt_mesh_msg_ctx and struct bt_mesh_model_pub;
- Allows to differentiate Label UUIDs in the model's Subscription List
by storing all subscribed UUIDs in struct bt_mesh_model.uuids field.
Signed-off-by: Pavel Vasilyev <pavel.vasilyev@nordicsemi.no>
If a user tries to enable TWT too early in the connection, then we might
enter TWT sleep even before DHCP is completed, this can result in packet
loss as when we wakeup we cannot receive traffic and completing DHCP
itself can take multiple intervals. Though static ip address can be
assigned too. Reject TWT till Wi-Fi interface has
a valid IP address.
Signed-off-by: Ajay Parida <ajay.parida@nordicsemi.no>
Add a new pinctrl driver for TI CC32XX SoC. The driver has not been
tested, just implemented following datasheet specs and checked that it
compiles. Consider this as a best-effort driver to remove custom pinmux
code in board files.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Some RISCV platforms shipping a CLIC have a peculiar interrupt ID
ordering / mapping.
According to the "Core-Local Interrupt Controller (CLIC) RISC-V
Privileged Architecture Extensions" Version 0.9-draft at paragraph 16.1
one of these ordering recommendations is "CLIC-mode interrupt-map for
systems retaining interrupt ID compatible with CLINT mode" that is
described how:
The CLINT-mode interrupts retain their interrupt ID in CLIC mode.
[...]
The existing CLINT software interrupt bits are primarily intended for
inter-hart interrupt signaling, and so are retained for that purpose.
[...]
CLIC interrupt inputs are allocated IDs beginning at interrupt ID
17. Any fast local interrupts that would have been connected at
interrupt ID 16 and above should now be mapped into corresponding
inputs of the CLIC.
That is a very convoluted way to say that interrupts 0 to 15 are
reserved for internal use and CLIC only controls interrupts reserved for
platform use (16 up to n + 16, where n is the maximum number of
interrupts supported).
Let's now take now into consideration this situation in the DT:
clic: interrupt-controller {
...
};
device0: some-device {
interrupt-parent = <&clic>;
interrupts = <0x1>;
...
};
and in the driver for device0:
IRQ_CONNECT(DT_IRQN(node), ...);
From the hardware prospective:
(1a) device0 is using the first IRQ line of the CLIC
(2a) the interrupt ID / exception code of the `MSTATUS` register
associated to this IRQ is 17, because the IDs 0 to 15 are reserved
From the software / Zephyr prospective:
(1b) Zephyr is installing the IRQ vector into the SW ISR table (and into
the IRQ vector table for DIRECT ISRs in case of CLIC vectored mode)
at index 0x1.
(2b) Zephyr is using the interrupt ID of the `MSTATUS` register to index
into the SW ISR table (or IRQ vector table)
It's now clear how (2a) and (2b) are in contrast with each other.
To fix this problem we have to take into account the offset introduced
by the reserved interrupts. To do so we introduce
CONFIG_RISCV_RESERVED_IRQ_ISR_TABLES_OFFSET as hidden option for the
platforms to set.
This Kconfig option is used to shift the interrupt numbers when
installing the IRQ vector into the SW ISR table and/or IRQ vector table.
So for example in the previous case and using
CONFIG_RISCV_RESERVED_IRQ_ISR_TABLES_OFFSET == 16, the IRQ vector
associated to the device0 would be correctly installed at index 17 (16 +
1), matching what is reported by the `MSTATUS` register.
CONFIG_NUM_IRQS must be increased accordingly.
Signed-off-by: Carlo Caione <ccaione@baylibre.com>
Currently, the L2 PPP subsystem is not using the network
interface subsystem appropriately. Here are the issues:
1. net_if_up hidden away internally in net L2 PPP
2. net_if_down not used at all...
3. net_if_carrier_on / off is not used, a workaround is
used instead, which results in duplicated code
4. L2 PPP does not listen for network events, instead
it needs the workaround callbacks from drivers.
5. The carrier_on workaround is delegated to a complex
and broken sys work queue item.
This commit fixes all above issues. net_if_up/down and
net_if_carrier_on/off now work as expected. workaround
for carrier_on/off has been removed.
Signed-off-by: Bjarki Arge Andreasen <baa@trackunit.com>
A mesh key type has been added to be able to choose the different
key representation for different security libraries.
The type as well as some functionality related to Mesh key
management has been added as a public API.
If tynicrypt is chosen then keys have representation
as 16 bytes array. If mbedTLS with PSA is used then keys are
the PSA key id. Raw value is not kept within BLE Mesh stack
for mbedTLS. Keys are imported into the security library
and key ids are gotten back. This refactoring has been done
for the network(including all derivated keys), application,
device, and session keys.
Signed-off-by: Aleksandr Khromykh <aleksandr.khromykh@nordicsemi.no>
This API introduced a new sensing subsystem, which is a high level
sensor framework inside the OS user space service layer.
It's a framework focus on sensor fusion, clients arbitration, sampling,
timing, scheduling and sensor based power management.
It's key concepts including physical sensor and virtual sensor objects,
and a scheduling framework reflects sensor objects' reporting relationship.
Physical sensors not depends on any other sensor objects for input, and
will directly interact with existing zephyr sensor device drivers.
Virtual sensors rely on other sensor objects (physical or virtual) as
report inputs.
The sensing subsystem relies on existing zephyr sensor device APIs or V2
zephyr sensor device APIs (#44098).
So it can leverage current existing zephyr sensor device drivers (100+).
And it's configurable, so, for some low cost IoT devices, may not need an
advanced sensor framework, but just need access some simple sensor
devices,can not configure and build this sensor subsystem, but just use
the exiting zephyr sensor device APIs to save memory resources.
Since the sensing subsystem is separated from device driver layer or
kernel space and could support various customizations and sensor
algorithms in user space with virtual sensor concepts. The existing
sensor device driver can focus on low layer device side works, can keep
simple as much as possible, just provide device HW abstraction and
operations etc. This is very good for system stability.
The sensing subsystem is decoupled with any sensor expose/transfer
protocols, the target is to support various up-layer frameworks and
Applications with different sensor expose/transfer protocols,
such as CHRE, HID sensors Applications, MQTT sensor Applications
according different products requirements. Or even support multiple
Applications with different up-layer sensor protocols at the same time
with it's multiple clients support design. For example can support CHRE
and other normal zephyr sensor application (can use HID etc) at
the same time.
Signed-off-by: Hebo Hu <hebo.hu@intel.com>
Signed-off-by: Guangfu Hu <guangfu.hu@intel.com>
The RSSI value in net_pkt (net_pkt_cb_ieee802154.rssi) was used
inconsistently across drivers. Some drivers did cast a signed dBm value
directly to net_pkt's unsigned byte value. Others were assigning the
negative value of the signed dBm value and again others were offsetting
and stretching the signed dBm value linearly onto the full unsigned byte
range.
This change standardizes net_pkt's rssi attribute to represent RSSI on
the RX path as an unsigned integer ranging from 0 (–174 dBm) to 254 (80
dBm) and lets 255 represent an "unknown RSSI" (IEEE 802.15.4-2020,
section 6.16.2.8). On the TX path the rssi attribute will always be
zero. Out-of-range values will be truncated to max/min values.
The change also introduces conversion functions to and from signed dBm
values and introduces these consistently to all existing call sites. The
"unknown RSSI" value is represented as INT16_MIN in this case.
In some cases drivers had to be changed to calculate dBm values from
internal hardware specific representations.
The conversion functions are fully covered by unit tests.
Fixes: #58494
Signed-off-by: Florian Grandel <fgrandel@code-for-humans.de>
