HWSPINLOCK drivers should be initialized very early in the kernel life, so
they can be used by other drivers without breaking their init dependencies.
Use the default init priority (40), instead of default device init
priority (50).
Signed-off-by: Pierrick Guillaume <pguillaume@fymyte.com>
HWSPINLOCK are used to protect accesses to shared resources between
clusters. However, we don't want to sleep or be preempted while holding
a HWSPINLOCK. This means we need to lock IRQs while holding the
HWSPINLOCK.
Moreover, in an SMP context, we need to protect accesses to a HWSPINLOCK
between cores, so it is not possible to lock or release the same
HWSPINLOCK multiple times.
Introduce a zephyr spinlock per HWSPINLOCK to counter those issues. User
facing APIs have been udpated to take into account those changes, and a new
hwspinlock_dt_spec struct has been added to align with other subsystem on
the management of complex device's data.
Take the opportunity to rename public APIs to match the naming scheme of
normal zephyr spinlocks:
hwspinlock_lock -> hw_spin_lock
hwspinlock_trylock -> hw_spin_trylock
hwspinlock_unlock -> hw_spin_unlock
Signed-off-by: Pierrick Guillaume <pguillaume@fymyte.com>
Add general hwlocks and hwlock-names to base.yaml to be utilized by any
clients that uses hwspinlocks.
Additionally add hwspinlock-controller.yaml for common properties shared by
all hwspinlock controller devices.
Signed-off-by: Pierrick Guillaume <pguillaume@fymyte.com>
HW spinlocks are not supposed to be used by user application. They are used
to protect shared resources between clusters. Consumer of such spinlocks
should be other drivers in kernel space.
Moreover, syscalls were actually not generated properly because of
a missing call to zephyr_syscall_header() in CMakeLists.txt.
Signed-off-by: Pierrick Guillaume <pguillaume@fymyte.com>
Add a configuration option to override the default BOOT_MAX_ALIGN value in
mcuboot. It defaults to the chosen flash's write block size to make it
backwards compatible.
This can be used when image slots are on different flash devices.
Signed-off-by: Pieter De Gendt <pieter.degendt@basalte.be>
1. Introduced GDMA support for efficient data transfer in the
NPCX FIU QSPI driver.
2. Refactor flash driver mutex handling to enhance concurrency safety,
preventing other threads from preempting erase/write operations during
UMA until mutex release.
Signed-off-by: Alvis Sun <yfsun@nuvoton.com>
Signed-off-by: Tom Chang <CHChang19@nuvoton.com>
Signed-off-by: Mulin Chao <mlchao@nuvoton.com>
1. Introduces boolean properties to enable support for different
versions of the FIU/Nor flash driver.
2. Add dmas phandle for DMA configuration.
Signed-off-by: Alvis Sun <yfsun@nuvoton.com>
Update the CAP initiator audio start unittest.
The reason for this, is that the recent change to endpoints
means that we need to properly generate a BAP unicast group
and call bt_bap_stream_config, instead of trying to manage
the bap_iso values directly.
As part of updating this, the tests were also updated to
significantly reduce code duplication, but generating
a default start param that can easily be reused.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
Adds assert for bt_bap_iso_bind_ep to ensure that endpoints
from different connection don't share a CIS (a bt_bap_iso).
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
Add a check that verifies that 2 endpoints from different
connection don't share a CIS, which would be invalid.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
We can actually already when we bind the endpoint to the stream
perform the call to bt_bap_iso_bind_ep, if the stream has been
added to a group.
If the stream has not yet been added to a group, then when
the stream is added to a group, then bt_bap_iso_bind_ep will
be done there (existing behavior).
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
Add condition to check whether the pin is set as output before
reading the output high or low for gpio driver of Renesas RZ
Signed-off-by: Quang Le <quang.le.eb@bp.renesas.com>
Signed-off-by: Nhut Nguyen <nhut.nguyen.kc@renesas.com>
The i.MX 91 Quick Start Board (MCIMX91-QSB board) is a platform designed
to display the most commonly used features of the i.MX 91 applications
processor. The MCIMX91-QSB board is an entry-level development board
with a small and low-cost package. The board can be used by developers
to get familiar with the processor before investing a large amount of
resources in more specific designs.
Signed-off-by: Hou Zhiqiang <Zhiqiang.Hou@nxp.com>
Change imx9111 and imx9131 use separate dts as they has different pin
configure.
Signed-off-by: Jiafei Pan <jiafei.pan@nxp.com>
This commit provides a CMakePresets.json which includes the sample's
CMakePresets.json file.
`west build` is extended to set `APP_DIR` in environment when sysbuild
is used. This allows sysbuild's CMakePresets.json to include the
sample's presets file.
Signed-off-by: Torsten Rasmussen <Torsten.Rasmussen@nordicsemi.no>
Use CONFIG_STM32_FLASH_PREFETCH config option to enable flash prefetch
in C0/F0/F1/F2/F3/F4/F7/G4/L0/L1/L4/U0/U3/WB/WBA/WBL for consistency
with other SoCs that use this configuration switch, default enabled
at SoC default config level.
Add SoC hidden config option HAS_STM32_FLASH_PREFETCH enabled for
SoCs that support the feature. STM32_FLASH_PREFETCH is default
enabled for all SoC that have HAS_STM32_FLASH_PREFETCH unless target
specific constraints (as for G0Bx/G0Cx SoCs).
Signed-off-by: Etienne Carriere <etienne.carriere@st.com>
For compatibility, do not use spare registers to store mask data.
RTC mask data can be generated through the calendar and time mask
registers at runtime.
Signed-off-by: cyliang tw <cyliang@nuvoton.com>
gcovr 8.0 introduces suspicious_hits detection, warn on such detection, but
do not stop coverage calculation at all.
Signed-off-by: Bartlomiej Buczek <bartlomiej.buczek@nordicsemi.no>
Move to the latest commit of the hal_adi, using the new msdk-export branch
with a clean commit history.
Signed-off-by: Pete Johanson <pete.johanson@analog.com>
Assert the reset pin for the duration specified by the modem
(`reset_pulse_duration_ms`) when booting, instead of immediately
de-asserting the pin.
Signed-off-by: Jordan Yates <jordan@embeint.com>
Prioritise asserting the reset pin when a script fails instead of the
power pin, as the reset pin is a "harder" reset mechanism.
Signed-off-by: Jordan Yates <jordan@embeint.com>
Commit eff5d02872 added
`net_socket_service_desc` to common-rom ld snippet but not to common-rom
CMake linker generator equivalent.
Bring common-rom.cmake up-to-date to ensure `net_socket_service_desc`
works correctly with the linker generator.
Signed-off-by: Torsten Rasmussen <Torsten.Rasmussen@nordicsemi.no>
Twister netif:eth was missing on stm32n6 based board and hence they
were not build against net samples.
Fix this.
Additionally, on dk, fix `pwm` to respect alphabetical order.
Signed-off-by: Erwan Gouriou <erwan.gouriou@st.com>
Compilers may not like that attributes are provides in actual and extern
definitions of symbols.
Removing them from extern definitions
Signed-off-by: Erwan Gouriou <erwan.gouriou@st.com>
commands should not have spaces around hyphens:
version - extended --> version-extended
build - version --> build-version
Fixes: zephyrproject-rtos/zephyr#96591
Signed-off-by: Benjamin Cabé <benjamin@zephyrproject.org>
If the process does an exec() (or fork, or..) all descriptors are kept
open by default, unless O_CLOEXEC is set when opening them.
This is usefull for stdin/out/err so that new process is connected to
them, but it is very rare for it to be usefull for any other descriptor.
In general this leads to descriptors being kept open unnecessarily,
which either will block other process from getting them (for example
if the child survives the parent but it does something else).
Or for a "leak" which unnecessarily uses descriptors and memory in the
child process.
Let's ensure we do not leak it for this component as we do not need it.
Note glibc fopen supports the "e" modifier as an extension to pass
O_CLOEXEC to the open syscall.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
If the process does an exec() (or fork, or..) all descriptors are kept
open by default, unless O_CLOEXEC is set when opening them.
This is usefull for stdin/out/err so that new process is connected to
them, but it is very rare for it to be usefull for any other descriptor.
In general this leads to descriptors being kept open unnecessarily,
which either will block other process from getting them (for example
if the child survives the parent but it does something else).
Or for a "leak" which unnecessarily uses descriptors and memory in the
child process.
Let's ensure we do not leak it for this component as we do not need it.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
If the process does an exec() (or fork, or..) all descriptors are kept
open by default, unless O_CLOEXEC is set when opening them.
This is usefull for stdin/out/err so that new process is connected to
them, but it is very rare for it to be usefull for any other descriptor.
In general this leads to descriptors being kept open unnecessarily,
which either will block other process from getting them (for example
if the child survives the parent but it does something else).
Or for a "leak" which unnecessarily uses descriptors and memory in the
child process.
Let's ensure we do not leak it for this component as we do not need it.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
If the process does an exec() (or fork, or..) all descriptors are kept
open by default, unless O_CLOEXEC is set when opening them.
This is usefull for stdin/out/err so that new process is connected to
them, but it is very rare for it to be usefull for any other descriptor.
In general this leads to descriptors being kept open unnecessarily,
which either will block other process from getting them (for example
if the child survives the parent but it does something else).
Or for a "leak" which unnecessarily uses descriptors and memory in the
child process.
Let's ensure we do not leak it for this component as we do not need it.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
If the process does an exec() (or fork, or..) all descriptors are kept
open by default, unless O_CLOEXEC is set when opening them.
This is usefull for stdin/out/err so that new process is connected to
them, but it is very rare for it to be usefull for any other descriptor.
In general this leads to descriptors being kept open unnecessarily,
which either will block other process from getting them (for example
if the child survives the parent but it does something else).
Or for a "leak" which unnecessarily uses descriptors and memory in the
child process.
Let's ensure we do not leak it for this component as we do not need it.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
If the process does an exec() (or fork, or..) all descriptors are kept
open by default, unless O_CLOEXEC is set when opening them.
This is usefull for stdin/out/err so that new process is connected to
them, but it is very rare for it to be usefull for any other descriptor.
In general this leads to descriptors being kept open unnecessarily,
which either will block other process from getting them (for example
if the child survives the parent but it does something else).
Or for a "leak" which unnecessarily uses descriptors and memory in the
child process.
Let's ensure we do not leak it for this component as we do not need it.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
If the process does an exec() (or fork, or..) all descriptors are kept
open by default, unless O_CLOEXEC is set when opening them.
This is usefull for stdin/out/err so that new process is connected to
them, but it is very rare for it to be usefull for any other descriptor.
In general this leads to descriptors being kept open unnecessarily,
which either will block other process from getting them (for example
if the child survives the parent but it does something else).
Or for a "leak" which unnecessarily uses descriptors and memory in the
child process.
Let's ensure we do not leak it for this component as we do not need it.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
If the process does an exec() (or fork, or..) all descriptors are kept
open by default, unless O_CLOEXEC is set when opening them.
This is usefull for stdin/out/err so that new process is connected to
them, but it is very rare for it to be usefull for any other descriptor.
In general this leads to descriptors being kept open unnecessarily,
which either will block other process from getting them (for example
if the child survives the parent but it does something else).
Or for a "leak" which unnecessarily uses descriptors and memory in the
child process.
Let's ensure we do not leak it for this component as we do not need it.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
If the process does an exec() (or fork, or..) all descriptors are kept
open by default, unless O_CLOEXEC is set for them.
(This default behaviour is usefull for stdin/out/err so that new process
is kept connected to them, but it is very rare for it to be usefull for
any other descriptor)
In general this leads to descriptors being kept open unnecessarily,
which either will block other process from using them (for example
if the child survives the parent but it does something else while keeping
a port open).
Or for a "leak" which unnecessarily uses descriptors and memory in the
child process.
Let's ensure we do not leak it for this component as we do not need it.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
