Fixing kernel crash caused by memory release
while having a scheduled work item pending.
Signed-off-by: Jeroen van Dooren <jeroen.van.dooren@nobleo.nl>
This polishes the doxygen doc on I3C API to make it,
hopefully, more usable. Fixed some typos too.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
The wording on deprecating arch_kernel_init() in favor of prep_c()
has never been materialized. Various architectures are using it to
perform initialization. So remove the wording.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
When compiling conn.c using arm-none-eabi-gcc version 11.3.1 20220712
with the -Wmaybe-uninitialized flag a warning is emitted due to
pending_no_cb not being initialized. I'm not sure if initializing it to
NULL is the "correct" fix, but it's certainly not any worse then it being
uninitialized, and it fixes the warning.
Signed-off-by: Dennis Grijalva <dennisgrijalva@meta.com>
This commit fixes the following issues with the PCM data format output
by the MCUX SAI driver:
- WS signal should be only one clock cycle in length for short PCM
format
- Word count should not be fixed to 2, except for classic I2S format
- BCLK polarity should be on falling edge for PCM long and short format
Additionally, the I2S_FMT_CLK_ constants now flip the frame and bit
clock polarity from the normal value expected for the selected I2S
format, as expected by the API.
Fixes#63041
Signed-off-by: Daniel DeGrasse <daniel.degrasse@nxp.com>
Update documentation for core clock frequency on LPC55xxx boards, to
indicate that they are clocked from PLL1 at 150MHz by default. Add a
note about the reduction to 96MHz required when using the flash
controller.
Signed-off-by: Daniel DeGrasse <daniel.degrasse@nxp.com>
Do not clock the LPC55xxx cores from PLL1 when CONFIG_FLASH is set. This
is required due to the following limitation of the flash controller
(documented in the reference manual):
Flash operations (erase, blank check, program) and reading a single word
can only be performed for CPU frequencies of up to 100 MHz. These
operations cannot be performed for frequencies above 100 MHz.
The PLL1 clock source will result in a core clock of 150MHz, which
violates this requirement.
Fixes#62963
Signed-off-by: Daniel DeGrasse <daniel.degrasse@nxp.com>
This sets the stack size appropriate for SPARC. Tested on the board
configurations qemu_leon3 and gr716a_mini.
Signed-off-by: Martin Åberg <martin.aberg@gaisler.com>
Specify only the bus corresponding to the current yaml file, as done in
the adi,adxl372-spi.yaml.
Signed-off-by: Antoniu Miclaus <antoniu.miclaus@analog.com>
When upgrading to version 3.4, our hardware lost ethernet connectivity.
Our hardware assumes a fixed link and does not communicate with the
underlying phy via the mdio bus.
I confirmed that the atsame54_xpro board also will lose ethernet
functionality when the atsame54_xpro configures the phy via to use a
fixed link...
The eth_sam_gmac driver changed the initialization behavior to call
net_if_carrier_off and notes to wait until phy link is up (via
callback.) However, when in a fixed link configuration, the callback is
never called. So the net_if_carrier_on event never occurs.
This patch adds a check to see if link is up already before calling
net_if_carrior_off. This check works because in fixed-link mode, link-up
is set synchronously during phy driver initialization.
I tested that atsame54_xpro with fixed-link configuration will now work
after this patch.
Signed-off-by: Thomas Chiantia <thomas.chiantia@gmail.com>
The commands showing west build had -b None which isn't quite right. Add
a :board: property with the tdk_robokit1 as its what I tested on myself.
Also noted that llext isn't supported on every architecture but
intentionally left this a bit vague as updating this little comment
every time something is added seems like a recipe for doc rot.
Signed-off-by: Tom Burdick <thomas.burdick@intel.com>
When the timer frequency is known at compile time, make sure we can use any
time conversion macro as a global initializer.
Signed-off-by: Keith Packard <keithp@keithp.com>
To be used in a global initializer, this must be a macro instead of a
static inline when the return value is constant.
Signed-off-by: Keith Packard <keithp@keithp.com>
clang emits a warning for this code:
foo.c:1:25: warning: division by zero is undefined [-Wdivision-by-zero]
int i = (10 > 100 ? (20 / (10 / 100)) : (20 * (100 / 10)));
^ ~~~~~~~~~~
The warning is generated for code whose value does not affect the
expression result, but technically it is still 'undefined behavior'.
Work around this warning by checking for a zero divisor and substituting
one to make the compiler happy.
Signed-off-by: Keith Packard <keithp@keithp.com>
Connection states are non-trivial to understand based on their names
only, so add short descriptions in the header file explaining what they
do.
Signed-off-by: Carles Cufi <carles.cufi@nordicsemi.no>
Remove excessive debug logging from the SJA1000 driver backend. Logging
each register access makes generic CAN debug logging unusable.
Signed-off-by: Henrik Brix Andersen <hebad@vestas.com>
Fix errenous return value in pacs_gatt_notify function. The function
would always return zero, and not forward the error correctly.
Signed-off-by: Fredrik Danebjer <frdn@demant.com>
Add padding to the header and remove unnecessary memset in order to fix
alignment faults in cores such as M0 or ones that support
CONFIG_TRAP_UNALIGNED_ACCESS
Signed-off-by: Yuval Peress <peress@google.com>
CONFIG_HAS_MCUX is also selected by ARM64 and XTENSA-based cores
which don't have any CMSIS files. As such, it's wrong to implicitly
select CONFIG_HAS_CMSIS_CORE when CONFIG_HAS_MCUX is selected. Since
the ARM32-based cores implicitly select CONFIG_HAS_CMSIS_CORE there's
no point in CONFIG_HAS_MCUX also selecting it too.
The old approach caused config-related warnings
in external projects such as SOF after the introduction of
Commit '8d5ed53' ("modules: remove uncessary source for
external cmsis kconfig"). The warnings were in the form of:
"HAS_CMSIS_CORE (defined at modules/cmsis/Kconfig:7) has
direct dependencies 0 with value n ..."
Signed-off-by: Laurentiu Mihalcea <laurentiu.mihalcea@nxp.com>
Don't reset values set in init callback as it is called only once by
bt_mesh_init call. The reset callback is called on every node reset.
Signed-off-by: Pavel Vasilyev <pavel.vasilyev@nordicsemi.no>
DFU/SR/FD/BV-59-C tests Distributor self-update. Because the actuall
firmware swapping doesn't happen, we need to just simulate the swapping
by setting new Firmware ID sent by LT. We use slot to take firmware id
as by procedure, Fw Update Server doesn't know about it.
Signed-off-by: Pavel Vasilyev <pavel.vasilyev@nordicsemi.no>
Fixes: #63610
Fix a variable spelling mistake which caused a required dependency to
not be set.
Together with this fix, then remove an unneeded CMake cache update.
The `ZEPHYR_GIT_INDEX-NOTFOUND` value is not really useful in this
case as a not set var will also ensure that Zephyr is checked to be
a git repo or not.
And when detected as a git repo, then `ZEPHYR_GIT_INDEX` is set
to correct value.
Signed-off-by: Torsten Rasmussen <Torsten.Rasmussen@nordicsemi.no>
The rpmsg_close() call uses the HCI reset command to reset the
controller. But when building as controller-only we do not bring in the
infrastructure to send HCI commands (nor should we) and rpmsg_close()
will not be called anyway.
Fixes#63534.
Signed-off-by: Carles Cufi <carles.cufi@nordicsemi.no>
A few of the length checks that deal with HCI packets coming from the
controller were using assert statements. But the recommended practice is
to drop invalid packets and continue execution whenever a malformed
packet arrives from an external source, so replace those assert
statements with branches that will drop the packet and return.
Signed-off-by: Carles Cufi <carles.cufi@nordicsemi.no>
When using multiple streams the seq_num provided to the
controller would be incorrect as we used a single global
seq_num rather than a proper seq_num per stream.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
Fix the correct path to net-tools directory.
Also default to Bash instead of /bin/sh as we use
Bash syntax.
Signed-off-by: Seppo Takalo <seppo.takalo@nordicsemi.no>
The code is designed to handle RF0L and RF1L in
line 1, but they were being sent to line 0. Becuase
they weren't handled, the interrupts would never
be handled which locked up the chip.
Signed-off-by: Abram Early <abram.early@gmail.com>
Fixed issue when reprovisioning is done on a device with
both RPR client and server on the same device.
Signed-off-by: Ingar Kulbrandstad <ingar.kulbrandstad@nordicsemi.no>
Update babblesim tests for testing chaining for advertising to ensure that
we spill over to the next PDU
Signed-off-by: Andries Kruithof <andries.kruithof@nordicsemi.no>
The existing formula for the amount of data that can be filled
in in the current PDU is correct but confusing. Replace it with
a simpler formula.
Signed-off-by: Andries Kruithof <andries.kruithof@nordicsemi.no>
Updates the chaining for advertising.
Instead of unconditionally adding a new PDU when new data
is added we now instead fill the last PDU in the chain with
the incoming data, only adding a new PDU when there is
not enough room.
This reduces the nr. of PDUs used for advertising, and also
fixes some qualification failures
Signed-off-by: Andries Kruithof <andries.kruithof@nordicsemi.no>
Prevent use of isochronous endpoints as bulk and/or interrupt endpoint.
The issue was observed when trying to use 4 CDC ACM instances where the
4th instance would claim the isochronous IN endpoint 0x88 to be bulk.
Because the isochronous endpoints cannot respond with handshake packet
the iso endpoints cannot be used as bulk or interrupt substitue.
Properly fail endpoint check and therefore make 4 CDC ACM instances not
enumerate at all because the hardware has endpoints that only allows up
to 3 CDC ACM instances.
Signed-off-by: Tomasz Moń <tomasz.mon@nordicsemi.no>
Update hal_silabs to pull request #45, fixing a file that
was copied from a wrong version.
Signed-off-by: Yonatan Schachter <yonatan.schachter@gmail.com>
Add the ESP32-C3-WROOM-02 modules with 4 or 8 MiB flash. The
temperature and antenna / connector variants are not mentioned
explicitely as they do not influence the software.
Signed-off-by: Detlev Zundel <dzu@member.fsf.org>
The PCA9633 i2c LED controller offers an All Call address in its nominal
operation, allowing simultaneous communication with all instances present
on the same i2c bus. The default address is 0x70. While this functionality
is convenient, it is possible that the board uses another i2c component
that also uses this address (for example, the shtcx). In such cases, the
address conflict prevents the proper functioning of the system.
The idea is to add a "disable-allcall" property to the device tree. If this
option is present, the initialization of the PCA9633 forces the bit 0
(ALLCALL) to be set to false, thereby disabling this function. It is
necessary to add this property to all PCA9633 devices on the bus to free up
the address 0x70.
Signed-off-by: Steve Jacot-Guillarmod <steve@piziwate.net>
The indexes were reset to 0 for each subgroup, which meant
that if we had a broadcast source with 2 subgroups and
1 BIS in each, both of them would use index 0x01.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
