Now that we a proper API, shell is just optional, so, make the
dependency optional by refactoring the code.
Also, add a build test combination in twister.
Signed-off-by: Krishna T <krishna.t@nordicsemi.no>
Initializing an event callback with net_mgmt_init_event_callback() just
sets some of the callback fields but do not propagate those masks in the
core. If we want to use the callback, it is necessary to also call
net_mgmt_add_event_callback().
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
The current association logic starts by changing the channel. The way it
is done is wrong because it dereferences req->channel which is simply
not initialized by the caller. But anyway, the command itself does not
support providing a channel so we must already be on the right one when
trying to associate. Hence, drop this channel change call.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
During an association, the peer we are trying to connect to will send us
an association response frame with the destination PAN ID set to the PAN
ID we try to join. If we do not update the hardware address filters
beforehands, it is likely that the hardware will just discard the
response and the association will fail.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
While the packet structures seem to always be reset when they are
allocated, it's apparently not the case of the data buffers. Indeed,
these are allocated differently and just attached to the packet
structure through a frag/buffer member.
Experience shows that we may get uninitialized buffers so let's set
all MAC parameter bits one by one, even the reserved ones.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
The spec clearly states: "association requests shall set the AR bit".
Even though Zephyr can currently only implement RFD devices which are
not expected to support incoming association requests, because this MAC
command is actually processed until being voluntarily ignored, let's
ensure the expected "ar" value is right to avoid failing because of a
wrong reason.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
When validating a MAC command, the "src" and "dst" fields may be set to:
- IEEE802154_ADDR_MODE_SHORT (0x2)
- IEEE802154_ADDR_MODE_EXTENDED (0x3)
- IEEE802154_ADDR_MODE_SHORT | IEEE802154_ADDR_MODE_EXTENDED (0x3)
Hence when the mode check happens, any times the mode is set to SHORT
the check will fail while in practice it was meant to be valid because
the check is:
if (src_mode == src || dst_mod == dst)
Use bitfields when relevant so that when checking capabilities we use
the bit offsets rather than the plain numbers.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
When validating a MAC command, there is sometimes a misunderstanding of
what "src" and "dst" length mean.
There are actually two fields in the MHR:
- One giving the type of address, if it is short or extended, it is the
value provided by the macros IEEE802154_ADDR_MODE_{SHORT,EXTENDED} and
their respective decimal values are 2 and 3.
- One giving the size of the address field, this is
IEEE802154_{SHORT,EXT}_ADDR_LENGTH and their value is actually 2 and 8
(bytes).
The function validate_mac_command() provides inputs to
validate_mac_command_cfi_to_mhr() which expects the former information.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Fix LwM2M rd client stop call hang when Queue client is at
RX_ON_IDLE_STATE. Added miossing connection resume for
de-register functionality.
Signed-off-by: Juha Heiskanen <juha.heiskanen@nordicsemi.no>
Check receiving buffer size and return error if
* buffer size is too small for opaque or string data type
* buffer size is not equal to data lenght when data type is
fixed size
Signed-off-by: Juha Ylinen <juha.ylinen@nordicsemi.no>
Instead of calling zperf shell initialization routine on the first
command execution, initialize it during system boot, along with other
zperf submodules.
Remove redundant IP address configuration on an interface. The default
configuration relies on NET_CONFIG module, so there's no need to set the
address manually in zperf.
Signed-off-by: Robert Lubos <robert.lubos@nordicsemi.no>
Zperf shell functionality is now encapsuled within a single file,
therefore it no longer makes sense to have a separate shell_utils
file.
Signed-off-by: Robert Lubos <robert.lubos@nordicsemi.no>
Make the TCP/UDP server functionality restartable. Provide a public API
to stop the TCP/UDP server.
Signed-off-by: Robert Lubos <robert.lubos@nordicsemi.no>
Add public API for zperf download functionality. The TCP/UDP server
modules are decoupled from shell, allowing to trigger download directly
from the application code. The shell submodule makes use of this new
public API.
Signed-off-by: Robert Lubos <robert.lubos@nordicsemi.no>
Add shell option to the UDP/TCP upload command, which allows to execute
the upload asynchronously. This allows to unblock the shell for other
commands during the upload.
Signed-off-by: Robert Lubos <robert.lubos@nordicsemi.no>
This commit defines a public API for zperf upload functionality. The
UDP/TCP uploader modules are decoupled from shell, allowing to perform
uploads directly from the application code. The shell submodule makes
use of this new public API.
Signed-off-by: Robert Lubos <robert.lubos@nordicsemi.no>
Clean up occurrences of "#if IS_ENABLED(CONFIG_FOO)" an replace
with classical "#if defined(CONFIG_FOO)".
Signed-off-by: Erwan Gouriou <erwan.gouriou@linaro.org>
websocket_recv_msg() is reworked with using fsm. Now the function
return 0 when payload is empty, -ENOTCONN if socket close. Receiving
empty ping and sending empty pong were added in tests.
Fixes#52327
Signed-off-by: Grixa Yrev <grixayrev@yandex.ru>
IEEE802154 L2 may modify the LL address during interface operation (when
processing MAC command). So far the L2 workaround the LL address update
protection by clearing the NET_IF_UP flag temporarily, but due to recent
changes it no longer works. Update this workaround to verify
NET_IF_RUNNING flag instead.
Signed-off-by: Robert Lubos <robert.lubos@nordicsemi.no>
Bluetooth IPSP L2 sets the LL address only after establishing Bluetooth
connection. As this can take place after the interface was brough UP by
the application, the network interface API should not block such
attempt. Instead, verify the NET_IF_RUNNING flag.
For the same reason, move the assert verifying that LL address is set
from net_if_up() (which only puts the interface in the admin UP state)
into the function that transitions the interface into operational UP
state.
Signed-off-by: Robert Lubos <robert.lubos@nordicsemi.no>
After network interfaces state handling rework, it's no longer correct
to verify the connection status on `net_if_up()`, as this only changes
the administrative state of the interface. The interface won't be put in
operational UP state until it's connected.
This check prevented the interface from being brought up during system
boot.
Signed-off-by: Robert Lubos <robert.lubos@nordicsemi.no>
Rewrite the "net ping" command handling, to allow the command to be
aborted during execution. This includes:
* Using shell bypass mode to capture input whilst the ping is active.
* Using system workqueue to send individual ping requestes, instead of
sending ping requests directly from shell thread, in a blocking
manner. This is needed because in order to receive input in the
registered bypass handler, the shell thread must be unblocked to
process the input.
* The bypass mode is left after receiveing `CTRL-C` character (which
cancels the ping), after receiving all expected Ping replies or after
the timeout occurs.
Signed-off-by: Robert Lubos <robert.lubos@nordicsemi.no>
Since 6c30c9ac47 (samples: net: zperf:
Rewrite upload part to use sockets), zperf uses SO_RCVTIMEO in UDP mode,
hence must depend on/enable support for it.
Without SO_RCVTIMEO support, zperf fails like this:
> nb_packets=47 delay=188964 adjust=-13
> setsockopt error (109)
> setsockopt error (109)
> -
> Upload completed!
> LAST PACKET NOT RECEIVED!!!
> Statistics: server (client)
> Duration: 0 us (10.19 s)
> Num packets: 0 (50)
> Num packets out order: 0
> Num packets lost: 0
> Jitter: 0 us
> Rate: 0 Kbps (9 Kbps)
Signed-off-by: Reto Schneider <reto.schneider@husqvarnagroup.com>
Setting the protocol type for raw IP packets to be sent so that they
can be passed to Ethernet.
Signed-off-by: Jani Hirsimäki <jani.hirsimaki@nordicsemi.no>
The sequence id from iperf starts with 1. The commit changes the
initial value of "next_id" in zperf from 0 to 1. In addition, the
error output of "error" and "out of order" packets is corrected.
Signed-off-by: Yanqin Wei <Yanqin.Wei@arm.com>
This patch reworks how fragments are handled in the net_buf
infrastructure.
In particular, it removes the union around the node and frags members in
the main net_buf structure. This is done so that both can be used at the
same time, at a cost of 4 bytes per net_buf instance.
This implies that the layout of net_buf instances changes whenever being
inserted into a queue (fifo or lifo) or a linked list (slist).
Until now, this is what happened when enqueueing a net_buf with frags in
a queue or linked list:
1.1 Before enqueueing:
+--------+ +--------+ +--------+
|#1 node|\ |#2 node|\ |#3 node|\
| | \ | | \ | | \
| frags |------| frags |------| frags |------NULL
+--------+ +--------+ +--------+
net_buf #1 has 2 fragments, net_bufs #2 and #3. Both the node and frags
pointers (they are the same, since they are unioned) point to the next
fragment.
1.2 After enqueueing:
+--------+ +--------+ +--------+ +--------+ +--------+
|q/slist |------|#1 node|------|#2 node|------|#3 node|------|q/slist |
|node | | *flag | / | *flag | / | | / |node |
| | | frags |/ | frags |/ | frags |/ | |
+--------+ +--------+ +--------+ +--------+ +--------+
When enqueing a net_buf (in this case #1) that contains fragments, the
current net_buf implementation actually enqueues all the fragments (in
this case #2 and #3) as actual queue/slist items, since node and frags
are one and the same in memory. This makes the enqueuing operation
expensive and it makes it impossible to atomically dequeue. The `*flag`
notation here means that the `flags` member has been set to
`NET_BUF_FRAGS` in order to be able to reconstruct the frags pointers
when dequeuing.
After this patch, the layout changes considerably:
2.1 Before enqueueing:
+--------+ +--------+ +--------+
|#1 node|--NULL |#2 node|--NULL |#3 node|--NULL
| | | | | |
| frags |-------| frags |-------| frags |------NULL
+--------+ +--------+ +--------+
This is very similar to 1.1, except that now node and frags are
different pointers, so node is just set to NULL.
2.2 After enqueueing:
+--------+ +--------+ +--------+
|q/slist |-------|#1 node|-------|q/slist |
|node | | | |node |
| | | frags | | |
+--------+ +--------+ +--------+
| +--------+ +--------+
| |#2 node|--NULL |#3 node|--NULL
| | | | |
+------------| frags |-------| frags |------NULL
+--------+ +--------+
When enqueuing net_buf #1, now we only enqueue that very item, instead
of enqueing the frags as well, since now node and frags are separate
pointers. This simplifies the operation and makes it atomic.
Resolves#52718.
Signed-off-by: Carles Cufi <carles.cufi@nordicsemi.no>
Introduce the Kconfig symbol `NET_PKT_BUF_USER_DATA_SIZE`. It is used to
set the user data size in the buffers used in the rx and tx network buffer
pools.
Signed-off-by: Christian Taedcke <christian.taedcke@lemonbeat.com>
This Kconfig setting was deprecated since
commit 8e99db5801 ("Kconfig: net: deprecate `NET_BUF_USER_DATA_LEN`")
and zephyr release 3.0.0.
This commit removes this deprecated and unused setting.
Signed-off-by: Christian Taedcke <christian.taedcke@lemonbeat.com>
When testing on qemu_x86_64 with e1000 Ethernet driver, there are
several crashes due to list management simultaneously executing on
different cores. Add mutexes similar to other parts on networking
stack, for example tcp_lock.
Signed-off-by: Andrei Emeltchenko <andrei.emeltchenko@intel.com>
As per RFC 2236 Section 9, the IGMP V2 membership reports should be sent
to the group multicast address being reported. Without this fix, when
IGMP snooping is enabled on a network a device is not able to receive
multicast from certain multicast addresses e.g. 239.255.x.x.
Fixes#52449
Signed-off-by: Chamira Perera <chamira.perera@audinate.com>
Allocation callback in net_buf_heap_cb and net_buf_var_cb
used for net_bufs with variable size payloads, defined by
NET_BUF_POOL_HEAP_DEFINE or NET_BUF_POOL_VAR_DEFINE,
allocate one more octet for internal variable ref_count,
used in function generic_data_ref(), which in turn is needed
for net_buf_clone()).
The user gets a buffer which is shifted by one octet in memory
block. This breaks alignment provided k_heap_alloc and k_malloc.
Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>
Add CONFIG_CRC for building CRC related routines.
CRC routines are now being built for each application, whether used or
not and are add in the build system unconditionally.
Keep CONFIG_CRC enabled by default for now and until all users have
converted to use the new option.
Partial fix for #50654
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
This aligns the conditional compilation logic with the one used for the
prototype in net_pkt.h, avoids build errors due to a undefined reference
to net_pkt_print_frags.
Signed-off-by: Reto Schneider <reto.schneider@husqvarnagroup.com>
net_pkt_get_frag() and a few other functions did not specify the
allocated fragment length, incorrectly assuming that fixed-sized
buffers are always used.
In order to make the function work properly also with variable-sized
buffers, extend the function argument list with minimum expected
fragment length parameter. This allows to use net_buf_alloc_len()
allocator in variable buffer length configuration, as well as verify if
the fixed-sized buffer is large enough to satisfy the requirements
otherwise.
Update the existing codebase to provide the expected fragment length,
based on the context.
Signed-off-by: Robert Lubos <robert.lubos@nordicsemi.no>
The "net mem" command handler did not take the variable buffer length
configuration, and failed to build in such case.
Signed-off-by: Robert Lubos <robert.lubos@nordicsemi.no>
In case variable buffer length feature was enabled, net_pkt code did
not build due to unconditional references to CONFIG_NET_BUF_DATA_SIZE,
which is not avaialable in this configuration.
Signed-off-by: Robert Lubos <robert.lubos@nordicsemi.no>
In case variable buffer length feature was enabled, the TCP stack did
not build due to unconditional references to CONFIG_NET_BUF_DATA_SIZE,
which is not avaialable in this configuration.
Signed-off-by: Robert Lubos <robert.lubos@nordicsemi.no>
In case Echo Request did not carry timestamp (i. e. requested payload
size was smaller than 4), there won't be a timestamp in the reply
either. Such replies should not be ignored. Instead, simply skip the
turnround time calculation if timestamp is missing.
Signed-off-by: Robert Lubos <robert.lubos@nordicsemi.no>
Add -s parameter to the net ping command, allowing to specify the
payload size for the Echo Request.
Signed-off-by: Robert Lubos <robert.lubos@nordicsemi.no>
Let net_icmpv4_send_echo_request() and net_icmpv6_send_echo_request()
autogenerate Echo Request payload, in case data length is specified but
no data pointer provided. The autogenerated payload includes timestamp
(if payload size permits) so that the turnround time can be calculated.
Signed-off-by: Robert Lubos <robert.lubos@nordicsemi.no>
This reverts changes introduced in commit
dd535f611d, as they broke the gsm_ppp
driver integration with PPP L2. Apparently, a more thorough
refactoring is needed to use the new interface management scheme with
PPP.
Signed-off-by: Robert Lubos <robert.lubos@nordicsemi.no>
