Newlib or Picolibc libraries for LLVM may be compiled or installed from
pre-built sources independently of LLVM itself.

This means that always indicating that TOOLCHAIN_HAS_NEWLIB=OFF and
TOOLCHAIN_HAS_PICOLIBC=OFF are wrong. But it could be just as wrong to
always indicate suport for newlib or picolibc.

Some pre-built LLVM toolchains are provided with default picolibc
support, such as LLVM for Arm embedded, but can also be used with newlib
be installing newlib add-on package.

Unfortunately it's not possible to query LLVM regarding newlib or
picolibc support.

Developers have the option of `-DTOOLCHAIN_HAS_<NEWLIB|PICOLIBC>=ON`,
but this is not widely known and cumbersome to do for each build.

An indication of newlib or picolibc support is the presence of library
specific headers, so to improve current situation we check for library
specific headers, and if those are present we assume support for the
library.

This commit improves the current support for LLVM in Zephyr when
cross-compiling, especially for users of LLVM for Arm embedded.

Signed-off-by: Torsten Rasmussen <[email protected]>
(cherry picked from commit 0274bcb)
(cherry picked from commit ce6eaf3)

Store informations regarding the current Zephyr build.
The following informations are stored during CMake configure:
- Board information
- Application source directory
- Application configuration directory
- Toolchain information
- Devicetree files
- Kconfig config files
- Zephyr version

Signed-off-by: Torsten Rasmussen <[email protected]>
(cherry picked from commit 09faf53)
(cherry picked from commit a9ba90a)

…le used.

Save information regarding SVD file in use in vendor-specific section
of the build info file.

Information is stored under Nordic section.

Signed-off-by: Torsten Rasmussen <[email protected]>
(cherry picked from commit 46a3e61)
(cherry picked from commit 22c1920)

Update build_info() calls to use `PATH` argument when values passed to
`build_info()` are user specified and thereby might use native path
separator, such as a single `\`.

Signed-off-by: Torsten Rasmussen <[email protected]>
(cherry picked from commit 2c1eae2)
(cherry picked from commit 7c1c2f4)

…specific gap fill

Deprecate BUILD_NO_GAP_FILL as it gives a false impression that gap
filling can be disabled in binary files.
Binary files are always gap filled due to the fact they contain no
address information. Only option for binary files is to control the gap
fill pattern. When no gap fill is enabled in binary files, then a
default pattern is used by the tool, which usually is 0x00.

Generally the pattern 0x00 leads to unnecessary flash writes, as a
flash generally contains 0xFF after an erase.
Therefore provide a gap fill pattern Kconfig setting instead, with
default value of 0xFF.

For hex-files, intel hex and s19, then gap filling is generally not
needed but in order to still support cases where gap filling is required
then this commit introduces BUILD_OUTPUT_HEX_GAP_FILL and
BUILD_OUTPUT_S19_GAP_FILL. Both settings are disabled per default.

Signed-off-by: Torsten Rasmussen <[email protected]>
(cherry picked from commit 2e8868c)
(cherry picked from commit d62198c)

The llvm-objcopy doesn't support a gap-fill argument, therefore clear
the property to remove the error:
> llvm-objcopy: error: unknown argument '--gap-fill'

Signed-off-by: Torsten Rasmussen <[email protected]>
(cherry picked from commit 71fd4d9)
(cherry picked from commit b249f97)

…0 bus lib

The nrf wifi drivers include code related to qspi/spi bus. This work
moves bus related files to a new nrf70 bus library which is now
independent of the core driver and facilitlates building applications
directly using the bus APIs, agnostic to enabling wifi configs.

Upstream PR #: 82372

Signed-off-by: Bansidhar P.M <[email protected]>
(cherry picked from commit b515453)

Add ztests for nrf70 bus lib to help validate functional integrity
of the QPSI/SPI interface to nRF70 device via appropriate reads/writes
from host processor to relevant memory blocks of nrf70 device.
Note that this will NOT serve as a memory test for nrf70 device,
rather just a functional verification of the wiring between the host
processor and the nrf70 device.

Upstream PR #: 82372

Signed-off-by: Bansidhar P.M <[email protected]>
(cherry picked from commit bf4df4a)

…nnect

This fixes an issue that occurred if deferred_work is queued by another
procedure before we run bt_conn_set_state(BT_CONN_DISCONNECTED).
bt_conn_set_state yields to the system work queue in bt_conn_tx_notify,
and then deferred_work runs and handles the disconnected callbacks.
bt_conn_set_state then enqueues another deferred_work which calls the
disconnected callbacks again and causes an assert. k_work_cancel_delayable
will clear queued deferred work when we call
bt_conn_set_state(BT_CONN_DISCONNECTED), guaranteeing that the disconnect
callbacks will only be called once.

Upstream PR #: 82543

Signed-off-by: Timothy Keys <[email protected]>
(cherry picked from commit 833b6d2)

…presence

Common symbol is defined by SoCs that support RAM control helper.

Upstream PR #: 82262

Signed-off-by: Nikodem Kastelik <[email protected]>
(cherry picked from commit 2e974fb)

…icit

Updated retained_mem driver variant exposes vendor-specific API,
rather than executing implicitly during boot.

Upstream PR #: 82262

Signed-off-by: Nikodem Kastelik <[email protected]>
(cherry picked from commit 0819c8e)

RAM retention is disabled for all RAM blocks
before entering System OFF to achieve lowest power consumption.
RAM retention can still be enabled using retained_mem driver.

Upstream PR #: 82262

Signed-off-by: Nikodem Kastelik <[email protected]>
(cherry picked from commit b0d7a30)

…argets

Using bare-metal approach for retention configuration
is no longer compatible with sys_poweroff() implementation.

Upstream PR #: 82262

Signed-off-by: Nikodem Kastelik <[email protected]>
(cherry picked from commit 15ffa30)

Updated manifest to include separate security types for 802.1X
in display of scan results when APs have EAP and PMF enabled.
BSS display limit based on user-configured non-zero values
set via CMD_INIT, along with support for few other regulatory
domain updates in firmware.

Upstream PR #: 81514

Signed-off-by: Karun Kumar Eagalapati <[email protected]>

EAP_TLS_SHA256 was being identified as EAP_TLS in display scan
results. Identify the security type in scan results sent by
RPU and display accordingly.

Upstream PR #: 81514

Signed-off-by: Ravi Dondaputi <[email protected]>

i2c_slave tests can now be run on nRF54H20 cpuppr target.

Upstream PR #: 82538

Signed-off-by: Michał Stasiak <[email protected]>

Add module for managing multiple requests for MRAM latency.

Signed-off-by: Krzysztof Chruściński <[email protected]>
(cherry picked from commit 980e0ac)

Add test for MRAM latency.

Signed-off-by: Krzysztof Chruściński <[email protected]>
(cherry picked from commit 20e28d2)

This reverts commit 8027617.

Signed-off-by: Krzysztof Chruściński <[email protected]>

…request no latency

Add CONFIG_MRAM_LATENCY_AUTO_REQ. When option is enabled then module
requests no latency during the initialization. This option might be
useful for cases where we want to achieve maximum performance and
want to avoid controlling MRAM in the code.

Upstream PR #: 82465

Signed-off-by: Krzysztof Chruściński <[email protected]>

…TXCHANNEL

The Kconfig `IEEE802154_SELECTIVE_TXCHANNEL` is added along with
the new capability `IEEE802154_HW_SELECTIVE_TXCHANNEL`.
The new capability of the ieee802154 drivers allows to schedule
CSL transmissions as stated in IEEE 802.15.4-2020 chapter 6.12.2.7
CSL over multiple channels. The benefit of the new API is that
additional call to `ieee802154_radio_api::set_channel()` is not
required. The drivers will switch to the new channel as late as
possible for CSL transmissions thus will not interrupt any reception
that might be in progress until the very late moment when the
transmission actually starts.
This improves reception performance when CSL transmissions are used.

Signed-off-by: Damian Krolik <[email protected]>
Signed-off-by: Andrzej Kuroś <[email protected]>
(cherry picked from commit 2fbba82)

…e transmission.

Added new API function to start modulated carrier wave transmission

Signed-off-by: Przemyslaw Bida <[email protected]>
(cherry picked from commit e0f94f8)

…odule.

Commit add `modcarrier` shell command for Openthread diagnostic mode.
Command can transmit modulated carrier out of device for test purposes.

Signed-off-by: Przemyslaw Bida <[email protected]>
(cherry picked from commit 63e1bb4)

Added pdm0 node to nrf54h20 devicetree with proper bindings.

Upstream PR #: 81588

Signed-off-by: Michał Stasiak <[email protected]>

Added support for DMM in PDM drivers in order to use it
with nRF54H20.

Upstream PR #: 81588

Signed-off-by: Michał Stasiak <[email protected]>

…ption

Add an alternate message output option in which messages are
ended with an additional zero data byte with marker and timestamp.
Can be used to maintain compatibility with certain decoders.

Signed-off-by: Juha Sunnari <[email protected]>
(cherry picked from commit 51bc60e)

fixup! [nrf noup] tree-wide: support NCS Partition Manager (PM) definitions

To read the secure memory we need to use soc_secure_mem_read
function instead of memcpy.

This commit allows to use the soc_secure_mem_read function
if the CONFIG_TRUSTED_EXECUTION_NONSECURE and PARTITION_MANAGER
is enabled.

Signed-off-by: Arkadiusz Balys <[email protected]>

…size option

Adds a --download-buffer-size option to the canopen runner to allow
specifying the buffer size of the SDO download.

Signed-off-by: Fabian Blatz <[email protected]>
(cherry picked from commit 11c83c8)

Signed-off-by: Tomasz Chyrowicz <[email protected]>

This adds support for J-Link tunnels, which run on top of an IP network
and therefore uses the -IP option. J-Link tunnels are identified by a
tunnel: prefix instead of a bare IP address. This change checks for the
presence of such a prefix, and choses the -IP transport option if the
tunnel prefix is found.

This has been tested with J-Link Remote Server v7.98g and the SEGGER
tunnel option.

Signed-off-by: Adam Dunkels <[email protected]>
(cherry picked from commit 221199e)

Signed-off-by: Tomasz Chyrowicz <[email protected]>

Fixes some issues with samples/tests by adding configuration
files to satisfy TF-M requirements

Signed-off-by: Jamie McCrae <[email protected]>

fixup! [nrf noup] tree-wide: support NCS Partition Manager (PM) definitions

Removes a condition in the partition manager enabled check which
was for child/parent images and keeps the sysbuild one only

Signed-off-by: Jamie McCrae <[email protected]>

Write-back cache after copying VPR's code and data to the destination
memory. It allows VPRs to run from RAM regions cacheable by its host.

It allows to use FLPR in nRF54H20.

Upstream PR #: 84104

Signed-off-by: Hubert Miś <[email protected]>

EngB+ uses 32bit bus-width stacking sequence for all VPR cores.

Signed-off-by: Karsten Koenig <[email protected]>
(cherry picked from commit 6201882)

…ing the first timeout

Ref: NCSDK-31290

This ensures that the system clock is correctly updated when the first
timeout is aborted, preventing unexpected early wake-up by the system clock
programmed previously.

Signed-off-by: Dong Wang <[email protected]>
(cherry picked from commit dd5f11c)

… responses

Adds a new error code that can be used to signify that a query was
valid but the response was not valid

Signed-off-by: Jamie McCrae <[email protected]>
(cherry picked from commit 5871037)

Update the HW models module to:
71bcaa88c97977647d387217dab99f7d6f026815

Including the following:
71bcaa8 hal/nrf_cracen_rng.c: Disable until released
69e9037 hal cracen_rng: Add replacements for proposed cracen_rng hal
71c2507 CRACEN RNG: Corrected one comment
dfbc93a 54 ECB: Corrected behaviour with premature end of input joblist
a46101f 54 AAR, ECB, CCM: Correct a few notes and ECB ERRORSTATUS
74aa2de 54L15: Correct IRQ 261 name
e12ab9c Add CRACEN RNG model
91066e7 docs: UART add PTY backend to shortlist of backends

Signed-off-by: Alberto Escolar Piedras <[email protected]>
(cherry picked from commit 86adcb6)
Signed-off-by: Herman Berget <[email protected]>

Update the HW models module to:
1069743a7fc06362e8f8d7f57225b8c7940da9f0

Including the following:
1069743 Fix in cracen hal conditional compile condition
3bbd3bd Enable build of hal/nrf_cracen_*.c
506da1f hal cracen_cm: Add const to prototype
b02d927 BLECrypt_if: For new aes_ecb scramble data if no libCrypto
e41f101 RRAMC nrf hal: Add replacements for new incoming RRAM APIs
4e933cf UART: Do not warn on invalid/unconfigured framesize
716b20c hal cracen_cm: Add replacements for new cracen_cm hal
fc597a4 CRACEN: Add CryptoMaster & CM AES (ECB only)
129f16c CRACEN RNG HAL: Remove senseless static inline

Signed-off-by: Alberto Escolar Piedras <[email protected]>
(cherry picked from commit a2d202f)
Signed-off-by: Herman Berget <[email protected]>

Update the HW models to the provisional branch zephyr_2025_01_14
which enables the use the not yet released CRACEN RNG & CM HAL

Signed-off-by: Alberto Escolar Piedras <[email protected]>
(cherry picked from commit ec17cc0)
Signed-off-by: Herman Berget <[email protected]>

…en HAL

Update the HW models to the provisional branch zephyr_2025_01_22
which enables the use the not yet released CRACEN RNG & CM HAL
while enabling the latest nrf HAL 3.10

Including the following:
4aac556 Support latest nrf HAL 3.10 / MDK 8.69.1 for 54L15 builds
2e75940 UART: Correct UART names printed out to the users

Signed-off-by: Alberto Escolar Piedras <[email protected]>
(cherry picked from commit 492ad96)
Signed-off-by: Herman Berget <[email protected]>

…request/release

Makes 802.15.4 use the API under CONFIG_CLOCK_CONTROL_NRF2
in nrf_clock_control.h to request/release HFCLK.

Previous HFCLK requesting/releasing on nRF54H20 was more
of a workaround and could produce issues when
comes to sharing the resources.

Signed-off-by: Piotr Koziar <[email protected]>
(cherry picked from commit b6ca9db)

…4 RD

Selects CLOCK_CONTROL config whenever nRF 802.15.4 Radio Driver
is enabled.

The driver requires HFCLK api available under the config switch.

Signed-off-by: Piotr Koziar <[email protected]>
(cherry picked from commit f6892a4)

Some time ago three Kconfigs dedicated for the child timeouts
configuration were added, but changing them does not apply when
using OpenThread libraries. Added setting these values using
openthread API on openthread start.

Signed-off-by: Kamil Kasperczyk <[email protected]>
(cherry picked from commit 2534dc1)

nrf-squash! [nrf noup] boards: Add non-secure target for nrf54L15dk

Enables forcing use of sysbuild to build nrf54l15dk
non-secure board target when using twister

Signed-off-by: Grzegorz Chwierut <[email protected]>

If BLE has already been enabled, `bt_enable` returns immediately
and does not call passed callback function.
In this case, call `bt_conn_cb_register` to ensure that callbacks
are registered correctly.

Signed-off-by: Maciej Baczmanski <[email protected]>
(cherry picked from commit 1bd8e44)

This fixes the compiler warning -Wstringop-truncation in
version 13 of gcc.

Signed-off-by: Timothy Keys <[email protected]>
(cherry picked from commit 614f55f)

This commit adds a test that checks if disconnecting
and restarting the IPC session works correctly. The
test is also focused on the "unbound" callback.

Signed-off-by: Dominik Kilian <[email protected]>
Co-authored-by: Radoslaw Koppel <[email protected]>

In some cases, CPUs that may need to reset or
temporary stop communication. This commit adds "unbound"
functionality that provides a callback to IPC service user
when connection was interrupted for some reason, e.g.
expected or unexpected CPU reset, closing the
endpoint. The "unbound" callback is optional to implement
by endpoints. This commit implements it in the ICMsg
backend.

Signed-off-by: Dominik Kilian <[email protected]>

Use hal for checking cache state.

Signed-off-by: Krzysztof Chruściński <[email protected]>

Add missing settings in LRCCONF needed before soft off.
Add releasing GRTC request.

Add enabling of DCACHE and ICACHE memory blocks during init. It is a
default register value but we can enter init when waken up from soft off
and in that case values there will coming from power off sequence and not
from HW device reset.

Signed-off-by: Krzysztof Chruściński <[email protected]>

Add option to not go to system off but instead go to sleep.
Extending sample to use second button (Button 1). Sample goes
to system off as before and when Button 0 is pressed it wakes
up and goes to system off again. If Button 1 is pressed then
it wakes up but goes to sleep again.

Signed-off-by: Krzysztof Chruściński <[email protected]>

Add support to nrf54h20dk/nrf54h20/cpuapp and cpurad in the sample.

Signed-off-by: Krzysztof Chruściński <[email protected]>