Doxygen docs for a macro need to be placed together with the macro being
documented (unless using `@def`).
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Documentation for STM32*_PINMUX macros was incorrectly placed above the
macro, so it was in practice documenting the next macro (e.g.
STM32_MODE_SHIFT).
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Since Doxygen 1.9.5 @ref enum::FIELD references will not work. Use the
`#FIELD` notation instead. Since enums place all of its entries to the
global namespace, adding the enum only adds redundant information
Doxygen can figure out on its own.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
This refactors how the BAP broadcast source handles the
extended and periodic advertising.
First it removes the start and stop of the extended
advertising, and instead expects the application
(or upper layers) to do this.
Second it exposes API functions to get the
necessary advertising data from BAP (service data and
the BASE), which the upper layers will then also
be responsible for setting and updating.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
Introduce support for Pinctrl driver on NXP S32Z/E SoC's.
The NXP S32 pin controller is a singleton node responsible for
controlling the pin function selection and pin properties, based on the
pin node group approach. The pinmux configuration is encoded in a
32-bit value.
Each S32 SoC implementing Pinctrl must create a `pinctrl_soc.h` header
which define SoC-specific macros to initialize the pinctrl structure.
Signed-off-by: Manuel Arguelles <manuel.arguelles@nxp.com>
Introduce a Kconfig (MP_MAX_NUM_CPUS) and an api arch_num_cpus() to
allow for systems that might determine the number of CPUs available to
Zephyr at runtime.
CONFIG_MP_MAX_NUM_CPUS is intented to be use for any array initialization
and such that need to occur at build time. For most systems
arch_num_cpus() will just report the value of CONFIG_MP_MAX_NUM_CPUS.
The intent is to phase out CONFIG_NP_NUM_CPUS.
Signed-off-by: Kumar Gala <kumar.gala@intel.com>
The macro STRUCT_SECTION_ITERABLE can only store items in ROM if a
const specifier is added to the struct declaration.
Also it does not create the iterable section itself, but adds an
element to the section.
This commit clarifies the documentation.
Signed-off-by: Martin Jäger <martin@libre.solar>
This function exposes list pointer, so that it allows the user to modify
the internal list. This adds bt_audio_foreach_capability iterator finction
that can be used instead.
Signed-off-by: Mariusz Skamra <mariusz.skamra@codecoup.pl>
This change re-orders attributes in struct ieee802154_context to
completely optimize struct padding away. This is done in a way that does
not impact readability.
Signed-off-by: Florian Grandel <jerico.dev@gmail.com>
Several attributes in the ieee802154_context struct may potentially be
accessed from different threads and/or ISR context. Only some of these
attributes were properly guarded against race conditions.
This may not have been to problematic in the past but as other changes
in this PR introduce additional attributes and mutate several attributes
in a single atomic transaction, leaving such changes unprotected seems
dangerous.
This change therefore introduces systematic locking of the
ieee802154_context structure.
Signed-off-by: Florian Grandel <jerico.dev@gmail.com>
IEEE 802.15.4 short address support is incomplete in several places.
This change improves short address support without claiming to fix
it everywhere. Future iterations will have to continue where this change
leaves off.
The purpose of this change was to:
* use the short address returned by association responses,
* automatically bind IEEE 802.15.4 datagram sockets to the short
address if available,
* use the short address in outgoing packages where applicable,
* improve validation of association/disassociation frames,
* model association more closely to the spec by tying it to the
existence of a short address in the MAC PIB thereby removing
redundancy in the PIB (which makes race conditions less probable),
* keep both, the short and extended addresses, of the coordinator.
Signed-off-by: Florian Grandel <jerico.dev@gmail.com>
This change introduces an additional function into the API that allows
to test and clear a net_if flag atomically. A similar function already
exists to test and set a flag. So this change actually improves symmetry
of the API.
The change is required in later commits of this PR.
Signed-off-by: Florian Grandel <jerico.dev@gmail.com>
The IEEE 802.15.4 L2 code stores representation of attributes like
PAN id, short address and extended address in different encodings:
* big endian for extended address and CPU byte order for everything
else whenever such attributes enter user space (except for IP/socket
link layer addresses which are always big endian - even in case of
short addresses - to maintain POSIX compatibility).
* little endian for everything that is close to the radio driver as
IEEE 802.15.4 frames are little endian encoded.
Endianness was almost nowhere documented which led to several bugs and
inconsistencies where assignments of different byte order were not
converted (or sometimes converted, sometimes not).
This change documents endianness wherever possible within the realm of
the IEEE 802.15.4 L2 code. Conversion bugs and inconsistencies that were
revealed by the improved documentation will be fixed in a separate
commit.
Signed-off-by: Florian Grandel <jerico.dev@gmail.com>
To maintain POSIX compatibility, link-layer addresses are encoded in big
endian in the IP stack and socket API.
The intended endianness was however not documented everywhere which led
to bugs and inconsistencies. The IEEE 802.15.4 L2 stack, for example,
sometimes stores addresses in little endian in structures that intend
them to be stored in POSIX-compliant big endian byte order.
This change documents intended endianness within the realm of the
IP and sockets stack. Conversion bugs are fixed in a separate commit.
Signed-off-by: Florian Grandel <jerico.dev@gmail.com>
The IEEE 802.15.4 radio driver encodes attributes in:
* little endian for everything that is close to the protocol as
IEEE 802.15.4 frames are little endian encoded.
* mixed big and little endian in its configuration where extended
addresses are being represented. These inconsistencies are unfortunate
but cannot be easily fixed in a backwards compatible way so will be
left untouched in this change.
Endianness was almost nowhere documented which explains these
inconsistencies and led to several bugs where assignments of different
byte order are not converted (or sometimes converted, sometimes not).
This change documents intended endianness within the realm of the
IEEE 802.15.4 radio driver code. Conversion bugs are fixed in a separate
commit.
Signed-off-by: Florian Grandel <jerico.dev@gmail.com>
Add support for setting the target mode for a PMIC regulator. Some
regulators support multiple modes, each with distinctive voltage and
current configuration data. This function allows the consumer to switch
the PMIC into a new mode. The PMIC can then be configured to use a new
set of voltages.
Signed-off-by: Daniel DeGrasse <daniel.degrasse@nxp.com>
Update regulator binding definitions for PCA9420 to support mode setting
features. This will allow the 4 voltage modes present on the PCA9420 PMIC
to be utilized.
Note that bindings have also been updated to indicate the default
register value and target voltage for each regulator on the PCA9420.
Signed-off-by: Daniel DeGrasse <daniel.degrasse@nxp.com>
Some minor housekeeping prior to adding an http server
implementation. There are already a number of http headers
and that number will likely increase with subsequent work.
Moving them into a common directory cleans up the
`include/net` directory a bit.
Signed-off-by: Christopher Friedt <cfriedt@meta.com>
Previously, HTTP method enumerations were only defined within
the `http_parser.h`, which may not be ideal for all use cases.
This commit moves the `enum http_method` definition to a
dedicated header in a dedicated `http` subdirectory.
Signed-off-by: Christopher Friedt <cfriedt@meta.com>
The can_set_timing_data() API function is optional and is not supported by
all CAN controller drivers.
On a board with only either classic CAN or CAN-FD capable controllers the
current behavior is fine, but on boards with a mix of classic CAN and
CAN-FD capable controllers/drivers, calling can_set_timing_data() on a
classic CAN controller will lead to a NULL pointer dereference.
To ensure a proper behavior when calling can_set_timing_data() on classic
CAN controller drivers, change the API wrapper function to return -ENOSYS
if the CAN driver does not implement CAN-FD support.
Note that the data phase timing calculation functions are fine as-is, as
the timing calculations will fail due to minimum/maximum data phase timing
parameters being all zeros for drivers only supporting classic CAN.
Signed-off-by: Henrik Brix Andersen <hebad@vestas.com>
All init entries are already prepended with `__init_`, there's no point
in pre-pending 2 prefixes.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Use the `init_id`, in line with recent device changes, as the argument
is being used as a unique identifier.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
- Create a sys_init entry, as init API is not strictly related to the
device model (device model re-uses it)
- Hide internals (Z_*) from public API docs.
- Give overview within the group definition of allowed levels, meaning,
etc.
- Properly document struct init_entry.
- Multiple style issues
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
The _SYS_INIT_LEVEL* definitions were used to indicate the index entry
into the levels array defined in init.c (z_sys_init_run_level). init.c
uses this information internally, so there is no point in exposing this
in a public header. It has been replaced with an enum inside init.c. The
device shell was re-using the same defines to index its own array. This
is a fragile design, the shell needs to be responsible of its own data
indexing. A similar situation happened with some unit tests.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
The function in charge of calling all init function was defined in
device.c, had a public prototype and was just used in init.c. Since this
is really an internal function tied to Kernel init code, move it to
init.c and make it static, there's no need to expose it publicly.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
The `ARCH` init level was added to solve a specific problem, call init
code (SYS_INIT/devices) before `z_cstart` in the `intel_adsp` platform.
The documentation claims it runs before `z_cstart`, but this is only
true if the SoC/arch takes care of calling:
```c
z_sys_init_run_level(_SYS_INIT_LEVEL_ARCH);
```
Which is only true for `intel_adsp` nowadays. So in practice, we now
have a platform specific init level. This patch proposes to do things in
a slightly different way. First, level name is renamed to `EARLY`, to
emphasize it runs in the early stage of the boot process. Then, it is
handled by the Kernel (inside `z_cstart()` before calling
`arch_kernel_init()`). This means that any platform can now use this
level. For `intel_adsp`, there should be no changes, other than
`gcov_static_init()` will be called before (I assume this will allow to
obtain coverage for code called in EARLY?).
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Instead of hardcoding alignment size for pass 2 device handles, use
Z_DECL_ALIGN. This makes sure gen_handles.py is always in sync with the
type defined in device.h. The build assert in device.h can be removed as
a result, since we do not hardcode handles size anywhere else.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
The file does not use any of the API/types defined in kernel.h or
types.h. Added stdint.h include as we're using int32_t, and stddef.h for
NULL.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Some platforms, e.g. NXP LPC, pollute the namespace with definitions
like ARRAY_SIZE, MIN, MAX, etc. Since they don't re-define if already
defined, we're lucky enough to be able to "fix" this problem by
re-ordering includes. This likely deserves a proper fix, either by
patching offending code or Zephyr using a namespace. The "don't
re-define" technique makes implementation dependent on the include
order, i.e. poor/bad solution.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Some headers made use of types defined in sys_clock.h (e.g. k_timeout_t)
without including it.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
With the incoming removal of kernel.h/types.h from init.h, lots of files
start to show compile errors because they relied on indirect
definitions, including errno.h.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
It turns out we could end up with a circular include dependency through
the following sequences:
zephyr/arch/arm/aarch32/irq.h:19 -> zephyr/irq.h
zephyr/arch/arm/aarch32/arch.h:27
zephyr/arch/cpu.h:19
zephyr/sys/arch_interface.h:33
zephyr/timing/timing.h:10
or
zephyr/arch/arm/aarch32/asm_inline_gcc.h:23 -> zephyr/irq.h
zephyr/arch/arm/aarch32/asm_inline.h
zephyr/arch/arm/aarch32/irq.h:19
zephyr/arch/arm/aarch32/arch.h:27
zephyr/arch/cpu.h:19
zephyr/sys/arch_interface.h:33
zephyr/timing/timing.h:10
The problem is that both aarch32 irq.h/asm_inline_gcc.h include
zephyr/irq.h (which in turn depends on the arch specific header, as it
should be). So arch_irq_connect_dynamic ended up being used before its
declaration.
This patch removes zephyr/irq.h from the aarch32 headers, as it makes no
sense to have such _reverse dependency_.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Doxygen comment containing <tt>__VA_ARGS__</tt> does not render
correctly, i.e. __VA_ARGS__ is completely missing from the html
output. The Markdown syntax `__VA_ARGS__` renders as expected.
Signed-off-by: Tomasz Moń <tomasz.mon@nordicsemi.no>
We have cases where some devices needs to be initialized very early and
before c_start is call, i.e. to setup very early console or to setup
memory. Traditionally this would be hardcoded as part of the soc layer
and not using device model or the init levels.
This patch adds a new level ARCH, which will be called in early
architecture code and before we jump to the kernel code.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
Allows filling up struct lwm2m_obj_path by a macro.
For example:
struct lwm2m_obj_path p1 = LWM2M_OBJ(MY_OBJ);
struct lwm2m_obj_path p2 = LWM2M_OBJ(MY_OBJ, 0, RESOURCE);
Similarly, some function calls accept the structure, so it can
be initialized from stack and given by a pointer
lwm2m_notify_observer_path(&LWM2M_OBJ(obj_id, 0, RESOURCE_ID));
Signed-off-by: Seppo Takalo <seppo.takalo@nordicsemi.no>
Bring back Kconfig option which was previously removed by
9cd5086407 (previously named LOG_MSG_PKG_ALWAYS_ADD_RO_STRING_IDXS).
Renamed to shorten and be more descriptive. It can be useful for
external backends or frontends.
Signed-off-by: Krzysztof Chruscinski <krzysztof.chruscinski@nordicsemi.no>
They were basically using the wrong instructions to traverse the cache
by index: `dhi`, `dhwb` and `dhwbi` instead of `dii`, `diwb` and `diwbi`
variants.
Fixes#49112.
Signed-off-by: Ederson de Souza <ederson.desouza@intel.com>
Follow up to e1e16640b5 adding model name
helpers to access generated macros based on a compatible's matching
entries in vendor prefixes.
Signed-off-by: Maureen Helm <maureen.helm@intel.com>
