Fixes: #70258
Specify NO_CMAKE_PATH to prevent CMake find_program from having a
preference for searching CMAKE_PREFIX_PATH.
Some tool like NixOS / Nix package manager sets this variable
to point inside the nix store, like this:
CMAKE_PREFIX_PATH=/nix/store/gpvnsgf8zhqjjgk63dd...-python3-3.11.6-env
thus causing `find_program()` to prefer this over the tool in the
default path.
Support Nix environments by specifying NO_CMAKE_PATH and in addition
support `VIRTUAL_ENV` environment setting for search path when set by
the virtual environment.
Signed-off-by: Torsten Rasmussen <Torsten.Rasmussen@nordicsemi.no>
The CMAC uses lp_clk as a sleep clock so it has to be updated if
frequency of lp_clk has changed. This happens either after XTAL32K
settling or RCX calibration.
Signed-off-by: Andrzej Kaczmarek <andrzej.kaczmarek@codecoup.pl>
Refactor the 54L board file to be able to better support out-of-tree
non-secure boards.
This aligns better with nrf53's DT.
Signed-off-by: Sebastian Bøe <sebastian.boe@nordicsemi.no>
Add missing nrf overlays to fix the sample at build time.
Copied from samples/drivers/watchdog/boards.
It is not clear why the build failure was not detected earlier.
Signed-off-by: Sebastian Bøe <sebastian.boe@nordicsemi.no>
Pins P0.06 and P0.07 are shorted on nRF5340 DKs prepared for on-target
tests because that is required for UART tests. That causes background
SPIM transfers in this sample to fail because SPIM is not able to drive
its SCK line configured as P0.06. Rearrange then the pins used in this
sample, so that SCK is not on a pin involved in a GPIO loopback.
Signed-off-by: Andrzej Głąbek <andrzej.glabek@nordicsemi.no>
Chosing by default in the prj.conf
PICOLIBC ensures we also use an embedded
libC even when built for a native target
and even if we dont set POSIX_API.
This ensures the test will build and pass
properly when built with default configuration
from command line without using the test yaml.
This change does not affect the build result
for other targets as those defaulted already
to PICOLIBC.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
Note that the Init of the alarm has no need for BCD2BIN conversion,
the LL_RTC_ALMx_Init will do. Add more log debug
The driver must clear the ALARM enable bit before wrting the alarm
registers.
Signed-off-by: Francois Ramu <francois.ramu@st.com>
Set a Calendar using the LL functions LL_RTC_TIME_Init
and LL_RTC_DATE_Init. Add more DBG info
On the stm32F2 serie, check RS Flag after setting calendar registers.
Signed-off-by: Francois Ramu <francois.ramu@st.com>
Add a timeout to the tests/drivers/rtc_api testcase so twister
can end properly when running with twister.
Signed-off-by: Francois Ramu <francois.ramu@st.com>
Implement a function bt_audio_get_chan_count that takes an enum
bt_audio_location and returns the number of channels in that value.
This PR fixes#69617
(https://github.com/zephyrproject-rtos/zephyr/issues/69617)
Signed-off-by: Babak Arisian <bbaa@demant.com>
Align grpcio version with logic2-automation package fixing its fail
on Channel.unary_unary() call with missing _registered_method argument..
Signed-off-by: Dmitrii Golovanov <dmitrii.golovanov@intel.com>
Make it a little bit easier to read using already existent
definitions for the registers used here.
Signed-off-by: Flavio Ceolin <flavio.ceolin@intel.com>
Secondary cores are not allowed to be power gated on
runtime-idle. They have to explicitely set off by host command.
Remove this state from secondary CPUs so power management logic
does not need workarounds to enforce this behavior.
Signed-off-by: Flavio Ceolin <flavio.ceolin@intel.com>
In twister, applying quarantine is a part of apply_filters() function.
However, this function is not called when --load-test is used.
Therefore, if one wants to use quarantines in combination with
dynamic scope from the test_plan.py script, one has to pass such
info through the script.
Signed-off-by: Maciej Perkowski <Maciej.Perkowski@nordicsemi.no>
This is a follow-up to commit e53e5448ee.
In the above commit, the calls to `wdt_setup()` were modified to use
the `WDT_OPT_PAUSE_HALTED_BY_DBG` option to prevent problems with
flashing of boards after this test was executed on them (the still
active watchdog could interrupt such flashing), but this option
is not essential to the test itself. And if a watchdog driver does not
support pausing when the CPU is halted by a debugger (this is the case
for example for wdt_counter), it is not able to pass the test because
of that option. Add then the possibility to continue the test when
pausing by debugger is not supported.
Signed-off-by: Andrzej Głąbek <andrzej.glabek@nordicsemi.no>
Two API calls that are supposed to set up the watchdog are not actually
causing the test to fail, only error messages are printed. This patch
adds the proper return statements.
Signed-off-by: Andrzej Głąbek <andrzej.glabek@nordicsemi.no>
Installation of single toolchain causes problems for customers who
are not familiar with Zephyr infrastructure. This tip is supposed
to help them.
Signed-off-by: Nikolay Agishev <agishev@synopsys.com>
When disconnect request is successful, the status
was taking as WIFI_REASON_DISCONN_UNSPECIFIED only.
Adding WIFI_REASON_DISCONN_SUCCESS which can be help
to determine status of disconnect request. If this status
is failed then reason can be useful.
Signed-off-by: Kapil Bhatt <kapil.bhatt@nordicsemi.no>
Get the preferred QoS settings via bt_bap_ep_get_info()
Therefore no need to use the internal header file to get it.
This PR fixes https://github.com/zephyrproject-rtos/zephyr/issues/72359
Signed-off-by: Ping Wang <pinw@demant.com>
When function sys_clock_set_timout() is called with small value
(i.e. 1) calculated time to be programmed to TIMER2 reload
register may be such that is expires before code set's it
up. In that case timer interrupt will be scheduled in far
future.
With this change, code checks after it sets reload value if
requested time already passed and if so TIMER2 interrupt
is marked as pending to avoid races.
Signed-off-by: Jerzy Kasenberg <jerzy.kasenberg@codecoup.pl>
With the addition of the STM32 hardware semaphore MBOX driver, this
sample is now supported by the board.
Signed-off-by: Celina Sophie Kalus <hello@celinakalus.de>
With the addition of the STM32 hardware semaphore MBOX driver, ICMsg is
now supported on the stm32h747i_disco board.
Signed-off-by: Celina Sophie Kalus <hello@celinakalus.de>
When sending a lot of data, a busy error code returned by the IPC is to
be expected. But if the last attempted send returns busy, this error
code is returned to the calling function, causing an error log message.
Fix this by setting return variable to 0 on this acceptable error.
Signed-off-by: Celina Sophie Kalus <hello@celinakalus.de>
The MBOX driver interface expects a device tree property '#mbox-cells'
which is not known by the IPM driver. This causes build problems when
both drivers are given for a single shared DT node.
To fix this problem for this driver specifically, add a dummy
'#mbox-cells' property to the bindings of the STM32 HSEM IPM driver.
This does not affect any other IPM driver, and the STM32 HSEM IPM driver
is still functioning with this dummy property.
Signed-off-by: Celina Sophie Kalus <hello@celinakalus.de>
Add a device tree binding for the new driver. Since there already
exists an IPM driver using the unsharable hardware semaphore interrupt,
the new driver is not added to any boards or SOCs per default to avoid
compatibility problems. See #37300 for the IPM driver.
Signed-off-by: Celina Sophie Kalus <hello@celinakalus.de>
This driver implements a simple MBOX device which supports a single
instance, two channels (one for each direction), and only signalling
mode with no data transfer. Signalling to another core is achieved by
taking and giving two hardware semaphores, similar to the STM32 HSEM
IPM driver.
Signed-off-by: Celina Sophie Kalus <hello@celinakalus.de>
The type of the resource table is known, casting to and from void* only
hides this type which can prevent the compiler from giving helpful
warnings. One warning would have been the accidental use of
"st_resource_table" in a cast, a struct which does not exist.
Use the fw_resource_table type when dealing with resource tables.
Signed-off-by: Andrew Davis <afd@ti.com>
This structure is shared between cores which may have different type
widths. Use fixed width types when defining the structure. We can
also use struct resource_table to help as it is already defined with
fixed width types and removes the need to redefine these elements.
Signed-off-by: Andrew Davis <afd@ti.com>
Check the return values of commands such as spi_nor_cmd_* and
spi_nor_wait_until_ready and ensure they are propagated back to the
caller on error.
Signed-off-by: Robert Hancock <robert.hancock@calian.com>
Torsten Rasmussen