For qemu_x86 and qemu_x86_64, it would seem that
heap-allocated thread stacks, as well as thread stacks
defined via K_THREAD_STACK_ARRAY_DEFINE(), must be at least
4096 bytes in size.
Using those sizes will ensure that these tests do not
fail with stack overflows or MMU faults.
Signed-off-by: Christopher Friedt <cfriedt@meta.com>
Create getter and setter functions for setting internal pthread
states to READY, RUN, or DONE queues.
Signed-off-by: Christopher Friedt <cfriedt@meta.com>
There may be race conditions when calling to_posix_thread()
from multiple threads.
Ensuing that the pthread pool lock is held when
to_posix_thread() is called will prevent those races.
Signed-off-by: Christopher Friedt <cfriedt@meta.com>
Rather than shadowing most of the fields from pthread_attr_t
(struct posix_thread_attr) inside of struct posix_thread, just
embed the structure itself.
Additionally, use pthread_attr_init() to allocate a thread
stack and pthread_attr_destroy() to destroy a thread stack,
dramatically simplifying pthread_create().
A fair bit of rework was needed already to mitigate bitrot
issues in the testsuite. More should probably be done later.
Signed-off-by: Christopher Friedt <cfriedt@meta.com>
The default pthread cancellation type is deferred.
Check that the canceltype is asynchronous (with
respect to cancellation points) before cancelling
a thread.
Signed-off-by: Christopher Friedt <cfriedt@meta.com>
Add a small wrapper to check if a pthread_attr_t has been
properly initialized (i.e. ready to pass to
pthread_create()).
Signed-off-by: Christopher Friedt <cfriedt@meta.com>
k_thread_stack_free() is designed to be called with any pointer
value. We return -EINVAL when an attempt is made to free an
invalid stack pointer.
This change reduces the verbosity in the degenerate case, when
the pointer is not obtained via k_thread_stack_alloc(), but
otherwise does not affect functionality.
If debug log verbosity is not enabled, we save a few bytes in
.text / .rodata / .strtab.
Signed-off-by: Christopher Friedt <cfriedt@meta.com>
The testsuite has been shifted to use dynamic thread
stacks (either statically allocated via a pool, or
dynamically allocated via the heap).
However, we definitely still need coverage for manually
specified thread stacks - but lets avoid duplicating
tests unnecessarily.
Signed-off-by: Christopher Friedt <cfriedt@meta.com>
Rather than fail when attempting to set an extraordinarily large
pthread stack size (much larger than most Zephyr platforms have)
consider the test non-fatal and skip it.
Signed-off-by: Christopher Friedt <cfriedt@meta.com>
Define the pinctrl-based pin controller instance
for the Mercury XU board and remove the old implementation
Signed-off-by: Jan Bylicki <jbylicki@antmicro.com>
Setting an extremely low value by default on two boards doesn't seem
like the right thing to do. The defaults were added with the v1 logging
subsystem in https://github.com/zephyrproject-rtos/zephyr/pull/8023.
Signed-off-by: Jordan Yates <jordan.yates@data61.csiro.au>
Export the minimal set of ztest_test_* symbols to llexts so that the
zassert_* macros can be used by the extension code to affect test
execution. Calling these also requires vprintk() to be exported.
Signed-off-by: Luca Burelli <l.burelli@arduino.cc>
soc\riscv\andes_v5\ae350\soc.h was empty and deleted,so revise
wdt_andes_atcwdt200.c to resolve
'fatal error: soc.h: No such file or directory'.
Signed-off-by: cyliang tw <cyliang@nuvoton.com>
Some very minor touch-ups for multi-level interrupt
wordings and documentations to better reflects its current
state.
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
If an ld file is included multiple times using `zephyr_linker_sources`,
only the last occurrence is actually used. Everything else is removed
again from the generated snippet files.
This allows to relocate certain blocks, e.g. the vector table, in an
application-specific CMakeLists.txt (or for an arch or soc).
Fixes: #66315
Signed-off-by: Greter Raffael <rgreter@baumer.com>
Use 'struct uc81xx_tres16' for 16-bit 'tres' setup, instead of 'struct
uc81xx_tres8'. This fixes a regression when support for 'uc8175' was added
and 'struct uc81xx_tres' was replaced with 'struct uc81xx_tres16'.
Fixes: 7c46b0b898 ("drivers: display: uc81xx: add support for uc8175")
Signed-off-by: Marcin Niestroj <m.niestroj@emb.dev>
`struct ieee802154_config config` is a struct containing an
union. Members of `config` were accessed incorrectly in
`otPlatRadioEnableCsl`. Fix by initializing `config` with `0`
and accessing one member at a time.
Signed-off-by: Maciej Baczmanski <maciej.baczmanski@nordicsemi.no>
- In `set_vendor_ie_header_lm`, case when
`link_metrics_data_len == 0` has been ignored.
This commit fixes that by setting `header_ie->length = 0`
before returning.
- current implementation of enh ACK header IE returns
`-ENOTSUP` when `ack_ie.header_ie == NULL` or
`ack_ie.header_ie->length == 0`. This commit fixes that by
refactoring checks in `nrf5_configure`.
Co-authored-by: Przemyslaw Bida <przemyslaw.bida@nordicsemi.no>
Signed-off-by: Maciej Baczmanski <maciej.baczmanski@nordicsemi.no>
Disabling the MPU doesn't clear regions configuration. There is a risk
in multi-image environment that there are some old region setting e.g.
stack guard. This may cause a memory fault, because of different images
layout e.g. RO/RW.
Just clear and disable all regions configuration before the new
initialization.
Signed-off-by: Dawid Niedzwiecki <dawidn@google.com>
Fix radio packet length configuration when first subevent
has empty PDU in the BIG event.
Signed-off-by: Vinayak Kariappa Chettimada <vich@nordicsemi.no>
Fix ISO Synchronized Receiver sequential packing BIS PDU
being dropped in the first repetation just after previous
BIS subevents where skipped.
Signed-off-by: Vinayak Kariappa Chettimada <vich@nordicsemi.no>
Fix Broadcast ISO Create Scheduling to consider reducing
Host requested RTN value. And also use maximum overhead
calculation when Extended Advertising and Periodic
Advertising have not been started before BIG create.
Signed-off-by: Vinayak Kariappa Chettimada <vich@nordicsemi.no>
Fix incorrect HCI ISO data length check for Broadcast ISO
transmissions.
Removed initially added code when preliminary support for
Broadcast ISO was implemented.
Signed-off-by: Vinayak Kariappa Chettimada <vich@nordicsemi.no>
MDIO driver header (mdio.h)is using errno values without including
errno header, this causes build errors depending on the order of inclusion
of this mdio header in other files, fix by including the errno header
in this mdio.h file.
Signed-off-by: Declan Snyder <declan.snyder@nxp.com>
Fix a build error with certain older Cadence XCC toolchains. These
are used e.g. for Intel Tiger Lake products for the audio DSP (thhe oldest
platform supported in Zephyr upstream for the audio DSP).
To keep all compilers happy, use C89 style initializers.
Error:
lib/os/p4wq.c:216: error: unknown field ‘dev’ specified in initializer
lib/os/p4wq.c:216: warning: missing braces around initializer
lib/os/p4wq.c:216: warning: (near initialization for
‘__init_static_init.<anonymous>’)
Compiler version XCC RG-2017.8-linux (xt-xcc version 12.0.8)
Fixes: 2438dbb613 ("init: add missing initialization of dev pointer
in SYS_INIT macro")
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
We should not have any dependencies on optional modules, for example on
Kconfigs from optional modules. Compliance has to pass without any
optional modules enabled.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
Add functions to unset, or remove, specific codec
LTV structure from codec_cfgs or codec_caps.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
Add support for CAP test cases.
Split btp_bap.c for better maintenance into unicast and broadcast.
Signed-off-by: Magdalena Kasenberg <magdalena.kasenberg@codecoup.pl>
Utilize a code spell-checking tool to scan for and correct spelling errors
in all files within the doc/releases directory.
The typo in the 'Issue Related Items' section will be left unchanged.
Signed-off-by: Pisit Sawangvonganan <pisit@ndrsolution.com>
In the code comments referring to section 3.6.1, changed the reference
from 'SD specification' to 'SD host controller specification'
for a clearer and more accurate context.
Signed-off-by: Pisit Sawangvonganan <pisit@ndrsolution.com>
After a review of the usage of csd[4] and cid[4],
it has been determined that zero-initialization can be omitted.
It can be expected from successful operations that data will be
appropriately written back from the underlying layer.
Signed-off-by: Pisit Sawangvonganan <pisit@ndrsolution.com>
The decision to omit zero-initialization is driven by a desire to enhance
Zephyr's compactness and efficiency.
This is achieved by omitting zero-initialization, thereby reducing the
instruction count and, as a byproduct, the code size.
After a thorough review of the usage of struct sdhc_command and sdhc_data,
it has been determined that zero-initialization can be omitted.
Only a portion of the fields need to be manually initialized.
(e.g. cmd.retries, data.block_addr)
For the uninitialized fields, it can be expected from successful
operations that data will be appropriately written back from
the underlying layer.
Signed-off-by: Pisit Sawangvonganan <pisit@ndrsolution.com>
Introduced the config SD_CMD_RETRIES option as a build-time parameter.
This allows the assignment of a specific value to cmd.retries.
The default value has been set to 0 to preserve the previous setting.
Signed-off-by: Pisit Sawangvonganan <pisit@ndrsolution.com>
Only when CONFIG_MP_MAX_NUM_CPUS > 1, then .bss is put in
uncached region. Otherwise, .bss is in cached region.
So the assertion that g_key_read_holder must be in uncached
region must take into account how many CPUs are enabled on
build.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
Updates in documentation support for `mq_notify` API
in `_POSIX_MESSAGE_PASSING` group option.
Signed-off-by: Adam Wojasinski <awojasinski@baylibre.com>
Adds test cases testing following features:
- function error handling
- basic notification type
- thread notification type
- adding notification to non-empty queue
Signed-off-by: Adam Wojasinski <awojasinski@baylibre.com>
The function was the last missing piece of the `_POSIX_MESSAGE_PASSING`
option group. Due to lack of signal subsystem in the Zephyr RTOS
the `sigev_notify` member of the `sigevent` structure that describes
the notification cannot be set to `SIGEV_SIGNAL` - this notification
type is not implemented, the function will return -1 and set `errno`
to `ENOSYS`.
`mq_notify` documentation:
https://pubs.opengroup.org/onlinepubs/9699919799/functions/mq_notify.htmlFixes#66958
Signed-off-by: Adam Wojasinski <awojasinski@baylibre.com>
Provide three basic examples to test the x-nucleo-iks4a1 shield:
- Test shield in standard mode
Acquire sensor data from shield with all MEMS sensors connected
to micro-controller
- Test shield in HUB1 mode
Acquire sensor data from shield with lis2mdl and lps22df
connected to LSM6DSV16X sensor hub
- Test shield in HUB2 mode
Acquire sensor data from shield with lis2mdl and lps22df
connected to LSm6DSO16IS sensor hub
Signed-off-by: Armando Visconti <armando.visconti@st.com>
x-nucleo-iks4a1 shield is an arduino compatible companion board
which can be used on top of Nucleo standard boards for industrial
applications. Following MEMS sensor are currently supported:
- LSM6DSO16IS: MEMS 3D accelerometer + 3D gyroscope
- LSM6DSV16X: MEMS 3D accelerometer + 3D gyroscope
- LIS2MDL: MEMS 3D magnetometer
- LPS22DF: Low-power and high-precision MEMS pressure sensor
(https://www.st.com/resource/en/data_brief/x-nucleo-iks4a1.pdf)
The board exports three overlays:
1. x_nucleo_iks4a1.overlay (standard mode)
All MEMS sensors are connected to micro-controller.
2. x_nucleo_iks4a1_shub1.overlay (HUB1 mode)
LSM6DSV16X IMU sensor act as a sensor_hub with LIS2MDL and
LPS22DF connected to it.
3. x_nucleo_iks4a1_shub2.overlay (HUB2 mode)
LSM6DSO16IS IMU sensor act as a sensor_hub with LIS2MDL and
LPS22DF connected to it.
Signed-off-by: Armando Visconti <armando.visconti@st.com>
Commit 3b99fb1b4a ("xtensa: do not imply atomic ops kconfig") removed
ATOMIC_OPERATIONS_ARCH at xtensa arch level. This triggers a bug in
intel_adsp cavs builds with XCC compiler as
CONFIG_ATOMIC_OPERATIONS_BUILTIN is not defined but neither is
CONFIG_ATOMIC_OPERATIONS_ARCH anymore, resulting in failed builds.
Fix the XCC build by defining CONFIG_ATOMIC_OPERATIONS_ARCH at
soc level.
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
