Previously, a call to set the net_context state was added here to
accommodate a new KConfig scenario that permitted disabling offload
sockets. This code path is used by both local and offloaded sockets.
While it is necessary to update the state of the net_context for a
locally managed socket, setting the net_context for an offloaded socket
is an error, as the net_context is invalid (in fact, it points to a
hard-coded dummy socket), as observed in #52346 and #38544.
A prior commit (a9ac0a88) attempting to remedy the offload socket
scenario removed this line, which resulted in problems for the local
socket scenario. The socket never got updated to the connected state,
and thus was unusable.
Adding this guard allows the eswifi socket to set the socket net_context
only if the socket is NOT offloaded.
Signed-off-by: Brian Dunlay <brian@nubix.io>
`spi_is_ready` function is being deprecated in favor of
`spi_is_ready_dt` so let's replace the old usage in the tree.
Signed-off-by: Bartosz Bilas <bartosz.bilas@hotmail.com>
Do not set wifi optimizations by default in order
to free up IRAM room region. In case higher throutput
is necessary, one can enable this flags to optmize it.
Signed-off-by: Sylvio Alves <sylvio.alves@espressif.com>
This change fixes a regression from commit 1cbc0acd.
Socket offload uses a dummy socket context, and setting
the socket state in this dummy context is invalid.
Fixes#52317
Signed-off-by: Brian Dunlay <brian@nubix.io>
In order to allow proper softAP mode, wifi
libraries and wpa supplicant requires this kconfig
to be enable.
Signed-off-by: Sylvio Alves <sylvio.alves@espressif.com>
Align Ethernet/Wi-Fi drivers/L2 with interface state handling update.
For drivers, that did not support carrier detection, no changes are
needed.
Driver that did support carrier detection, are updated to set the
carrier state to OFF by default, instead of setting the
NET_IF_NO_AUTO_START flag. This allows to postopne the actual
NET_EVENT_IF_UP notification until driver detects that carrier is ready.
Signed-off-by: Robert Lubos <robert.lubos@nordicsemi.no>
This fixes a random crash caused by race condition in the eswifi
driver used by the disco L475 iot1 board.
Signed-off-by: Michał Zając <m.zajac@avsystem.com>
Instead of using "select" on certain EC configurations, which is
considered unsafe for various reasons, use a "depends on" and rely on
the user to set a proper configuration in the config file.
Update the respective project configurations to comply with the new
configuration scheme.
Signed-off-by: Robert Lubos <robert.lubos@nordicsemi.no>
Rename Z_DEVICE_DT_DEV_NAME to Z_DEVICE_DT_DEV_ID to make it more clear,
as the macro is generating a token that identifies the created struct
device.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Those Kconfig options are no longer used after ESP32 WiFi driver was
converted to Zephyr WiFi mgmt and the removal of platform specific
`samples/boards/esp32/wifi`.
Fixes: d524015f82 ("samples: boards: remove esp32 wifi sample code")
Signed-off-by: Marcin Niestroj <m.niestroj@emb.dev>
Driver is not ready to handle NET_REQUEST_WIFI_CONNECT wifi_mgmt events
before ESP32_WIFI_EVENT_STA_START event is received. Revert back to setting
NET_IF_NO_AUTO_START flag for interface and call net_eth_carrier_on() after
ESP32_WIFI_EVENT_STA_START event is received. This makes it possible for an
application to wait for network interface to be ready, e.g. with:
while (!net_if_is_up(iface)) {
k_sleep(K_MSEC(100));
}
before making a connect request with:
net_mgmt(NET_REQUEST_WIFI_CONNECT, iface ...).
Fixes: 7d9edc8bf0 ("wifi: esp32: add support to wifi api mgmt")
Signed-off-by: Marcin Niestroj <m.niestroj@emb.dev>
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>
Due to Zephyr's WiFi updates, common private
enums used in `hal_espressif` are now being used.
This updates private internal values to meet hal_espressif
changes.
Signed-off-by: Sylvio Alves <sylvio.alves@espressif.com>
ESP32 wifi connection uses Zephyr's net stack. Once WiFi connects
in station mode, is requires the application to handle DHCPv4
negotiation. This PR adds support to automatic negotiation by
handling the dhcpv4 calls in driver layer.
Signed-off-by: Sylvio Alves <sylvio.alves@espressif.com>
Moved all MBEDTLS dependencies from prj.conf
to Kconfig as WiFi depends on it.
Update esp32 wifi driver to enable `samples/net/wifi`
to work. Commands as such as `wifi connect` and `wifi scan` are now
available.
Signed-off-by: Sylvio Alves <sylvio.alves@espressif.com>
* reduced cyclomatic complexity
* group validation by family to make the validation easier to understand
and extend
* change preprocessor markup where possible to allow for complete code
elimination when features (esp. IP) are disabled
* renamed net_context_get/set_ip_proto() to net_context_get_proto()
While the latter is formally part of the public API and might therefore
have to be deprecated rather than renamed, it is considered internal API
by the net developers, see
https://github.com/zephyrproject-rtos/zephyr/pull/48751#discussion_r942402612
Signed-off-by: Florian Grandel <jerico.dev@gmail.com>
As of today <zephyr/zephyr.h> is 100% equivalent to <zephyr/kernel.h>.
This patch proposes to then include <zephyr/kernel.h> instead of
<zephyr/zephyr.h> since it is more clear that you are including the
Kernel APIs and (probably) nothing else. <zephyr/zephyr.h> sounds like a
catch-all header that may be confusing. Most applications need to
include a bunch of other things to compile, e.g. driver headers or
subsystem headers like BT, logging, etc.
The idea of a catch-all header in Zephyr is probably not feasible
anyway. Reason is that Zephyr is not a library, like it could be for
example `libpython`. Zephyr provides many utilities nowadays: a kernel,
drivers, subsystems, etc and things will likely grow. A catch-all header
would be massive, difficult to keep up-to-date. It is also likely that
an application will only build a small subset. Note that subsystem-level
headers may use a catch-all approach to make things easier, though.
NOTE: This patch is **NOT** removing the header, just removing its usage
in-tree. I'd advocate for its deprecation (add a #warning on it), but I
understand many people will have concerns.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Supplicant start call is part of esp_wifi_init, which
would become duplicated if kept in here.
Signed-off-by: Sylvio Alves <sylvio.alves@espressif.com>
Update Wifi drivers to use DT_HAS_<compat>_ENABLED Kconfig symbol
to expose the driver and enable it by default based on devicetree.
Signed-off-by: Kumar Gala <galak@kernel.org>
wifi drivers that depends on native ethernet stack cannot perform
wifi API calls missing availability. This changes adds the ethernet_api
interface in wifi_mgmt so that it becomes possible.
Naming "offload" in "struct net_wifi_mgmt_offload" is kept because
Zephyr still has no supplicant to handle a full non-offloaded driver.
Signed-off-by: Sylvio Alves <sylvio.alves@espressif.com>
This change reverts a portion of a prior change that results
in an uninitialized eswifi_spi_data struct to be passed to
the `is_spi_ready` function.
See commit 48c87f2bf8
Signed-off-by: Brian Dunlay <brian.dunlay@gmail.com>
This PR adds a KConfig option that allows moving
all .noinit content related to wifi a net stack into external
ram. This free dram space to application.
Linker script files are also modified so that the content
are mapped into external ram.
Signed-off-by: Sylvio Alves <sylvio.alves@espressif.com>
Following zephyr's style guideline, all if statements, including single
line statements shall have braces.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
Replace eswifi_gpio with gpio_dt_spec. We can replace devicetree
label based access with this change as we try and phase out the
usage of the 'label' property.
Signed-off-by: Kumar Gala <galak@kernel.org>
If the preprocessor constant ESP_MAX_DNS is defined as 0,
Coverity reports a condition that can never be true, and some
dead code as well. This also violates MISRA rules on dead code.
This is caused by this segment in drivers/wifi/esp_at/esp.h:
#if defined(CONFIG_WIFI_ESP_AT_DNS_USE)
#define ESP_MAX_DNS MIN(3, CONFIG_DNS_RESOLVER_MAX_SERVERS)
#else
#define ESP_MAX_DNS 0
#endif
Fix this by never setting ESP_MAX_DNS to 0, as is the currently
case. Define ESP_MAX_DNS only if it is configured through
CONFIG_WIFI_ESP_AT_DNS_USE and CONFIG_DNS_RESOLVER_MAX_SERVERS:
#if defined(CONFIG_WIFI_ESP_AT_DNS_USE)
#define ESP_MAX_DNS MIN(3, CONFIG_DNS_RESOLVER_MAX_SERVERS)
#endif
Since CONFIG_DNS_RESOLVER_MAX_SERVERS is 1 or greater, ESP_MAX_DNS
will always be greater than 0.
If, on the other hand, ESP_MAX_DNS is not defined, relevant
functions will be reduced to empty stubs, since in that case
they do not make sense.
There could be a cleaner solution to this, but this one is the
least intrusive (involves less code changes).
Coverity-CID: 219490
Coverity-CID: 219513
Coverity-CID: 219520
Coverity-CID: 219524
Signed-off-by: Aleksandar Markovic <aleksandar.markovic.sa@gmail.com>
Logging v1 has been removed and log_strdup wrapper function is no
longer needed. Removing the function and its use in the tree.
Signed-off-by: Krzysztof Chruscinski <krzysztof.chruscinski@nordicsemi.no>
Files including <zephyr/kernel.h> do not have to include
<zephyr/zephyr.h>, a shim to <zephyr/kernel.h>.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
This driver uses stack buffers to hold AT command strings which are
generated at runtime using sprintk. The buffers are only sized for the
expected range of values, not the full possible range given the datatypes
involved. Values outside this expected range could cause a buffer overflow.
To mitigate this, increase the size of each buffer to hold the full range
of each parameter type.
Signed-off-by: Keith Packard <keithp@keithp.com>
Move away from 'modem_pin' abstraction as it has not obvious value compared
to generic 'gpio_dt_spec'.
Signed-off-by: Marcin Niestroj <m.niestroj@emb.dev>
UART device pointer is already used directly when configuring modem
interface. Convert the case when UART is reconfigured when 'target-speed'
DT property is specified.
Signed-off-by: Marcin Niestroj <m.niestroj@emb.dev>
PHY TX power configuration must be added into soc level.
It was previously hardcoded in hal_espressif, which was removed.
Signed-off-by: Sylvio Alves <sylvio.alves@espressif.com>
Some sockets (UDP sockets at least) do not generate "<N>,CLOSED"
messages when the WiFi network drops. As a result the networking stack
thinks these sockets are still open after the network has dropped, and
after any subsequent reconnections.
This affects the DNS resolver library in particular, which leaves UDP
sockets open permanently by default.
Manually close these sockets when the network drops to ensure a clean
state the next time the network connects.
Signed-off-by: Jordan Yates <jordan.yates@data61.csiro.au>
`esp_close_work` can be queued from the `on_cmd_closed`, which clears
`ESP_SOCK_CONNECTED` and sets `ESP_SOCK_CLOSE_PENDING`, but does no
further work. The receive callback should still be run with no data when
the socket is closed through this mechanism.
Signed-off-by: Jordan Yates <jordan.yates@data61.csiro.au>
Log a message when the modem asynchronously closes a link. This is
useful information to the user as it can explain the root cause of later
failures.
Signed-off-by: Jordan Yates <jordan.yates@data61.csiro.au>
Use the proper `%p` printf specifier when printing memory addresses,
instead of casting to an integer, which may not be the same size as a
pointer.
Signed-off-by: Jordan Yates <jordan.yates@data61.csiro.au>
In order to bring consistency in-tree, migrate all drivers to the new
prefix <zephyr/...>. Note that the conversion has been scripted, refer
to #45388 for more details.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Add a separate macro for registering offloaded sockets implementation,
along with information in the structure whether the implementation is
offloaded or not. This allows to differentiate between native and
offloaded socket implementations, which is critical for binding socket
API with an interface.
Signed-off-by: Robert Lubos <robert.lubos@nordicsemi.no>
Instead of keeping a boolean informing whether a network interface is
offloaded at socket layer or not, keep a pointer to a function which
allows to create an offloaded socket. Native interfaces keep this as
NULL, while for offloaded interfaces it allows to connect an offloaded
socket implementation with an interface.
Signed-off-by: Robert Lubos <robert.lubos@nordicsemi.no>
