axis1bit16 and axis3bit16 unions are no longer used
and can be deleted from all .h files that are referencing
them.
Signed-off-by: Armando Visconti <armando.visconti@st.com>
Align all sensor drivers that are using stmemsc (STdC) HAL i/f
to new APIs of stmemsc v2.00.
Requires https://github.com/zephyrproject-rtos/hal_st/pull/7
(merged as 575de9d461aa6f430cf62c58a053675377e700f3)
Signed-off-by: Armando Visconti <armando.visconti@st.com>
Reprocess keyscan event processing in case of an I2C bus read
error. Otherwise, the HT16K33 will be stuck (when used in IRQ mode, not
in polling mode) and no new IRQs will be delivered.
Signed-off-by: Henrik Brix Andersen <henrik@brixandersen.dk>
Convert the keyscan portion of the Holtek HT16K33 driver to adhere to
the kscan API instead of the GPIO API.
When this driver was introduced the kscan API was not present. The
keyscan driver was therefore implemented as a GPIO interrupt driver.
Signed-off-by: Henrik Brix Andersen <henrik@brixandersen.dk>
The was a missing comma in the DEVICE_DT_INST_DEFINE macro and the
SAME71 HAL tweaks the name of a struct so we have to work around that.
Fixes#36095
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
Increase main and system workqueue stack sizes to avoid overflows
as the LoRaMac-node library is quite stack-hungry.
Signed-off-by: Ilya Tagunov <tagunil@gmail.com>
Update the LoRaMac-node library to the last stable release and fix
the Zephyr glue code to match it. Move CMakeLists.txt to the main
Zephyr repository to simplify loramac-node module maintenance.
Signed-off-by: Ilya Tagunov <tagunil@gmail.com>
Modify the check from checking the feature bit to
checking the command bit. This ensures that we
don't send the read buffer size V2 to a controller
that does not support it.
This also moves the entire ISO init procedure into
a separate function to avoid having a large
ISO-only block in `le_init`.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
Adds a mention of the Kconfig hardening script hardenconfig to
the Kconfig documentation to promote awareness of its existence.
Signed-off-by: Lauren Murphy <lauren.murphy@intel.com>
This commit removes a superfluous null pointer check.
The device is accessed via a label.
so in case it is not available a build error arises,
but the pointer won't ever be null.
Signed-off-by: Thomas Stranger <thomas.stranger@outlook.com>
This check is present in att_read_rsp(), but att_read_mult_req and
att_read_mult_vl_req do not use it. Add this check to these functions.
This was affecting GATT/SR/GAS/BV-08-C
Signed-off-by: Krzysztof Kopyściński <krzysztof.kopyscinski@codecoup.pl>
Make kscan init priority higher, since it is lower than some buses like
I2C, making some devices to fail initialization if not tuning
priorities.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Allow caller to specify microsecond accuracy and not convert
to milliseconds.
Fixes#36072
Signed-off-by: Jukka Rissanen <jukka.rissanen@linux.intel.com>
Make zsock_select() a syscall so that the following commit
can call the internal poll implementation directly. This is
needed as zsock_select() will not call zsock_poll() directly
in order to allow select to use microsecond timeout accuracy.
Signed-off-by: Jukka Rissanen <jukka.rissanen@linux.intel.com>
GPIOTE1 on the nRF5340 SoC is always accessible as a non-secure
peripheral. However, it is only defined in the non-secure DTSI file.
This is therefore a missing node in the secure DTSI file, since
non-secure addresses are accessible by secure software.
Move the node definition to a common include file and pull it into the
app core DTSI as well. To keep things clean, adjust the node labels so
that:
- 'gpiote0' and 'gpiote1' are defined in the secure DTSI
- 'gpiote0' is not defined in the non-secure DTSI
- 'gpiote' is defined in both secure and non-secure DTSIs
- 'gpiote' refers to the same node as 'gpiote0' in the secure DTSI
- 'gpiote' refers to the same node as 'gpiote1' in the non-secure DTSI
Signed-off-by: Martí Bolívar <marti.bolivar@nordicsemi.no>
The sram1 node in nrf5340_cpuapp.dtsi represents the SRAM accessible
to the network core on the SoC.
However, while the network core can access the app core's SRAM, the
app core cannot access the net core's SRAM. Therefore, the sram1 node
should only appear in DTSI files for the net core, and not the app
core.
Signed-off-by: Martí Bolívar <marti.bolivar@nordicsemi.no>
Because it is not always possible to determine whether
a request is for Audio class or not, all requests are passed
to it. This can lead to the requests, to e.g. HID interface,
being hijacked by Audio class.
Ignore return value of Audio custom handler as it is not relevant here.
Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>
Since the settings are stored in the core now,
we can remove REQ_GET_INTERFACE part from custom handler
and simplify the code. Return value is no longer relevant
since custom callback does not need to interfere with the
request handling process.
Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>
Store alternate interface settings and
return alternate on GetInterface request.
Fixes: #24200
Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>
This option is deciding whether bootutil_public library supports
multi-image procedures. MCUboot have its own definition which is
unavailable zephyr application. Introduced option declares similar
option for the application.
Signed-off-by: Andrzej Puzdrowski <andrzej.puzdrowski@nordicsemi.no>
Add boot_write_confirm_multi() and boot_request_upgrade_multi() APIs so
that the user can set the image with given index as pending, confirmed.
This is needed for enabling the mcuboot multi-mage boot feature by
the zephyr.
Signed-off-by: Andrzej Puzdrowski <andrzej.puzdrowski@nordicsemi.no>
Change Highlights:
- Fix error check after `k_work_reschedule_for_queue`. A value of 1
means job was scheduled which was resulting in a ton of "<err>
wifi_eswifi: Rescheduling socket read error" logs getting printed
due to the erroneous check
- When using the B-L475E-IOT01A, attempts to use a TLS socket result
in a hang when socket offload is enabled so I'd like to have a way
to disable the option. To accomplish this, I I switched the
`CONFIG_NET_SOCKETS_OFFLOAD=n` Kconfig option from `select` to
`imply`.
- There was a missing `net_context_set_state()` call when
`CONFIG_NET_SOCKETS_OFFLOAD=n`. I applied the same fix from #30664
for this case to fix the issue.
Signed-off-by: Chris Coleman <chris@memfault.com>
This commit adds the flash driver for nucleo_f207zg platform.
This has been tested with flash test application.
Signed-off-by: Krishna Mohan Dani <krishnamohan.d@hcl.com>
This commit enables flash support for nucleo-f207zg in device tree.
The last sector of 128KB is used as storage partition.
Signed-off-by: Krishna Mohan Dani <krishnamohan.d@hcl.com>
This is a follow-up to commit 60d9988401.
Add a third option for the HFXO capacitance that keeps the default
value of the XOSC32MCAPS register untouched. The message in the above
commit incorrectly claimed that external load capacitors for HFXO
(the reset value of the XOSC32MCAPS register) was the configuration
in effect before. In fact, the register value was modified during
the device trimming in the SystemInit() function to use the internal
capacitors, and that is the configuration required for proper RADIO
operation on nRF5340 DK, for instance.
Signed-off-by: Andrzej Głąbek <andrzej.glabek@nordicsemi.no>
Sphinx provides a way to persist data across builds: the
BuildEnvironment. The build environment is automatically managed by
Sphinx, so there is no need to take care of loading/dumping cache files.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Use file hashes instead of paths and modification times. This method is
fits better on systems using cache mechanisms. Note that hash is
computed using content obtained in UTF-8 text mode for portability
reasons.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Document that calling settings_load is required when the application is
not creating and managing the identities of the stack itself.
The application will not be able to use all features of the stack before
the identities have been loaded.
Signed-off-by: Joakim Andersson <joakim.andersson@nordicsemi.no>
We do not need the return value of k_sem_take() so ignore it.
Coverity-CID: 236602
Fixes#36313
Signed-off-by: Jukka Rissanen <jukka.rissanen@linux.intel.com>
We were testing modified_tags.args before we called get_twister_opt.py
so it would never have been created and thus twister_exclude_tag_opt
would never get set correctly. Move the check to after we call
get_twister_opt.py
Also we need to remove modified_tags.args once we are done with it.
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
CONFIG_DEBUG_OPTIMIZATIONS should not be enabled by default in
prj.conf for tests or samples. Remove it from a few places that it
is.
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
When the host toolchain is used, `-nostdlib` option is not specified
and therefore all default libraries, including the libgcc and its
components, are automatically linked -- so it is not necessary to
manually link them here.
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
