If we could not fragment the IPv4 packet, then drop it and do
not try to send it. Let the upper layer re-send the packet if
needed. It is causing more trouble if we try to send the packet
and not honor the MTU setting.
Fixes#81021
Signed-off-by: Jukka Rissanen <jukka.rissanen@nordicsemi.no>
SNTP simple runs request iterations with exponential backoff.
If the net interface is a slower connection (ie. CAT M1 modems)
then the request will be sent but the response may take time to
be received, thus causing a timeout and another request to be sent.
Because of the nature of UDP and the fact that the same socket
(source IP/port combo) is being used for both requests, a delayed
response to the first request can be received as the response to the
second request, causing -EINVAL to be returned when the timestamps
mismatch (see subsys/net/lib/sntp/sntp.c). The solution provided
retries receiving the response when the timestamp is mismatched
(without sending an additional request).
Signed-off-by: Marcus Penate <marcus.penate@ellenbytech.com>
Add condition check so that the 11k flag will be updated for set
operation only.
Fix print log error when getting 11k status.
Signed-off-by: Hui Bai <hui.bai@nxp.com>
Using strncpy gives this warning
warning: 'strncpy' output may be truncated copying 32 bytes from
a string of length 32 [-Wstringop-truncation]
strncpy(roaming_params.neighbor_rep.neighbor_ap[idx].bssid_info,
bssid_info,
sizeof(roaming_params.neighbor_rep.neighbor_ap->bssid_info));
There is '\0' at the end of the allocated buffer so we can safely
use memcpy() here to avoid any warnings.
Signed-off-by: Jukka Rissanen <jukka.rissanen@nordicsemi.no>
Make sure we are not able to overwrite string variables in sscanf
call. Allocate also one extra byte for null terminator character.
Fixes#80644
Signed-off-by: Jukka Rissanen <jukka.rissanen@nordicsemi.no>
Because we might get answers in capital letters, convert the answer
to small case letters and also make sure we send query in small case
latters. This makes sure that our query_hash is properly calculated
regardless of how the resolver gets the data.
Signed-off-by: Jukka Rissanen <jukka.rissanen@nordicsemi.no>
Allocate one extra pointer for the DNS server list so that
DNS resolving code can detect the end of the list.
Signed-off-by: Jukka Rissanen <jukka.rissanen@nordicsemi.no>
In receiving thread, continuing the loops is based on
has_ongoing_exchanges() so it does not need atomic
coap_client_recv_active variable.
When idling, it wakes from semaphore. But there was potential
deadlock when coap_client_schedule_poll() would not signal the
semaphore, if atomic variable was already showing that it runs.
Removing the atomic variable removes this deadlock.
Signed-off-by: Seppo Takalo <seppo.takalo@nordicsemi.no>
It is error prone to flag separate booleans, so try to use
reset_internal_request() every time we release the internal request
structure.
Also refactor the reset_internal_request() so that we reset the
timeout value so it does not trigger again.
Signed-off-by: Seppo Takalo <seppo.takalo@nordicsemi.no>
If send() fails, we have not technically send the CoAP retry yet, so
restore the same pending structure, so our timeouts and retry counters
stay the same.
This will trigger a retry next time the poll() return POLLOUT, so we
know that we can send.
Signed-off-by: Seppo Takalo <seppo.takalo@nordicsemi.no>
Refactor the CoAP retry handling into the handle_poll() function,
so that we only try to send retries if the socket reports POLLOUT.
Also move the receiving into same loop, so when poll() reports POLLIN
we recv() the message and handle it before proceeding to other sockets.
Also fix tests to handle POLLOUT flag and add support for testing
multiple clients.
Signed-off-by: Seppo Takalo <seppo.takalo@nordicsemi.no>
When transmission of first request fails, reset the internal request
buffer as there is no ongoing CoAP transaction.
Application can deal with the failure.
Signed-off-by: Seppo Takalo <seppo.takalo@nordicsemi.no>
If POSIX_API is not configured the close function is not available.
Use zsock_close instead.
Signed-off-by: Andreas Huber <andreas.huber@ch.sauter-bc.com>
Add TLS_ECDHE_ECDSA_WITH_AES_128_CCM_8 to the list for use with x509
certificates. The LWM2M v1.1 specification says that a LWM2M client
which used X509 certificates must support this ciphersuite and
additional ciphersuites may be supported.
Signed-off-by: Jeroen Broersen <jbroersen@interact.nl>
The second condition needs to check ARP probes only
The ACD is not properly implemented as described in RFC5227 ch. 2.1.1
The implementation incorrectly detects an IP conflict, if an ARP request
is received for the target IP.
The reason is that the current implementation checks for ARP requests
instead of ARP probes.
Signed-off-by: Andreas Huber <andreas.huber@ch.sauter-bc.com>
Before this patch, any unexpected socket error during poll (caused by
LTE disconnects for instance) would lead to a infinite loop because the
error state was never cleared, handled or even signaled to the user.
Signed-off-by: Benjamin Lindqvist <benjamin@eub.se>
As the socket service API is currently used by mutiple applications
(dhcpv4 server, dns, telnet), it should be marked as unstable,
according to the docs:
https://docs.zephyrproject.org/latest/develop/api/api_lifecycle.html
Signed-off-by: Fin Maaß <f.maass@vogl-electronic.com>
Originally, the net_mgmt task priority is very low. Based on roaming
implementation, roaming is triggered in net_mgmt task. When running UDP
test while doing roaming, the net_mgmt task won't have much chance to
run and roaming can't be triggered.
Increase it to 3, which is same value of supplicant task.
Signed-off-by: Hui Bai <hui.bai@nxp.com>
Added new ops and events in glue layer to support roaming.
Added new flag WIFI_NM_WPA_SUPPLICANT_ROAMING to control roaming
feature.
Signed-off-by: Hui Bai <hui.bai@nxp.com>
Generated PEM headers need to be created before we build the WiFi shell
when WIFI_NM_WPA_SUPPLICANT_CRYPTO_ENTERPRISE is enabled, so add an
explicit dependency on the app target to be certain these headers are
created
Signed-off-by: Daniel DeGrasse <daniel.degrasse@nxp.com>
The "wifi ap config" command uses the station interface via
net_if_get_first_wifi(), and should use the ap interface.
Signed-off-by: Gang Li <gang.li_1@nxp.com>
remove k_work related code and change
the argument of the callback to `struct net_socket_service_event`.
Signed-off-by: Fin Maaß <f.maass@vogl-electronic.com>
Switch from using direct `strtol` calls to `shell_strtol`.
This change leverages the extensive error handling provided
by `shell_strtol`.
Signed-off-by: Pisit Sawangvonganan <pisit@ndrsolution.com>
Since `<host>` is a mandatory argument, the (_mand, _opt) values
should be adjusted to 2 and 12, respectively.
Note that `_mand` includes the number of mandatory arguments,
including the command name (`ping` itself).
Signed-off-by: Pisit Sawangvonganan <pisit@ndrsolution.com>
If CONFIG_NET_PKT_ALLOC_STATS is enabled, then "net mem" command
can show net_pkt allocation statistics like succeed / failed
allocation counts, average sizes and allocation time.
Signed-off-by: Jukka Rissanen <jukka.rissanen@nordicsemi.no>
This patch adds a feature to directly connect to stored Wi-Fi
credentials without having to compose the NET_MGMT commands yourself.
Signed-off-by: Maximilian Deubel <maximilian.deubel@nordicsemi.no>
Upstream NCS's library for storing Wi-Fi credentials.
This library allows storage of Wi-Fi credentials
using different backends.
Either the Zephyr settings subsystem
or the PSA secure backend can be used.
For testing purposes, credentials can be defined statically.
Signed-off-by: Ravi Dondaputi <ravi.dondaputi@nordicsemi.no>
Signed-off-by: Kapil Bhatt <kapil.bhatt@nordicsemi.no>
Signed-off-by: Gregers Gram Rygg <gregers.gram.rygg@nordicsemi.no>
Signed-off-by: Kaja Koren <kaja.koren@nordicsemi.no>
Signed-off-by: Simen S. Røstad <simen.rostad@nordicsemi.no>
Signed-off-by: Maximilian Deubel <maximilian.deubel@nordicsemi.no>
Currently GCC complains that temp64 may be used uninitialized in this
function. Adds a check to ensure time is valid before assignining
and fixes GCC warning.
Signed-off-by: Brandon Allen <brandon.allen@exacttechnology.com>
The library provides Prometheus metrics
types, collector and exposion formatter.
The library isn't thread-safe for now.
The next first pull request will support
that. Can be use exposion formatted
output with Zephyr Http server.
Signed-off-by: Mustafa Abdullah Kus <mustafa.kus@sparsetechnology.com>
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>
