gcc prints this warning message
'strncat' specified bound 1 equals source length [-Wstringop-overflow=]
58 | strncat(fd, "C", 1);
There was no error in the code but avoid the warning by not using
strncat().
Signed-off-by: Jukka Rissanen <jukka.rissanen@nordicsemi.no>
If CONFIG_POSIX_API is enabled, then the socket.h is found under
zephyr/posix/sys/socket.h etc. This allows one to compile the
socket test applications without error prints.
Signed-off-by: Jukka Rissanen <jukka.rissanen@nordicsemi.no>
The socket service provides a similar functionality as what
initd provides in Linux. It listens user registered sockets
for any activity and then launches a k_work for it. This way
each application does not need to create a thread to listen
a blocking socket.
Signed-off-by: Jukka Rissanen <jukka.rissanen@nordicsemi.no>
For a CLIC the vector table has to be aligned by 512 bytes, if there are
between 65 and 128 interrupts, which is the case for the gd32vf103.
`isr_wrapper` has to be aligned to 64 bytes, s.t. the lower 6 bits of
mtvec are 0.
Signed-off-by: Greter Raffael <rgreter@baumer.com>
For a proper initialisation, the soc-specific `__nuclei_start` has to
be executed before the common `__start`. To ensure that `__nuclei_start`
is linked first, I added the linker section init.
Signed-off-by: Greter Raffael <rgreter@baumer.com>
A lot of the entry.S is again implemented in common/vector.S.
I removed everything redundant and changed the jump to the common
symbol __start at the end.
Signed-off-by: Greter Raffael <rgreter@baumer.com>
Ignore communication faults of the TLE9104 which are reported
before the communication watchdog is configured.
Fixes#67370
Signed-off-by: Benedikt Schmidt <benedikt.schmidt@embedded-solutions.at>
The function did not move data correctly when increasing
or decreasing the size of a value that was not the last value
in the LTV array.
Added a few tests to verify the fix, using the CCID list as the
main way of verifying it.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
In order to make state more manageable among the other
posix api tests, create a separate ZTEST_SUITE() called pthread
and another ZTEST_SUITE() called pthread_attr.
The tests in the pthread suite mainly operate on the pthread_t
type, while tests in the pthread_attr suite mainly operate on
the pthread_attr_t type.
The pthread_attr testsuite is moved to a separate file
(pthread_attr.c).
Having separate test suites allows us to narrow the scope of
the tests and have a better degree of control over the state.
Some effort went in to highlighting undefined behaviour at the
API level. This was quite intentional, as we will need to be
be able to test / verify that we do in fact have deterministic
outcomes when users attempt operations that result in undefined
behaviour (e.g. assert).
Signed-off-by: Christopher Friedt <cfriedt@meta.com>
Instead of relying on a fault occurring when NULL
parameters are passed in, explicitly check and return
EINVAL.
Signed-off-by: Christopher Friedt <cfriedt@meta.com>
No point marking the phy API to be callable from usermode,
the device cannot be accessed from usermode anyway so this
is pointless. User can call the phy API from supervisor mode
thread just fine.
Signed-off-by: Jukka Rissanen <jukka.rissanen@nordicsemi.no>
Reduce some of the arithmetic required to decode a sample. Math is
documented in bma4xx_convert_raw_accel_to_q31.
This improved the accuracy from roughly 0.865mm/s2 to 0.001mm/s2 when
stationary and using a range of 4g.
Signed-off-by: Yuval Peress <peress@google.com>
ssize is a POSIX.1-2001 extension, which may or may
not be provided by the C library, or may be defined
to a different size in the host and embedded C library.
Two internal functions were returning ssize.
As these functions were just trampolines
into the same host API, which are already provided
by the native simulator let's just use those.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
ssize is a POSIX.1-2001 extension, which may or may
not be provided by the C library, or may be defined
to a different size in the host and embedded C library.
Two internal functions were returning ssize, but
one of them was a trampoline into the same host API,
which is already provided by the native simulator
so let's just use that instead.
The other is only carrying data that fits into an
int and is anyhow being cast in/to ints, so let's just
avoid the trouble by defining it as returning int.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
The current driver initializes the IADC with the default configuration
(IADC_INITSINGLE_DEFAULT), which aligns the data to the right.
To correctly read the 12-bit sample, it should be masked from the right
instead.
Signed-off-by: Paulo Santos <pauloroberto.santos@edge.ufal.br>
There was a few places where the scan delegator returned an
errno value instead of a valid GATT error codes when rejecting
write request. These have been modified to send
BT_ATT_ERR_WRITE_REQ_REJECTED to the client instead.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
Add additional debug information in some ASCS debug statements,
so it is easier to follow the state when there is a potentional
state change.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
This adds notifications to do_first_track only when changing group.
Before the change was made, client was never notified about
track change when changing group in Media Player.
Signed-off-by: Piotr Narajowski <piotr.narajowski@codecoup.pl>
Add overlay to the blinky pwm sample to run the application on the
Wio Terminal. The USER LED on PA15 is used to setup the PWM LED.
Signed-off-by: Joel Guittet <joelguittet@gmail.com>
Some of the ASCS response codes should not be used by the
unicast server, as those will be checked and validated by the
stack.
They are kept in the enum, as the enum is also used by the
unicast client.
Besides updating the documentation, this commit also adds
a function to validate the response codes as well as the
combination of response codes and reason/metadata values,
and log a warning if the application does not adhere to the
documented behavior.
A warning is used instead of an error/assert/modification
of the response codes, as there may be unforseen reasons
why an application provide a non-approved response code,
and since it may be spec-valid, we should not fully
prevent it.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
According to the POSIX specification, null pointer
is a valid value for the `address` argument
of the `accept` function.
This commit adds a check to prevent a null pointer
dereference inside `z_impl_zsock_accept`.
Signed-off-by: Mirko Covizzi <mirko.covizzi@nordicsemi.no>
This PR makes the modem_chat wait until a chat script chat
request has been sent before accepting responses to said
request.
This helps ensure that an unsolicitet response is not mistaken
for a response to a request, which is especially problematic
if the chat script chat is using an any match.
For example, start chat script sending AT+CGMI, expecting the
modem name as a response followed by OK
> start script, waiting for response immediately
> start sending "AT+CGMI"
> receive "+CEREG 1,0" while sending
> script accepts "+CEREG 1,0" as the response to "AT+CGMI"
> "AT+CGMI" sent
> receive "QUECTEL BG95"
> receive "OK"
script handler got "+CEREG 1,0" instead of "QUECTEL BG95"
After this PR:
> start script
> start sending AT+CGMI
> receive "+CEREG 1,0" while sending
> "AT+CGMI" sent
> start waiting for response
> receive "QUECTEL BG95"
> script accepts "QUECTEL BG95" as response to "AT+CGMI"
> receive "OK"
Signed-off-by: Bjarki Arge Andreasen <bjarki@arge-andreasen.me>
The bt_le_adv_start function can (currently) not be used
together with BT_LE_ADV_OPT_EXT_ADV, and it will
return an -EINVAL error if attempted, but this behavior
is not documented, and several people have asked about
why it does not work.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
Calling I2S write with less bytes than I2S TX memory block size makes it
possible to synchronize the time when next block is used against some
other, possibly externally sourced, signal. Example use case includes
USB Audio where audio sink and/or source has to be synchronized against
USB SOF. In Asynchronous synchronization type the rate matching is
achieved by varying the number of samples sent/received by 1 sample
(e.g. for 48 kHz audio, 47 or 49 samples are transmitted during frame
instead of 48).
Signed-off-by: Tomasz Moń <tomasz.mon@nordicsemi.no>
Fix an issue where updating a `zephyr_code_relocate()` call in CMake
didn't trigger regeneration of `code_relocation.c` and linker scripts.
The generation command was missing a dependency on the aforementioned
input text file, where the outcome of each call is cached.
Signed-off-by: Grzegorz Swiderski <grzegorz.swiderski@nordicsemi.no>
When using the code and data relocation feature, every relocated symbol
would be marked with `KEEP()` in the generated linker script. Therefore,
if any input files contained unused code, then it wouldn't be discarded
by the linker, even when invoked with `--gc-sections`.
This can cause unexpected bloat, or other link-time issues stemming from
some symbols being discarded and others not.
On the other hand, this behavior has been present since the feature's
introduction, so it should remain default for the users who rely on it.
This patch introduces support for `zephyr_code_relocate(... NOKEEP)`.
This will suppress the generation of `KEEP()` statements for all symbols
in a particular library or set of files.
Much like `NOCOPY`, the `NOKEEP` flag is passed to `gen_relocate_app.py`
in string form. The script is now equipped to handle multiple such flags
when passed from CMake as a semicolon-separated list, like so:
"SRAM2:NOCOPY;NOKEEP:/path/to/file1.c;/path/to/file2.c"
Documentation and tests are updated here as well.
Signed-off-by: Grzegorz Swiderski <grzegorz.swiderski@nordicsemi.no>
There are some situations like #67035 where sparse aborts and returns an
error code before the compiler has generated the .obj file; without any
clear indication that the .obj is missing (in normal situations sparse
prints warnings and _does_ creates the .obj file)
Also, builds are parallel by default and sparse runs tend to be massive
walls of text which all conspires to make it totally impossible to find
the relevant error message. Instead, we get an link-time error.
The only clear indication is the exit code. So catch it and abort the
build ASAP thanks to COMMAND_ERROR_IS_FATAL.
More generally speaking, the default behavior of execute_process() to
ignore errors is crazy. How frequently does a build system run commands
that do NOT matter?
Signed-off-by: Marc Herbert <marc.herbert@intel.com>
Since the pins of bt-spi instance are wired internally in the chip, it will
make sense to move the definition to soc dts so no need for every board
using the chip to redefine the same.
Signed-off-by: Aaron Ye <aye@ambiq.com>
Fix the log level of two LOG_ERR() statements which should have always
been LOG_INF(). As confirmed by the author Adrian in #60172
Fixes commit 3fbaed4de9 ("dai: intel: ace: dmic: Refactor of
dai_nhlt_dmic_dai_params_get function")
Signed-off-by: Marc Herbert <marc.herbert@intel.com>
Employ a code spell checking tool to scan and correct spelling errors
in all files within the drivers/i2c directory.
Signed-off-by: Pisit Sawangvonganan <pisit@ndrsolution.com>
.bss and .data are uncached in Zephyr builds for intel_adsp. No need
to try to manipulate cache of objects in those sections.
Signed-off-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
It seems like a nice idea at first, but leads to hard-to-debug
situations for the application.
The previous behavior can be implemented by the app by defining
`alloc_seg` and allocating from the same pool as `buf`.
Signed-off-by: Jonathan Rico <jonathan.rico@nordicsemi.no>
When OpenThread application is built with CONFIG_ASSERT
and CONFIG_ASSERT_NO_MSG_INFO, OT_ASSERT() prints a location
that points to the otPlatAssertFail() function instead of
the code that actually failed an assertion.
This is confusing and CONFIG_ASSERT_NO_MSG_INFO sometimes
cannot be disabled because of flash size limitations.
Make otPlatAssertFail() always print the actual assert
location.
Signed-off-by: Damian Krolik <damian.krolik@nordicsemi.no>
