psram: xip_psram support on c5/c61, also fixed cache writeback/invalidate not work issue on c61
Closes IDF-8688, IDF-9292, and IDF-11008
See merge request espressif/esp-idf!33265
Add AP-STA mode support in FTM example
Update wifi libs with below changes -
- Use ex-AP's bandwidth in FTM request when connected
- Allow Bandwidth downgrade in FTM negotiation
- Use separate calibration when Initiator performs 20MHz FTM
while in-AP is in 40MHz
Currently, USB Host unit tests that require a software triggered disconnection/
reconnection rely on the 'usb_phy_action()' function.
This commit replaces those calls with 'hcd_port_command()' or
'usb_host_lib_set_root_port_power()'.
Note: Also removed 'test_usb_common.h/c' as it is no longer necessary are the
function call replacements.
This commit adds the usb_host_lib_set_root_port_power() function. This provides
a public API for users to power the root port OFF or ON at runtime, thus trigger
a disconnection or allow connections respectively.
In addition, the usb_host_config_t.root_port_unpowered install configuration is
provided to allow users to install the USB Host Library without automatically
powering ON the root port.
test(esp_flash): Enable test for spi_flash component for c61/c5/p4
Closes IDF-10313, IDF-8715, IDF-9314, IDF-10991, and IDF-8985
See merge request espressif/esp-idf!33238
On p4, the clock was configured in the wrong order causing a dropped
count on the first transmission. And gdma eof event fails to trigger
parlio_tx eof event when transmitting single byte.
This feature serves as an immediate fix in case
the new export approach causes issues.
The user can simply set the environment variable `ESP_IDF_LEGACY_EXPORT`,
and the old export script will be used.
fix(wifi): Add PSRAM failure fallback in WiFi OSI API's
Closes WIFIBUG-700, WIFIBUG-705, WIFIBUG-713, WIFIBUG-736, and WIFI-6592
See merge request espressif/esp-idf!32713
The select function wrapper was rewritten to be non-blocking
on Linux systems, as it was stealing all the CPU time
from lower priority tasks when called from a higher priority task.
This is because the FreeRTOS scheduler does not know
that the task thread is sleeping during the system call.
This issue manifests all "slow" system calls on the Linux target,
but handling the case of select fixes the problems for most ESP-IDF components.
The FreeRTOS POSIX port documentation lists this as a known issue,
so user code is responsible handling this case if other system calls are used,
even if unknowingly.
This closes GH issue #14395 "select() blocks the FreeRTOS scheduler on Linux target"
