The shim files should be part of the Zephyr module directory within
Zephyr, as they are coupled with the module implementation which os OS
agnostic.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
Following the naming convention add an underscore and update path in the
maintainers file too.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
`_current` is now functionally equals to `arch_curr_thread()`, remove
its usage in-tree and deprecate it instead of removing it outright,
as it has been with us since forever.
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
Signed-off-by: Yong Cong Sin <yongcong.sin@gmail.com>
If a character is received immediately after modem_iface_uart_init called,
the modem_iface_uart_isr function will not read the data in the serial port
fifo register as the context has not been registered at this time,
which will result in the program not being able to exit the interrupt,
so the context should registered before the serial port is initialised.
Signed-off-by: Hongquan Li <hongquan.prog@gmail.com>
As long as MBEDTLS_ENTROPY_C is enabled, Mbed TLS needs to
poll some entropy source to gather data that will then be
processed by CTR/HMAC-DRBG modules. This means that in most
of the cases, once MBEDTLS_ENTROPY_C is enabled then also
MBEDTLS_ENTROPY_POLL_ZEPHYR needs to be enabled. This was
done manually until now, as the long list of samples/tests
demonstrate.
This commit solves this dependency by defaulting
MBEDTLS_ENTROPY_POLL_ZEPHYR to on as soon as
MBEDTLS_ENTROPY_C is set. As a consequence, all manual
enablement of MBEDTLS_ENTROPY_POLL_ZEPHYR in samples/tests
are removed.
Signed-off-by: Valerio Setti <vsetti@baylibre.com>
Remove CONFIG_GPIO from defconfigs for Infineon boards.
Applications, drivers will enable GPIO if need.
Added 'select GPIO' from spi/Kconfig.ifx_cat1
Added 'select GPIO' from wifi/infineon/Kconfig.airoc
Signed-off-by: Nazar Palamar <nazar.palamar@infineon.com>
Some spi drivers do not allow the send buffer
and receive buffer to be empty at the same time,
if this happens it will cause the spi to be unable
to communicate with the nrf7002, so add the receive
buffer for the discard byte in the spim_xfer_rx.
Fix#80686
Signed-off-by: Hongquan Li <hongquan.prog@gmail.com>
nRF70 recovery relies on power-management feature which isn't applicable
for SAP, so, we keep seeing false recovery causing interface restart.
Disabled recovery for SAP to fix the issue.
Fixes#81130.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
With the newly introduced settling time for PLLS, we are seeing a peak
performance drop of 3Mbps in Zperf benchmarks, and also other
performance tests are also impacted.
This settling time was introduced based on observations in lab and not
because of any real problem or bug reported, so, for now, use zero
settling time (same as earlier) till we fully investigate and understand
the impacts and need.
Fixes#80951.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
This works because undefined macro in conditional is treated as zero,
but could end up choosing the wrong divider.
Fix the macro with the new name.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
With introduction of Raw modes, nRF70 driver now advertises get_c
onfig OP, but doesn't implement all types.
This causes problems two-fold with checksum calculations:
1. The "config" isn't uninitialized, so, every call returns differnet
values. So, for UDP header checksum would be done and
pkt->chksumdone would be set. But for IPv4 header checksum might be
skipped.
2. Even if we initialize to zero, then network stack gets all zeros
and calculates checksum by itself rendering offload moot.
There is another problem in #1, as there is only single flag for pkt for
all checksum, nRF70 driver sees this and tells UMAC to skip checksum for
the entire packet. The design isn't coherent, and should be converted to
communicate per-type checksum status (some are filled by network stack
and some HW).
But as nRF70 support all checksum offloads, advertise all types for both
RX and TX.
Upstream PR #: 80882
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
In case the driver UP fails, the FMAC context will be NULL, so, add a
NULL check in the DOWN.
Fixes a crash seen when working with unprogrammed OTP (no MAC) that
fails the interface UP.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
During secure association, if we get any data packets from the
networking stack (as we switch off the dormancy) then they will be
dropped till the 802.1x port is authorized, but with recent changes the
corresponding network buffer isn't freed causing a memory leak.
Handle this case and free the network buffer (Using an OSAL API though
it's Zephyr to avoid duplication and also getting access to "struct
nwb");
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
Make sure all kconfig related to Wi-Fi is
in its driver area.
This commit also removes esp_timer_init() call
from Wi-Fi driver.
Signed-off-by: Sylvio Alves <sylvio.alves@espressif.com>
Fix a set of warnings like:
drivers/wifi/winc1500/wifi_winc1500.c:332:25: error: cast from pointer
to integer of different size [-Werror=pointer-to-int-cast]
332 | SOCKET socket = (int)context->offload_context;
when int and void* do not have the same size, by casting the content
of context->offload_context to intptr_t which is whichever integer
size matches the pointer size.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
Recent WPA supplicant changes have broken nRF regulatory support,
implement the new set/get country WPA supplicant ops.
WPA supplicant: Regulatory SET through WPA supplicant, GET is direct to
the driver
Scan only: SET and GET direct calls to the driver
Fixes#79916.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
The eswifi driver uses an integer to identify the socket used for each
connection, which must be cast to and from a `void *`. This causes
warnings on 64 bit platforms, as precision is lost when casting from a
`void *` to `int`. Use a `intptr_t` type to store the socket value to
resolve this warning. Also, fix a function signature for the `accept_cb`
to use `size_t` instead of `int` for the length
Fixes#80242
Signed-off-by: Daniel DeGrasse <daniel.degrasse@nxp.com>
The defines should explicitly be passed to nordic HAL as they dont' use
the CONFIG_ prefix.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
Add CONFIG_NXP_WIFI_BUILD_ONLY_MODE, which allows the NXP WiFi driver to
be built without a binary blob. The built application will not be
functional, but this enables continuous integration to run upstream.
Signed-off-by: Daniel DeGrasse <daniel.degrasse@nxp.com>
set_btwt API implementation is not complete. Remove the structure
definition and implementation from the NXP WiFi driver to fix a build
error.
Signed-off-by: Daniel DeGrasse <daniel.degrasse@nxp.com>
ap_bandwidth API implementation is not complete. Remove the structure
definition and implementation from the NXP WiFi driver to fix a build
error.
Signed-off-by: Daniel DeGrasse <daniel.degrasse@nxp.com>
Though nRF7000 is a special Scan only chipset, scan only can work with
any nRF70 platforms, no need for this enforcement and we can keep things
flexible.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
Fix couple of bugs in exit path:
- In case of calloc failure, return's without unlocking
- memory is freed outside lock, in case of a tiny window of race, this
can cause a crash when this function is called from two threads.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
During Wi-Fi connection UMAC sends an unsolicited regulatory change
event but the driver code always assumes that this event is solicited
hence doesn't free the memory for the event.
Fix this by dropping the unsolicited event as it's not supported yet.
Fixes#79733.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
This adds promiscuous mode support functions into build for filter
operation when the configuration CONFIG_NRF70_PROMISC_DATA_RX
is enabled.
Signed-off-by: Vivekananda Uppunda <vivekananda.uppunda@nordicsemi.no>
This set of changes brings in promiscuous mode filtering support in
driver. Since, firmware would be unable to filter packets due to
connection issues, the filtering support for promiscuous mode is
moved to the driver.
Signed-off-by: Vivekananda Uppunda <vivekananda.uppunda@nordicsemi.no>
The mutex is used to protect RPU zephyr context which gets modified for
every interface down and up (including recovery), so, it was being
re-initialized but also used to protect down and up which is a bug.
Move the re-initialization to the driver entry so that it happens only
once and we can properly use the mutext for down and up protection.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
The RPU context lock is not unlocked this is causing recovery to be
stuck waiting for the lock.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
There is a race condition when recovery is in progress and in parallel
Wi-Fi util commands are being executed (CTF), where the RPU context is
de-initialized as part of recovery but no checks are present in the
Wi-Fi util command processing causing a crash.
This needs a proper fix for all commands, but for maintenance branch
fix is added only for commonly used commands.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
Dynamically set power save exit strategy runtime configuration that
allows to switch b/w stratgies depending on conserving power and
low-latency traffic download.
Signed-off-by: Ajay Parida <ajay.parida@nordicsemi.no>
Ideally we should be using Zephyr spinlock APIs but that requires
changes to shim API, so, for this maintenance release just replace with
mutex to keep the context same and no API changes.
This solves the locking issue that we see when control and data path are
excited concurrently due to locking semantics of semaphores, the issue
is not root caused but mutex enforce strict locking semantics for
multiple threads and solve the issue.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
Add a null check for HAL context in the interrupt handler, this was
causing locking issue operating on null. The root cause of null is not
known, but this solves the locking issue when data and control paths are
excited in parallel.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
To handle interoperability issue with few APs, add a feature to keep
sending keepalive frames periodically to avoid AP disconnecting the STA.
This is disabled by default to avoid unnecessary power consumption as
it's only seen with few old APs.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
In crowded environments RPU is active for more than 10s due to too many
retries and this triggers a false RPU recovery. To avoid this, increase
the default to 50s to handle corner cases, as this will only impact the
recovery triggered case, higher timeout doesn't have any impact in
normal cases.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
With this offload, host doesn't need to manage RX buffers for management
frames, and this saves Host-RPU comms and thus giving RPU to sleep more
often and is essential to test RPU recovery.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
These are very frequent, so, a separate debug is added for debugging
host RPU recovery logic.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
During interface down in case TX has pending buffers in either TXQ or
Pending_Q then they are not freed instead the Q itself is freed.
Fix by traversing the Q and freeing all members.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
Fix RPU recovery protection to solve build failures when RPU recovery is
disabled.
As recovery is primarily based on power-management, add a Kconfig
dependency to enforce, this simplies the macros to protect the code.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
In case RPU is stuck in consecutive recovery over a time period then
that means it's not recoverable through RPU recovery, only thing left to
do is to trigger a system reboot. This feature is disabled by default,
so, either application can do their own implementatio or enable this
feature in the driver along with configurable retries and window period.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
Check for RPU context as well.
To fix this properly we need more fixes to be backported, but this
should suffice for now.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
This is to avoid successive recoveries in case we get successive
watchdog interrupts from the RPU.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
During recovery we might get further watchdog interrupts causing
multiple recovery requests, ignore them if a recovery is already in
progress.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
In order for the interface down to propagate and cleanup it needs more
time, using Shell 10ms was working due to human delay, but
programatically this needs higher delay.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
Before proceeding with RPU bringup, do a sanity check by reading a known
signature to make sure the Host-RPU comms are operational.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
In case RPU is stuck and need a recovery, the failures in interface down
should be ignored as they are expected and we should proceed with device
removal that in turn removes power to the RPU.
TODO: This works for single VIF, but needs more thought for multi-VIF.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
During watchdog (or any) interrupt processing, RPU accesses are being
made and they assert the wakeup_now flag this causes RPU recovery to not
trigger.
New false or true recovery detection algo:
Check the time difference b/w last de-assert and assert, and if it
exceeds minimum time needed for RPU to enter sleep, then not the
timestamp. This timestamp will be used to compare when a watchdog
interrupt is received and see if during the last window if host has
given a chance for RPU to attempt sleep, if yes, then attempt recovery
else ignore watchdog.
Also, add a Kconfig for the 10s active time that triggers recovery, this
needs to be passed to the FW (once we have enough patch memory).
Also, add a Kconfig for the minimum time needed for RPU to attempt sleep
in positive case.
Also, add a new _ms API for time stamp fetch, this is to avoid
precision loss when converting to and from ms to us and also makes code
readable by avoiding *1000 and /1000.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
This commit applies several changes in the way "heap_runtime"
feature is used. It can't be split due to bisectability issues.
Whenever the feature is enabled, a new heap is created and
custom malloc/calloc/free functions are added into the build
system. Those functions are currently used for internal Wi-Fi and BLE
drivers only.
Such changes are described below:
1) Rename heap.c to esp_heap_runtime.c for better readability.
2) Rename RUNTIME_HEAP to HEAP_RUNTIME to make it similar to what is
available in Zephyr.
3) Add runtime heap to BT as such as Wi-Fi.
Fixes#79490Fixes#79470
Signed-off-by: Sylvio Alves <sylvio.alves@espressif.com>
Update both Wi-FI and BLE init codes to return proper
error code and logging when it is missing heap.
Signed-off-by: Sylvio Alves <sylvio.alves@espressif.com>
`eswifi-uart` uses the ring buffer library, but the dependency is
not indicated.
Specify `select RING_BUFFER` to resolve the dependency.
Signed-off-by: TOKITA Hiroshi <tokita.hiroshi@gmail.com>
`_POSIX_C_SOURCE` must be specified to use the `strnlen()` function
in string.h. Make sure to specify this explicitly.
Signed-off-by: TOKITA Hiroshi <tokita.hiroshi@gmail.com>
The FW blobs are contentious topic in OSS, so, add a provision to use
external tooling (e.g., NCS) to handle FW blobs for advanced usecases
(e.g., storing in external flash).
This won't be used upstream, only built-in or build-only are supported.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
Instead of relying on INCBIN macros which do not properly add
dependencies, e.g., modifying FW file doesn't trigger rebuild.
Use the Zephyr cmake tooling to load the FW patch file as a header.
This also improves memory report where the patch target is clearly
visible instead of a hidden section.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
This reverts commit e2e96acebf.
This will be properly fixes by adding a target for nRF70.bin and
removing INCBIN approach.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
Random MAC address setting can never be configured as the two defaults
cover all cases. Fix the defaults, now the order is
* Fixed
* OTP (default, in case of no config)
* Random
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
Whenever FW blobs are updated manually, we need to tell cmake to
auto-build the source files to use the latest firmware without doing a
pristine build.
This adds a custom target to be run with nRF Wi-Fi driver and updates
timestamp of fw_load.c to rebuild whenevr the blob is updated.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
This script is needed only when external flash is used to store nRF70
firmware patches. This also uses PM which is supported only in NCS.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
For using external flash the tooling is only available in NCS, so, use
the existing Kconfig option to override the FW loading.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
Checks blobs presence as dependency to allow driver build.
Goal is to avoid building in environment where blobs won't
be present (e.g GitHub CI environment).
Signed-off-by: Raffael Rostagno <raffael.rostagno@espressif.com>
Currently a user, or a shield definition, must select a Kconfig to
signal that the iovdd_ctrl and bucken GPIOs are connected to the same
pin. We can instead check this in the driver by simply checking if
the iovdd_ctrl_spec and bucken_spec are the same pin.
Update the nrfwifi driver and remove the redundant
NRF_WIFI_COMBINED_BUCKEN_IOVDD_GPIO config.
Signed-off-by: Bjarki Arge Andreasen <bjarki.andreasen@nordicsemi.no>
For code clarity, this commit adjusts the use of `return` statements
in functions with a void return type as follows:
- Transform `return foo();` into separate statements:
`foo();`
`return;`
- Remove unnecessary `return` statements when
they don't affect control flow.
Signed-off-by: Pisit Sawangvonganan <pisit@ndrsolution.com>
There is a regression caused by #76177, which
causes build to fail due to missing includes and others.
This wraps it with proper checks and fixes wifi scan call.
This also remove unused variable present in the same driver.
Signed-off-by: Sylvio Alves <sylvio.alves@espressif.com>
Add missing braces to comply with MISRA C:2012 Rule 15.6 and
also following Zephyr's style guideline.
Signed-off-by: Pisit Sawangvonganan <pisit@ndrsolution.com>
Remove all entries that as not being used.
This also update hal to re-enable warning flags
as such as -Wno-unused-variable.
Signed-off-by: Sylvio Alves <sylvio.alves@espressif.com>
Removes constant error logging when parsing IPD headers in active mode and
waiting on remote IP and port. Currently an error message is logged for
every character in the remote IP and port until the full length obtained.
Signed-off-by: Daniel Mangum <georgedanielmangum@gmail.com>
The ESP32 Wi-Fi driver previously did not support APSTA mode,
enabling AP mode would disable STA mode and vice versa.
Support for APSTA has been added by introducing additional
network interface: one for STA (Interface 2) and one for
AP (Interface 1). The CONFIG_ESP32_WIFI_AP_STA_MODE option in
Kconfig now allows enabling or disabling this support.
Signed-off-by: Muhammad Haziq <muhammad.haziq@zintechnologies.com>
When build with the psa_crypto_driver, it will try to include common.h
of mbedtls, but actually include the common.h from hostap, and causes
build error of undefined symbol. Now link the library of hostap to use
the header file in /utils path, to fix the build error.
Signed-off-by: Maochen Wang <maochen.wang@nxp.com>
Updated condition to enable NET_STATISTICS. Changed
CONFIG_NET_STATISTICS_ETHERNET to CONFIG_NET_STATISTICS_WIFI
in airoc_wifi.c
Signed-off-by: Nazar Palamar <nazar.palamar@infineon.com>
Add a mode to build the driver without FW blobs, this is mainly for
twister to run and catch any build issues without depending on any FW
blobs.
Signed-off-by: Chaitanya Tata <Chaitanya.Tata@nordicsemi.no>
1) After the wifi driver successfully connects to the AP,
the dormant state should be set to OFF.
Once a deauth/disassoc frame is received or link lost occurs,
the dormant state should be set to ON.
2) Enable CONFIG_NET_TC_SKIP_FOR_HIGH_PRIO and CONFIG_NET_CONTEXT_PRIORITY
to push high priority packets directly to network driver.
Signed-off-by: Gang Li <gang.li_1@nxp.com>
