Add boolean properties to the binding for the nRF RADIO peripheral
to indicate whether the following features are supported:
- IEEE 802.15.4 mode
- 2 Mbps BLE mode
- coded BLE PHY
- high TX power settings
Set these properties appropriately in devicetree radio nodes for all
nRF SoCs.
Add also such properties and set them in appropriate nodes for nRF
flash controllers to indicate whether they support partial erase.
Signed-off-by: Andrzej Głąbek <andrzej.glabek@nordicsemi.no>
Change dependencies of the Kconfig options that enable particular nrfx
drivers so that the drivers and their instances are available when
a devicetree node with a corresponding compatible property exists,
no matter if it is enabled or not.
Change the options that enable compilation of multi-instance drivers
to be promptless helpers only (not directly configurable for users),
as they need to be enabled only together with some options that enable
particular instances of a given driver, and those options select them
as needed.
Signed-off-by: Andrzej Głąbek <andrzej.glabek@nordicsemi.no>
For consistency with the scheme used for other nRF peripherals,
use the peripheral name that nRF Product Specifications use.
In this case, it is WDT, not WATCHDOG.
Also remove the requirement for the label property in the binding.
It is no longer needed.
Signed-off-by: Andrzej Głąbek <andrzej.glabek@nordicsemi.no>
The function in its current form is confusing because unlike other
similarly named functions (dt_nodelabel_has_prop(), dt_node_has_prop())
or devicetree macros (DT_NODE_HAS_COMPAT(), DT_NODE_HAS_PROP()), this
function takes into account the status of the checked node and returns
"y" only when the node is enabled.
This commit redefines dt_nodelabel_has_compat() so that it no longer
checks the node status, and for cases where the previous functionality
is needed, a new function named dt_nodelabel_enabled_with_compat()
is introduced as a replacement.
Signed-off-by: Andrzej Głąbek <andrzej.glabek@nordicsemi.no>
Add a function for checking if any node with a given compatible exists,
no matter if its status is "okay" or not.
Signed-off-by: Andrzej Głąbek <andrzej.glabek@nordicsemi.no>
Prevent "unique_unit_address_if_enabled" warnings from being reported
for nRF52 Series SoCs, where certain nodes need to be enabled with
the same base addresses. These can be (depending on a given SoC):
- power@40000000 & clock@40000000
- power@40000000 & clock@40000000 & bprot@40000000
- acl@4001e000 & flash-controller@4001e000
Signed-off-by: Andrzej Głąbek <andrzej.glabek@nordicsemi.no>
Prevent "unique_unit_address_if_enabled" warnings from being reported
for nRF51 Series SoCs, where certain nodes need to be enabled with
the same base addresses. These are:
- power@40000000 & clock@40000000 & mpu@40000000
Signed-off-by: Andrzej Głąbek <andrzej.glabek@nordicsemi.no>
Add missing devicetree nodes representing the following peripherals
in nRF SoCs:
- ACL
- BPROT
- CCM
- COMP/LPCOMP
- CTRLAP
- DCNF
- MPU (nRF MPU peripheral in nRF51 Series, not ARM MPU)
- MUTEX
- MWU
- NFCT
- OSCILLATORS
- POWER (in nRF51 and nRF52 Series)
- PPI
- RESET
- SWI
- USBREG
Add also corresponding bindings and validation of base addresses of
these nodes.
Signed-off-by: Andrzej Głąbek <andrzej.glabek@nordicsemi.no>
This is a follow-up to commit cf6a58d3f6.
Restore the "nordic,nrf-ipc" compatible property in mbox nodes for both
nRF5340 cores and use it together with the new "nordic,mbox-nrf-ipc"
one. This way either the MBOX or the IPM driver can be used for these
nodes without further modifications. This eliminates the need to use
overlays in quite a few cases, so remove all those no longer needed
ones (which are also a bit confusing now as they refer to no longer
existing ipc@2a000 and ipc@41012000 nodes).
Restore also the ipc node label removed in the commit mentioned above,
as the label is used in validation of base addresses of nRF DT nodes.
Signed-off-by: Andrzej Głąbek <andrzej.glabek@nordicsemi.no>
The NXP SDK defines CodeQuickAccess and DataQuickAccess sections
for locating critical items that need faster access. Centralize
the handling of these sections instead of doing it per SOC.
Fixes#44453
Signed-off-by: David Leach <david.leach@nxp.com>
an area without maintainer is still considered active, calling it orphan
is a bit extreme. We have some areas that can be considered "orphaned",
those now will be covered with 'odd fixes' status, meaning that they
might have one or more collaborator and getting some changes from time
to time, but nothing beyond fixes and nobody driving the area beyond
where it is right now.
Even an area with a dedicated maintainer can be have the status of 'odd
fixes', i.e. there is a maintainer but the area is stale and no further
development is happening.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
Each board needs to have a maintainer to allow automatic assignment and
adding reviewers when pull requests concerning the platform are being
submitted.
This is now a placeholder so we can start adding maintainers early, the
data is not being used anywhere yet.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
Indicate wether pointer parameters are input or output. This information
is useful for API users.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Some values are limited to 32-bit, UINT32_MAX can be used to check if
they overflow.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Group all capture APIs together for better readability. Previous to this
patch, get_cycles() as placed at the end.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Do not convert get_cycles_per_sec() failures to -EIO, just propagate the
failure code since the failure is not handled. As a consequence, all
Doxygen docstrings have been updated.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Hide all API ops (typedefs and struct) under INTERNAL_HIDDEN. This is
information required by drivers only, not API clients.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
ESP32C3 requires master init call to enable its clock
gate. Without this, SPI interface may not initialize
properly.
Signed-off-by: Sylvio Alves <sylvio.alves@espressif.com>
Don't leave idle state if soc isn't waked-up by an interrupt.
(We change to check interrupt controller register)
Signed-off-by: Dino Li <Dino.Li@ite.com.tw>
Verify the get_state API call in z_vrfy_can_get_state(). Adjust sizeof()
arguments to match the rest of the file.
Signed-off-by: Henrik Brix Andersen <hebad@vestas.com>
Verify the add_rx_filter API call in z_vrfy_can_add_rx_filter_msgq() as
this is used by the underlying implementation. Remove unnecessary
typecasts.
Signed-off-by: Henrik Brix Andersen <hebad@vestas.com>
Do not attempt to verify that the current thread has access to the void
*user_data argument to z_vrfy_can_send(). The size of the data is not
known and no driver code will try to dereference it (it may not even be
a valid pointer).
Remove unnecessary typecasts from z_vrfy_can_send().
Signed-off-by: Henrik Brix Andersen <hebad@vestas.com>
On dual-core architecture the platform time can be not aligned with
radio time. This happens e.g. for nRF53 devices. Unaligned times imply
mulfunction in CSL windows scheduling.
This PR fixes it by adding/subtracting the time offset in functions
which return the plaform time. The changes have no impact on platforms
where the times are the same.
Signed-off-by: Lukasz Maciejonczyk <lukasz.maciejonczyk@nordicsemi.no>
Add build only tests to cover Periodic Advertising ADI
support in advertising state and in Periodic
Synchronization state.
Signed-off-by: Vinayak Kariappa Chettimada <vich@nordicsemi.no>
Fix crash in TF-M when non-secure IRQ is interrupting the secure
processing and using the FPU.
The FPU context must be saved when
ARM_NONSECURE_PREEMPTIBLE_SECURE_CALLS is disabled since an IRQ handler
can still access the FPU registers.
Fixes: #43587
Signed-off-by: Joakim Andersson <joakim.andersson@nordicsemi.no>
A reproducible case to see what happens when the TCP stack runs out
of buffers. It transfers a block, bigger then the number of buffers
available.
Also test by introducing packet loss in the loopback driver.
Signed-off-by: Sjors Hettinga <s.a.hettinga@gmail.com>
To allow for high level robustness tests on protocols, add an interface
to control the packet drop rate. A rate of 0 means no packet dropped, a
rate of 1 means all packets being dropped.
Signed-off-by: Sjors Hettinga <s.a.hettinga@gmail.com>
When run the test_mutex_lock_timeout(), we need more time (>300us)
for the testing thread to finish the job, or we will get a fail
(we didn't run the test before).
Because our event timer doesn't handle the float part, and 32768 is
divisible by 8192 which is closest to kernel tick default 10000, I add
CONFIG_SYS_CLOCK_TICKS_PER_SEC = 8192 for all board of it8xxx2 series.
If the CONFIG_SYS_CLOCK_TICKS_PER_SEC = 32768,
the 10 tick of timeout = 300us.
If the CONFIG_SYS_CLOCK_TICKS_PER_SEC = 8192,
the 10 tick of timeout = 1200us.
So we can get more time to finish the job.
Verified by follow test pattern:
west build -p auto -b it8xxx2_evb tests/kernel/sleep
west build -p auto -b it8xxx2_evb tests/subsys/portability/cmsis_rtos_v2
fixes#43513fixes#42847
Signed-off-by: Ruibin Chang <Ruibin.Chang@ite.com.tw>
ITE RTOS timer HW frequency is fixed at 32768Hz, because this
clock source is always active in any EC mode (running/doze/deep doze).
Signed-off-by: Ruibin Chang <Ruibin.Chang@ite.com.tw>
We don't need to convert the free run clock count,
that will be converted by the kernel
(base on CONFIG_SYS_CLOCK_TICKS_PER_SEC),
so we should return the HW register count value directly.
Signed-off-by: Ruibin Chang <Ruibin.Chang@ite.com.tw>
This is a strange one: The printing code pushes a floating point
register, and is called during the mpu falt. If the floating point
registers are lazily stacked, this fp push can cause another mpu
fault to be pending during the current mpu fault, and tail chained
without returning to PendSV. Since we're already cleaning up the
fp execption reason, we might as well also clean up thisp pending,
spurious mpu exception.
Signed-off-by: Jimmy Brisson <jimmy.brisson@linaro.org>
If an SVC was pending during the stack overflow, it will run
after the return of the memory manage fault. To the SVC's misfortune of
the SVC handler, the it's invariant, that PSP point to the
hardware-stacked context is no longer valid. When the user has a
k_sys_fatal_error_handler that tries to kill the thread that caused a
stack overflow, this manifests as the svc reading the memory of whatever
is on the stack after being adjusted by the mem manage fault handler, and
that leads to unending, spurious hard faults, locking up the system.
This patch prevents that.
Signed-off-by: Jimmy Brisson <jimmy.brisson@linaro.org>
For some reason when the manifest action runs and removes the DNM label,
the Do Not Merge workflow doesn't run automatically. Force it running
whenever the Manifest workflow completes to fix this.
Signed-off-by: Carles Cufi <carles.cufi@nordicsemi.no>
Add several new functions to the IPC service API:
- ipc_service_get_tx_buffer()
- ipc_service_drop_tx_buffer()
- ipc_service_send_nocopy()
- ipc_service_release_rx_buffer()
- ipc_service_hold_rx_buffer()
This set of function is used to support backends with nocopy capability.
Signed-off-by: Carlo Caione <ccaione@baylibre.com>
When NO_OPTIMIZATIONS is set and asynchronous API is used but
HW counting is not enabled then linking fails because of lack of
nrfx_timer code. When optimization is enabled, linker is smart
enough to figure out that nrfx_timer is not used.
Converting decision function from static inline function to macro
which is handled correctly with optimization off.
Signed-off-by: Krzysztof Chruscinski <krzysztof.chruscinski@nordicsemi.no>
