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>
Test was not covering nrf54l15 which had Synth source bitfield
renamed. Changing the test to use Kconfig to determine if
Synth is present.
Signed-off-by: Krzysztof Chruściński <krzysztof.chruscinski@nordicsemi.no>
The commit removes unneeded check for enter_4byte_addr_exist in
the function, as the utility function is supposed to set the address
by parsing enter_4byte_addr parameter, while the enter_4byte_addr_exist
used to exist only for DTS entries that directly set enter-4byte-addr
parameters for a flash node.
The change also moves LOG_DBG reporting enter_4byte_addr before
it is processed, to allow logging what actually function will be working
with, before it does so.
The enter_4byte_addr_exist is removed as it is no longer used anywhere.
Setting of data->flag_access_32bit has been moved in the same code block
as call to the spi_nor_cmd_write, as it can be only consequence of
successful call to that function.
Signed-off-by: Dominik Ermel <dominik.ermel@nordicsemi.no>
Subghzspi instances cannot have any pinctrl configs. This causes a
failure of the power management suspend and resume operations for the
subghzspi instance because no "default" pinctrl is found.
Fix that by skipping the pinctrl parts on subghzspi instances.
At the same time fix a copy and paste in the suspend error message.
Fixes: b567a7db83
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
Temporary solution as there is no such harness as remote.
However console regex are not defined yet.
Signed-off-by: Piotr Kosycarz <piotr.kosycarz@nordicsemi.no>
Add a definition of SUIT storage, so there will be a common source of
the SUIT storage location for both SDFW and scripts generating SUIT
storage areas, assigned to local domains.
Signed-off-by: Tomasz Chyrowicz <tomasz.chyrowicz@nordicsemi.no>
Add new channel: `SENSOR_CHAN_POS_DXYZ`, so that it is
consistent with other 3-axis channels.
Updated pytest, `sensor_shell` & `fake_sensor` accordingly.
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
Update the pytest script to exercise the entire
`sensor_channel_name` table with `parse_named_int()`.
The `gauge_desired_charging_current` is selected because
it is the last one in the table before `all`, `all` is not
used because `sensor get sensor@0 all` doesn't return anything.
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
If `CONFIG_SENSOR_INFO` is enabled, use the `sensor_info`
section to validate that the argument is a sensor before using,
otherwise the shell command will hang the application.
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
The `SENSOR_CHAN_VSHUNT` was added in #60717 but was never
added to the `sensor_channel_name[SENSOR_CHAN_COMMON_COUNT]`
table. Since the length of `sensor_channel_name` is fixed to
`SENSOR_CHAN_COMMON_COUNT`, this means that the index at
`SENSOR_CHAN_VSHUNT` points to `NULL`. When we use the
`sensor get` command for anything bigger than
`SENSOR_CHAN_VSHUNT`, we will deref that `NULL` pointer
when we do `strcmp` in the for-loop of `parse_named_int`.
Fix this by defining `SENSOR_CHAN_VSHUNT` in the table.
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
This patch removes all uses of the adv auto-resume feature in the audio
bsim tests, and instead makes all adv starts explicit.
The auto-resume feature is planned for deprecation. And, explicit
starting of adv makes what happens in the test more explicit as well.
Signed-off-by: Aleksander Wasaznik <aleksander.wasaznik@nordicsemi.no>
This commit should optimize the way the device is allowed
to enter the suspended state. Instead of returning a PM
error code to abort the PM process, the standby power state
is constrained as long as the device is not allowed to enter
suspension. With that approach, acquiring PD_SYS is not needed
when in PM device runtime mode.
Signed-off-by: Ioannis Karachalios <ioannis.karachalios.px@renesas.com>
This commit should optimize the way the device is allowed
to enter the suspended state. Instead of returning a PM
error code to abort the PM process, the standby power state
is constrained as long as the device is not allowed to enter
suspension. With that approach, acquiring PD_SYS is not needed
when in PM device runtime mode.
Signed-off-by: Ioannis Karachalios <ioannis.karachalios.px@renesas.com>
This commit should optimize the way the device is allowed
to enter the suspended state. Instead of returning a PM
error code to abort the PM process, the standby power state
is constrained as long as the device is not allowed to enter
suspension. With that approach, acquiring PD_SYS is not needed
when in PM device runtime mode.
Signed-off-by: Ioannis Karachalios <ioannis.karachalios.px@renesas.com>
New nRF54H20 SoC series expose more ports, e.g. P9, so reserve more bits
for the Port+Pin field.
Signed-off-by: Gerard Marull-Paretas <gerard@teslabs.com>
In some relatively rare conditions, the DFU detach/attach happens to
fast, which cause the host to not notice it. It seems to to be the case
for instance on STM32 when no endpoint are being used.
Adding a 1 ms delay (i.e. one SOF period) between the detach and the
attach fixes the issue.
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
d599e2b670 removed early return from sys_clock_timeout_handler if
current counter value is less than cc_val. It seems that this return is
needed as after the removal some stress tests which were using system
timers heavily started to hang.
Signed-off-by: Krzysztof Chruściński <krzysztof.chruscinski@nordicsemi.no>
Allocate the GPIOTE channels that the Bluetooth Controller needs to use
with the dedicated function provided by the nrfx_gpiote driver instead
of using hard-coded indexes of these channels, as now, when the driver
supports multiple GPIOTE instances, it would be much more difficult to
properly inform the driver which channels should be reserved.
Signed-off-by: Andrzej Głąbek <andrzej.glabek@nordicsemi.no>
This definition is no longer supported by the nrfx_gpiote driver (since
nrfx 3.2.0, reserved channels need to be defined per GPIOTE instance),
so remove it to not mislead users.
Signed-off-by: Andrzej Głąbek <andrzej.glabek@nordicsemi.no>
Several places we simply had
/* Create a non-connectable non-scannable advertising set */
regardless of the type of advertising it was.
Modified the comment to omit the scannable part
and corrected the connectable part.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
Several places we simply had
/* Create a non-connectable non-scannable advertising set */
regardless of the type of advertising it was.
Modified the comment to omit the scannable part
and corrected the connectable part.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
Not necessarily related to the namespacing changes, but just
establishing proper indentation formatting throughout the whole file.
Signed-off-by: Luis Ubieda <luisf@croxel.com>
MCS OTS support doesn't implement OTS checksum and if this is globally
enabled assert is triggered.
[00:00:03.409,484] <dbg> bttester_gap: set_io_cap: io_cap: 0
ASSERTION FAIL [ots_init->cb->obj_cal_checksum]
@ WEST_TOPDIR/zephyr/subsys/bluetooth/services/ots/ots.c:476
Callback for object calculate checksum is not set
This should be reverted when Zephyr issue #72471 is fixed.
Signed-off-by: Szymon Janc <szymon.janc@codecoup.pl>
Since 7eaca455fa, pyserial is required for
building docs because `DeviceAdapter` from the `pytest-twister-harness`
plugin, which is imported by autodoc, now depends on it.
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
By default PSA crypto library copies all the buffers that are passed
in into the heap before further processing them. This is required
when the PSA crypto library is on the secure side and it accepts
buffers from a non-secure code. However this case in Zephyr
is already handled by TF-M, so when the PSA crypto library is
built (CONFIG_MBEDTLS_PSA_CRYPTO_C) then it will be on the
non-secure side. Therefore this malloc copy is useless and
it consumes heap memory. Setting MBEDTLS_PSA_ASSUME_EXCLUSIVE_BUFFERS
at Mbed TLS build time helps in removing this mechanism and
it saves heap memory.
Signed-off-by: Valerio Setti <vsetti@baylibre.com>
This block of code was in the original commit that added
CONFIG_SPI_NOR_IDLE_IN_DPD but later modifications
added acquire_device() and release_device() calls
earlier in spi_nor_configure() and the release_device()
function will have already done the enter_dpd().
Signed-off-by: Mike J. Chen <mjchen@google.com>
The BIG handle for an ISO sync receiver is provided by the host. The
valid range is [0x00..0xEF], and this requires the controller to allow
non-consecutive handle numbering. This means that even if only one sync
set is supported, a BIG handle of ex. value 2 must be supported.
As the code uses indexing and range checking directly by handle, this
needs to be changed. Introducing the handle index, which sequnces the
handles such that index = 0 is always the "first handle", regardless
of value.
Signed-off-by: Morten Priess <mtpr@oticon.com>
This commit adds a custom target to tests/crypto/mbedtls in
order to verify that MbedTLS auto-generated files are up to
date.
Signed-off-by: Valerio Setti <vsetti@baylibre.com>
Instead of selecting legacy MBEDTLS symbols, use corresponding
PSA_WANT ones when possible (note: some legacy symbols do not
have a PSA correspondance).
Signed-off-by: Valerio Setti <vsetti@baylibre.com>
Add "sha3.c" source file that was previously missing from
CMakeLists.txt and that has been added in MbedTLS 3.6 release.
Signed-off-by: Valerio Setti <vsetti@baylibre.com>
Using MBEDTLS_PSA_CRYPTO_CLIENT to guard all PSA_WANT symbols is
not completely correct because:
1. the prefix MBEDTLS suggests that it's something related to
MbedTLS, while actually PSA APIs can be provided also
by other implementations (ex: TFM)
2. there might applications which are willing to use PSA APIs
without using MbedTLS at all. For example computing an hash
can be as simple as writing psa_hash_compute() and, if the
PSA function is provided thorugh TFM, then MbedTLS is not
required at all
Therefore this commit:
- moves MBEDTLS_PSA_CRYPTO_CLIENT to Kconfig.tls-generic since
that symbol belongs to MbedTLS
- adds a new symbol named PSA_CRYPTO_CLIENT as a generic way
to state that there is "some" PSA crypto API provider
- let MBEDTLS_PSA_CRYPTO_CLIENT automatically select
PSA_CRYPTO_CLIENT, since the former is an implementation of
the latter.
Signed-off-by: Valerio Setti <vsetti@baylibre.com>
Now that support for all PSA_WANT symbols is in place, we can
enable MBEDTLS_USE_PSA_CRYPTO when MBEDTLS_PSA_CRYPTO_C is
enabled as well.
Note: this commit also moves USE_PSA_CRYPTO out of CRYPTO_C
dependency in config-tls-generic.h because TLS/DTLS/X509 modules
of MbedTLS can rely on *any* implementation of PSA crypto APIs
not only the MbedTLS one. TFM is for example an alternative
to this.
Signed-off-by: Valerio Setti <vsetti@baylibre.com>
Since PSA has several different symbols defined and since it's
better to keep them aligned with the current MbedTLS release,
a python script is used to generate both:
- Kconfig.psa
- config-psa-generic.h
Signed-off-by: Valerio Setti <vsetti@baylibre.com>
Adding a Kconfig and a header file to control which PSA features
are to be used from MbedTLS:
- new kconfig symbols are placed in a separate header file
(Kconfig.psa) and are guarded by MBEDTLS_PSA_CRYPTO_CLIENT. The
reason for this is that TLS/X509 can either rely on PSA functions
provided by MbedTLS (when MBEDTLS_PSA_CRYPTO_C is defined) or
TFM (when BUILD_WITH_TFM is selected). Therefore we could
not make these new Kconfigs depending on MBEDTLS_PSA_CRYPTO_C.
- by default all PSA symbols are enabled, but they can be
disabled by respective Kconfigs in order to reduce the image
size.
- the new header file (config-psa-generic.h) mimics what
config-tls-generic.h does for MbedTLS builtin symbols: it
enables a build symbol for every Kconfig one. The name is
kept identical in the 2 cases (a part from the initial CONFIG_
in order to simplify the mechanism).
- MBEDTLS_PSA_CRYPTO_CONFIG is finally enabled whenever there
is any PSA crypto provider (either MBEDTLS_PSA_CRYPTO_C or
TFM)
Signed-off-by: Valerio Setti <vsetti@baylibre.com>
