Spilt the cavs server and FW loader, to make the cavstool.py
a pure FW loader and runner as it used to be. The reasons is
try to keep the FW loader won't be affected by the client-
server-based HW service as possible, and more easy to debug.
After splitting them, the service program spawns another
process to run the FW loader. The advantage is no matter when
the FW loader is stuck, the service can detect and stop it,
then continue to next test.
Fixes#47652.
Signed-off-by: Enjia Mai <enjia.mai@intel.com>
Copy cavstool.py to remote-fw-service.py in order to track changes
for following splitting work of cavstool.py.
Signed-off-by: Enjia Mai <enjia.mai@intel.com>
Limit allowed build platforms.
Building sample.bootloader.mcuboot for many platforms
is not possible (for instance a qemu).
The limit is need as otherwise zephyr-rtos/zephyr CI is
failing on any push to main branch or nightly CI run.
Signed-off-by: Andrzej Puzdrowski <andrzej.puzdrowski@nordicsemi.no>
This commit updates the twister workflow to check out the west modules
before running the test plan script for the pull request runs because
the twister test plan logic resolves the the module dependencies and
filters tests based on the local availability of the required modules.
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
Missing include to compile correctly when the
CONFIG_PM && IS_UART_WAKEUP_FROMSTOP_INSTANCE are defined
Signed-off-by: Francois Ramu <francois.ramu@st.com>
this changes adv_new_legacy to adv_get_legacy.
without this fix the function would return NULL if BT_EXT_ADV is enabled.
Signed-off-by: Jonas Woerner <jonas.woerner@online.de>
This update contains the version update which was reverted plus
relevant bugfixes.
The revert take places here: #50120
Synch up to the mcu-tools/mcuboot SHA:
73d69e9b566daac1a91724aa1237bfc1b9d51fa3
- Switched zephyr port from using FLASH_AREA_ macros to FIXED_PARTITION_
macros
- Made flash_map_backend.h compatible with a C++ compiler
- Enabled watchdog feed by default
- Allowed to get the flash write alignment basing on the zephyr,flash
DT chosen node property
- fixed watchdog device typos
- serial recovery: fixed build failure if CONFIG_BOOT_SERIAL_DETECT_PIN
has undefined value
Signed-off-by: Andrzej Puzdrowski <andrzej.puzdrowski@nordicsemi.no>
Sharing the same pool for both, local and relayed messages has some
disadwantages. In a high-traffic environment the relayed messages may
occupy the whole pool without letting the node to send own messages,
neither to process segmented messages. The vice versa situation is
possible when local messages takes the whole pool so that the node is
not able to relay other messages. Another thing is that it doesn't let
to controll amount of local and relayed messages that the node can send
at once. Controlling the number of messages that the node can relay may
help to reduce the message latency.
This change adds separate net_buf_pool for the relayed messages that
helps to avoid problems described above.
Signed-off-by: Pavel Vasilyev <pavel.vasilyev@nordicsemi.no>
This PR removes information regarding previous
west extension related to OpenOCD download.
This extension is no longer available and manual download
is required until upstream OpenOCD receives
all xtensa and riscv patches.
Signed-off-by: Sylvio Alves <sylvio.alves@espressif.com>
Add BUILD_ASSERT to check max val of ngpios. If someone sets ngpios to
more than 32, and acts on pin 32 or higher, memory corruption will
result. This is only a problem with the emulator, real hardware
will have limits that are probably much smaller.
Signed-off-by: Jeremy Bettis <jbettis@google.com>
Since db11fcd174 [net/net_pkt: Fully
separate struct net_pkt from struct net_buf], net_pkt_compact() can not
fail anymore.
This commit acknowledges this and simplifies the API accordingly.
Signed-off-by: Reto Schneider <reto.schneider@husqvarnagroup.com>
This update introduces some fixes required to build with Zephyr 3.1
and the associated Zephyr SDK update.
It also enables running the zscilib test suite as part of the normal
CI build process, referencing the 'tests' folder in `module.yml`.
Signed-off-by: Kevin Townsend <kevin.townsend@linaro.org>
After magic is read during bound procedure, trigger an
initial data read. This is to ensure data is not missed
when piggybacked on magic's exchange interrupt.
Signed-off-by: Pawel Dunaj <pawel.dunaj@nordicsemi.no>
Add bsim tests to check tx callback behavior for advertiser.
Feature coverage:
TX Complete callback execution
Callback order
Packets are freed before the end callback is called
Signed-off-by: Aleksandr Khromykh <aleksandr.khromykh@nordicsemi.no>
The peripheral_hids sample currently implements the HID descriptor and
characteristics, but no notify report. Add some code for implementing a
simple one button click report for boards that have sw0 defined, so that
the sample actually does something beside connecting and enumerating the
device.
Signed-off-by: Fabio Baltieri <fabiobaltieri@google.com>
The default report_map for peripheral_hids mouse sample descriptor is
missing a report id field. This seems to be confusing Android hosts into
somehow parsing reports incorrectly.
Adding a report id seems to fix the problem and still works correctly on
bluez hosts.
Signed-off-by: Fabio Baltieri <fabiobaltieri@google.com>
The return values of esp_wifi_set_mode(), and esp_wifi_set_config() were
overwritten once esp_wifi_start() is executed.
This commit fixes that by ORing the return value to the others.
Signed-off-by: Thomas Stranger <thomas.stranger@outlook.com>
PR #49550 missed adding proper RTCIO pin mapping
in a few rtcio functions calls, which causes board
to crash. This fixes it.
Closes#50238
Signed-off-by: Sylvio Alves <sylvio.alves@espressif.com>
GCC doesn't actually need this treatment, only clang. (GCC is smart
enough to use "const" as the needed clue to suppress instrumentation
it seems like). And it turns out older gcc's emit a warning
(attribute ignored, which is... sort of weird to warn about?) in that
circumstance, which breaks builds. Just turn it off rather than try
to find a way to suppress that one warning.
Signed-off-by: Andy Ross <andyross@google.com>
It turns out that SOF is already using a symbol named
"zephyr_app_main()", so this produces a collision. Pick something
that looks more relevant to "posix", and put an underscore on it (it's
a "system" symbol, after all).
Signed-off-by: Andy Ross <andyross@google.com>
It seems like libfuzzer wants to relocate 32 bit instrumented code
sections at runtime at addresses different than the ones in the ELF
file. This is problematic, because Zephyr files are compiled
statically and so will crash the first time they try to jump to an
absolute .text address (basically at the first function call after a
fuzzer entry point).
It seems that building with -fPIC is enough to defeat this (we use the
host linker script, which will manage the GOT/PLT entries for us),
which will work as long as the fuzzer isn't playing games with data
other than text. None of this seems to be documented, so... I guess
it's as good as we can get. It works, at least.
(x86_64 binaries don't show the same behavior, they run where they
were linked)
Signed-off-by: Andy Ross <andyross@google.com>
With the move to start/stop functions the transceiver is not expected
to be enabled after driver initialization.
Signed-off-by: Thomas Stranger <thomas.stranger@outlook.com>
This overlay shows the LPUART at 9600 with wakeUp capability
to exit the nucleo_wl55jc target plaform from the low Power
stop mode, EXTI_Line 28 is the corresponding wakeUp source.
The LPUART is sourced by the LSE (32768Hz) and max baudrate
is 9600. Set the console to 9600.
Signed-off-by: Francois Ramu <francois.ramu@st.com>
The USART or LPUART, especially on stm32wl, must set the EXTI
lines corresponding to the wakeUp source.
This LL_EXTI_LINE_xx is given by the Device Tree property.
This makes the system exiting lowpower stop mode on EXTI irq.
The LL_EXTI_LINE_x is exactly the bit(X) set.
Signed-off-by: Francois Ramu <francois.ramu@st.com>
When the EXTI Line nb is not set as interrupt by default
(at reset value), the uart instance cannot waekup the
system from its low power stop mode.
The EXTI line must be specified, like with stm32WL55
Signed-off-by: Francois Ramu <francois.ramu@st.com>
When necessary, enable QSPI and Modem drivers
by default if relevant higher level kconfigs are used in an app.
Signed-off-by: Ryan Erickson <ryan.erickson@lairdconnect.com>
Removed duplicated configuration items for CONFIG_ARM64_VA_BITS and
CONFIG_ARM64_PA_BITS, and modify them to be 36bit as SoC supports up
to 34-bit address.
Signed-off-by: Jiafei Pan <Jiafei.Pan@nxp.com>
Extracted common function for erase the image trailer which
decrease code size and increase legibility.
Signed-off-by: Andrzej Puzdrowski <andrzej.puzdrowski@nordicsemi.no>
Inappropriate page was erased during test as its absolute offset was
calculated with omission of the the fixed-partition offset to which
it belongs.
fixes#48115
Signed-off-by: Andrzej Puzdrowski <andrzej.puzdrowski@nordicsemi.no>
When CONFIG_ZTEST_SHUFFLE is enabled and ASSERTS are disabled
`start_pos` becomes an unused variable leading to a compile error.
Cleaned-up shuffling algorithm to not need a `start_pos` check.
Signed-off-by: Al Semjonovs <asemjonovs@google.com>
Adding guidelines for writing samples which includes definition
and minimum criteria to qualify as a sample.
Signed-off-by: Aastha Grover <aastha.grover@intel.com>
Fix the issue that the initialization will fail when both
CONFIG_SPI_NOR_SFDP_RUNTIME and CONFIG_SPI_NOR_IDLE_IN_DPD are enabled.
The cause of this problem is simply calling the wrong function.
Fixes#33015
Signed-off-by: Xinyang Tan <xinyang.tan@delve.com>
Allocated, but undersized packets must not just be logged, but also
unreferenced before returning an error.
Signed-off-by: Reto Schneider <reto.schneider@husqvarnagroup.com>
If `struct bt_mesh_prov::unprovisioned_beacon` callback is not set,
don't process the beacon.
Signed-off-by: Pavel Vasilyev <pavel.vasilyev@nordicsemi.no>
