Add ipm driver to use Inter Processor Interrupts
on Xilinx ZynqMP platform. This patch also adds sample
application that shows use of xlnx ipm driver.
This driver uses default arm gic interrupt controller
and works only for lockstep mode of cortex-r5f
cluster for now.
In split mode the cortex-r5 cluster will
have two r5f cores and they are expected to work in AMP
mode. If both r5f cores run simultaneouly, only one of
the core is able to receive IPI interrupts at this time
and it will be the one that started later. In future
this limitation shall be removed.
Signed-off-by: Tanmay Shah <tanmay.shah@amd.com>
openamp shared memory regions can be anywhere in DDR memory
withing 2G range. Current SRAM_PRIV region is 64M which
prevents access of shared memory (vrings) by RPU (cortex-r5)
if it is out of 64M range. This patch allows vrings to be in DDR within
2G address space.
Developed-by: Dan Millea
Commited by: Tanmay Shah
Signed-off-by: Dan Milea <dan.milea@windriver.com>
Signed-off-by: Tanmay Shah <tanmay.shah@amd.com>
the dev pointer is already dereferenced before this function is
called, so this check does not make any sense.
Signed-off-by: Jacob Siverskog <jacob@teenage.engineering>
Add `yd_esp32` board:
- Model name: YD-ESP32
- Manufacturer: VCC-GND® Studio
- Espressif module: ESP32-WROOM-32E
Signed-off-by: Julio Cesar <hi@jcsx.dev>
There is no reason why the capabilities
should be modified by the stack after registration,
nor any reason why a the unicast client application
needs to modify the reported remote capabilities, so
all have been marked as const.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
Refactor the bt_audio_codec_cap to use flat arrays to store
metadata and codec specific capabilities.
The purpose of this is to make it easier to copy the data
between layers, but also to support non-LTV data for non-LC3
codec capabilities.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
This diff removes a build assert that no longer makes
sense due to the fact the MAX_IRQ_PER_AGGREGATOR number
does not need to fit into only the first level interrupt
bits. Before the extension of the interrupt bit masks if
the MAX_IRQ_PER_AGGREGATOR bits was larger than the first
level of interrupts then the first level bits could overflow
into the second level. This is no longer necessary to check
since the user can set the correct number of bits needed per
level.
Signed-off-by: Joshua Lilly <jgl@meta.com>
Instead of sending output strings to stdout via fputs(), just
print them using printk(). This allows the output to be detected
by twister.
Fixes#60676
Signed-off-by: Peter Mitsis <peter.mitsis@intel.com>
Various fixes to LwM2M doxygen documentation
- document all macros
- mount path helpers in lwm2m_api doxygen group
- hide internal macros
- created proper header definitions for macros that belong together
(ex. result codes)
- fixed capitalization for Objlnk
Signed-off-by: Benjamin Cabé <benjamin@zephyrproject.org>
Following #56576, the `cs` field in `struct spi_config` is of type
`struct spi_cs_control` instead of a pointer to the same type.
This PR updated the driver to use `spi_cs_is_gpio()` helper to
check if SPI CS is controlled using a GPIO.
Signed-off-by: Manuel Argüelles <manuel.arguelles@nxp.com>
This adds refactor of bt_vocs object that is common interface for client
and server commands. As the bt_vocs_client is significantly larger than
bt_vocs_server the memory usage is not optimized when bt_vocs has an
union of those. Thus RAM consumption has been minimized by using
CONTAINER_OF macro to reference either to bt_vocs_client or
bt_vocs_server instance that bt_vocs is a member of.
Signed-off-by: Mariusz Skamra <mariusz.skamra@codecoup.pl>
The VOCS API uses `bt_vocs` object to reference both client and server
instance. This adds missing instance type checks to avoid invalid API
usage.
Signed-off-by: Mariusz Skamra <mariusz.skamra@codecoup.pl>
Time between the start of 2 consequtive advertising events consists
of advertising interval and delay. If delay is zero and the measured
interval is the interval between the last or prelast packet and the
first packet of the next adv event, the measured time can be less than
adv interval.
Signed-off-by: Pavel Vasilyev <pavel.vasilyev@nordicsemi.no>
When settings_load() is not called from system workqueue, but still
called from a cooperative thread, rescheduling during the settings
loading can wake up system workqueue and try to register the Mesh
Proxy Service, which can not be allowed yet. This is the case for bsim
tests.
Postpone registering the service with some small timeout.
Signed-off-by: Pavel Vasilyev <pavel.vasilyev@nordicsemi.no>
This fixes possible buffer overflow that may happen when `buf->len`
equals `buf->size` while adding null terminator.
Signed-off-by: Mariusz Skamra <mariusz.skamra@codecoup.pl>
The font existence is checked at the time of linking,
The check at initialization is no longer necessary.
Signed-off-by: TOKITA Hiroshi <tokita.hiroshi@fujitsu.com>
The nRF HW models have been updated to include the following
fixes and improvements:
* TIMER: Fix bogus events/interrupts if shorts where used
* PPI: Handle well an event triggering a task twice
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
When building the 32bit native board targets variants
for x86(-64) hosts, gcc will promote float literals to double
(See https://gcc.gnu.org/bugzilla/show_bug.cgi?id=92875 )
This can result in unexpected comparison differences.
This is due to the compiler using the 8087 float mode by
default.
Instead let's tell the compiler to use the SSE float path,
which is the default for 64 bit x86-64 builds.
The assumption that any x86 host used for development
will have SSE support should be safe enough.
For more background see
https://github.com/zephyrproject-rtos/zephyr/issues/61345
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
This change introduces a new PWM driver for all CC13/26xx SoC.
See the documentation in ti,cc13xx-cc26xx-timer-pwm.yaml for detailed
usage instructions.
Signed-off-by: Florian Grandel <fgrandel@code-for-humans.de>
According to section 4.2.2.4, Compositiion Data page 128 shall be
present if the node supports Remote Provisioning Server (section
4.2.2.4).
Signed-off-by: Pavel Vasilyev <pavel.vasilyev@nordicsemi.no>
According to sections 7.2.2.2.4 and 7.2.2.2.5 of MshPRTd1.1v20,
when starting Private Node Identity or Private Network Identity, the
node shall use either RPA or NRPA and the advertising address. The
address shall be regenerated when Random field is regenerated.
Use BT_LE_ADV_OPT_USE_NRPA options to make advertiser generate and use
NRPA on every start of advertising Private Node Identity or Private
Network Identity.
Signed-off-by: Pavel Vasilyev <pavel.vasilyev@nordicsemi.no>
Currently zephyr_code_relocate() will exit with an error if the
library does not exist. This commit adds a macro that only calls
zephyr_code_relocate() when a specific Kconfig symbol is enabled.
Signed-off-by: Andriy Gelman <andriy.gelman@gmail.com>
This removes the scanning and PA sync capabilities from the
broadcast sink implementation, moving the responsibility of this
to the application layer.
The reason for this is to add more flexibility and choice to the
upper layers, while simplifying the Broadcast Sink implementation.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
The broadcast sink API has been modified so that all scanning, etc.
will be done by the application, rather than the stack. This commit
updates the sample to use this updated API.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
The broadcast sink API has been modified so that all scanning, etc.
will be done by the application, rather than the stack. This commit
updates the sample to use this updated API.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
The broadcast sink API has been modified so that all scanning, etc.
will be done by the application, rather than the stack. This commit
updates the tests to use this updated API.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
The broadcast sink scan API is no longer used by the BAP
shell, which handles everything related to PA sync and scan
itself now.
It has also been made easier to sync to a broadcast
source by simply using create_broadcast_sink.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
The number of subgroups the scan delegator supports should
match the number of subgroups the broadcast sink supports.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
Add support for selecting a periodic advertising
sync object for use. This will also be available outside
the scope of bt.c via the extern declaration.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
This will effectively test both the BAP and CAP TX sync functions
as well as the CAP send function.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
Modify audio_send_timeout to use bt_bap_stream_get_tx_sync
instead of bt_iso_chan_get_tx_sync to avoid using an internal
function to get the ISO channel from the BAP stream.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
The @note in bt_bap_stream_send did not make any sense, and has
been replaced with a new note.
Signed-off-by: Emil Gydesen <emil.gydesen@nordicsemi.no>
Configure ENET_REF_CLK direction as input when
CONFIG_ETH_MCUX_RMII_EXT_CLK is set to allow Ethernet external clock
usage.
Signed-off-by: Benjamin Lemouzy <blemouzy@centralp.fr>
This is following the same convention that was recently introduced to Alder
Lake boards (intel_adl). Additionally, an "S" suffix is added to the
boards, since what's currently supported is in fact the Raptor Lake S
variant.
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
A sneaky whitespace was causing inconsistent rendering in the list of
available overlays.
Fixed by switching to a list table which is arguably now even easier to
read.
Signed-off-by: Benjamin Cabé <benjamin@zephyrproject.org>
Add a config entry specifically for RISC-V runtime.
Memory usage is 7066 bytes.
Also adjust some library options. This make the ARC runtime binary
smaller as well:
before:
FLASH: 8808 B
SRAM: 3672 B
after:
FLASH: 5596 B
SRAM: 3672 B
Signed-off-by: Nicolas Pitre <npitre@baylibre.com>
Update ARC support status page as now we have scalar ARC VPX
port upstreamed.
Signed-off-by: Eugeniy Paltsev <Eugeniy.Paltsev@synopsys.com>
Signed-off-by: Evgeniy Paltsev <PaltsevEvgeniy@gmail.com>
