If OA read fails, then the essential thread is terminated.
Correct the behavior so that the driver re-tries the read and
continues to work.
Refactor offload thread OA and Generic SPI parts into
if/else statement.
Add missing `is_adin2111` to OA port 2 RX status check.
Signed-off-by: Georgij Cernysiov <geo.cgv@gmail.com>
Fix to setup correct short prepare timeout when preempt
callback does not find the prepare in the head of the
pipeline. In this case, find the short prepare and setup
a fresh preempt timeout for the determined short prepare
present in the pipeline.
Relates to commit 7d1bc1789e ("Bluetooth: Controller: Fix
short prepare preempt timeout start").
Signed-off-by: Vinayak Kariappa Chettimada <vich@nordicsemi.no>
Fix delayed ULL scheduling an auxiliary PDU reception by
skipping it. When many bufferred ADV_EXT_IND are to be
processed, the ULL scheduling could get late and then cause
the scan aux LLL to detect increased overhead start timing.
Signed-off-by: Vinayak Kariappa Chettimada <vich@nordicsemi.no>
Fix to reschedule at window start when there are no more
tickers in the list to check for overlap.
Signed-off-by: Vinayak Kariappa Chettimada <vich@nordicsemi.no>
Remove ticks_elapsed use in reschedule_in_window as it has
already been used in the bottom half processing.
Signed-off-by: Vinayak Kariappa Chettimada <vich@nordicsemi.no>
During use of tests/bluetooth/shell application to perform
extended scanning with periodic advertising and broadcast
audio source devices around, it was observed that the
application crashed due to too low BT_RX_STACK_SIZE, hence
increase by 128 bytes.
Signed-off-by: Vinayak Kariappa Chettimada <vich@nordicsemi.no>
Fix implementation of single switch timer minimum compare
value requirement. Do not always add the maximum possible
radio latency duration, but check if the required compare
value is smaller, then use the remainder required as the
extra compare value. This fix reduces latencies to setup
radio receptions and fixes an assertion in lll_scan_aux
when radio_tmr_start_us() is checked for latencies.
Relates to commit bcd28e0a86 ("Bluetooth: Controller:
Fix sw switch single timer for spurious TXEN/RXEN").
Signed-off-by: Vinayak Kariappa Chettimada <vich@nordicsemi.no>
Consistently use the nRF54L15 as the non-TF-M integration platform
in addition to `native_sim`, and the nRF9151 as the TF-M one.
In addition, add the `integration_platforms` specification to the
tests that were missing it.
Signed-off-by: Tomi Fontanilles <tomi.fontanilles@nordicsemi.no>
add `ALLOW_EMPTY TRUE` property to silence cmake warnings in case
when no sources are added to subsys__net library
Signed-off-by: Maciej Baczmanski <maciej.baczmanski@nordicsemi.no>
Seen when having multiple network interfaces that the sock_error
was 0 even if the socket was serviced properly. So if SO_ERROR
returns 0, just ignore it.
Signed-off-by: Jukka Rissanen <jukka.rissanen@nordicsemi.no>
We must register all listening sockets under one file descriptor
array when calling dns_dispatcher_register() which then calls
net_socket_service_register(). The socket services expects all
the registered sockets under one service context to be in one
array. If this is not done, the latest socket array wins and
the earlier registrations are forgotten.
Signed-off-by: Jukka Rissanen <jukka.rissanen@nordicsemi.no>
As the DNS might listen to multicast addresses (like in mDNS) in
different network interfaces, make sure to check the network
interface index when registering the dispatcher context. This
allows two mDNS registrations to more than one network interface.
Signed-off-by: Jukka Rissanen <jukka.rissanen@nordicsemi.no>
Initial support for SSD1322 OLED display driver. Only 1 bit color mode is
supported.
Most options map directly to values documented in the datasheet,
except segments-per-pixel, which I had to add to support the Newhaven
NHD-2.7-12864WDW3. This is a slightly odd feature, but in practice
it is a lot nicer to support it in the driver, and since we're currently
remapping pixels anyway, it makes sense to implement here.
This driver also has a configurable buffer size for the pixel conversion.
By using a larger buffer, we can potentially use DMA for the SPI transfer.
The default is set to 8, which is the smallest value that supports
segments-per-pixel = 2
Initial driver implementation by Lukasz Hawrylko <lukasz@hawrylko.pl>.
Additional options and configurability by Tobias Pisani <mail@topisani.dev>
Signed-off-by: Lukasz Hawrylko <lukasz@hawrylko.pl>
Signed-off-by: Tobias Pisani <mail@topisani.dev>
Co-authored-by: Lukasz Hawrylko <lukasz@hawrylko.pl>
Co-authored-by: Tobias Pisani <mail@topisani.dev>
Like 2ca179c, check and warn if incoming HCI event parameters exceed
the specification. This helps debugging controllers by detecting the
out-of-spec value that shouldn't appear.
Signed-off-by: Sungwoo Kim <iam@sung-woo.kim>
The RxFIFO is already flushed any time a transceive operation
finishes, there is no need for doing it also before the
transaction start.
The aim of this change is to simplify the logic and to
(potentially) reduce the minimum time between transactions.
Signed-off-by: Daniel Gaston Ochoa <dgastonochoa@gmail.com>
Fix errors found in CI for missing clock_t and clockid_t types
when building the tests/posix/net testsuite against newlib.
Also use angle-brackets and prefixed path with
`<zephyr/posix/posix_types.h>` since it is not in the default
search path and is analagous to `<sys/types.h>`.
Signed-off-by: Chris Friedt <cfriedt@tenstorrent.com>
Adds support for cooperative locking/unlocking the VPX vector registers.
Provided that all VPX enabled threads use these routines to control
access to the VPX vector registers, it will allow multiple threads to
safely use them without the need for saving/restoring them upon each
context switch.
Signed-off-by: Peter Mitsis <peter.mitsis@intel.com>
To check if DVFS can be used, the CONFIG_NRFS_DVFS_LOCAL_DOMAIN symbol
needs to be used, not CONFIG_NRFS_HAS_DVFS_SERVICE which only indicates
that DVFS is technically possible, not that its local domain part is
actually included in the build.
Signed-off-by: Andrzej Głąbek <andrzej.glabek@nordicsemi.no>
Current nPM1300 charger driver does not work with batteries
without NTC thermistor. Added supported for this feature.
Signed-off-by: Javier Santos <jasr93@outlook.es>
This commit adds support for INA236 into the existing INA230 driver.
These two chips are similar enough to share most of the code.
The device can be defined the same way as INA230 and we only have
the extra option to select the high-precision mode.
```
ina236: ina236@40 {
status = "okay";
compatible = "ti,ina236", "ti,ina230";
reg = <0x40>;
adc-mode = "Bus and shunt voltage continuous";
vbus-conversion-time-us = <1100>;
vshunt-conversion-time-us = <1100>;
avg-count = <1>;
current-lsb-microamps = <1000>;
rshunt-micro-ohms = <15000>;
alert-gpios = <&gpiod 0 GPIO_ACTIVE_LOW>;
high-precision;
};
```
Signed-off-by: Tomáš Juřena <jurenatomas@gmail.com>
Enforce some minimum delay between enabling the voltage divider with a
GPIO and sampling the analog voltage. Without this delay the ADC can
easily sample the transient power up curve instead of the steady state
output.
Signed-off-by: Jordan Yates <jordan@embeint.com>
Implement the parallel mode in the powertrain switch TLE9104.
This allows that OUT1 and OUT2 are controlled together, as well
as OUT3 and OUT4.
Signed-off-by: Benedikt Schmidt <benedikt.schmidt@embedded-solutions.at>
Inform the executor of a submissions completion with -ENOMEM
if the size of the workq is not big enough.
Signed-off-by: Florian Weber <Florian.Weber@live.de>
This function can be used to parse HCI-formatted subevent results
regardless of whether they come from CS Test or not
Signed-off-by: Olivier Lesage <olivier.lesage@nordicsemi.no>
Add some defines for packet quality and rssi for RTT,
these are useful for filtering out steps where the RTT exchange
was unsuccessful
Signed-off-by: Olivier Lesage <olivier.lesage@nordicsemi.no>
These options are in microseconds but not fully configurable,
document what options are allowed in 6.0
Signed-off-by: Olivier Lesage <olivier.lesage@nordicsemi.no>
It's more common (and more readable) convention to use lower-case
names for string-based device tree property values. Convert the HCI bus
and quirks to follow this convention. Also take advantage of the
recently added support for string-array enums to enforce that the
correct values are used.
Signed-off-by: Johan Hedberg <johan.hedberg@silabs.com>
To enable breakpoints on this board, remove the `dgdb_memory_map disable`,
which is a leftover from early development steps.
Signed-off-by: Erwan Gouriou <erwan.gouriou@st.com>
Georgij Cernysiov