mirror of
https://github.com/liquidraver/ZephCore.git
synced 2026-09-03 01:33:50 +00:00
29 lines
1.6 KiB
Plaintext
29 lines
1.6 KiB
Plaintext
/*
|
|
* Device-tree half of esp32s3_usb.conf — auto-paired by the
|
|
* zephcore_auto_pair_overlay() helper in CMakeLists.txt, so it is applied if
|
|
* and only if the build actually passes boards/common/esp32s3_usb.conf.
|
|
*
|
|
* esp32s3_usb.conf disables USB Serial/JTAG and hands the shared D+/D- pins to
|
|
* the USB OTG controller (the CDC-ACM companion channel). Any board whose
|
|
* console rides usb_serial would go silent at that moment, so the console is
|
|
* rerouted to uart0 here — in the paired overlay rather than in board.overlay,
|
|
* because board.overlay applies to EVERY build of the board regardless of role.
|
|
*
|
|
* That is the bug this file exists to prevent: xiao_esp32s3 and station_g2 used
|
|
* to include the reroute unconditionally, which silently moved the console (and
|
|
* with it the observer/repeater serial CLI, which falls back to the chosen
|
|
* console) onto GPIO43/44 in builds that never enabled USB OTG at all.
|
|
*
|
|
* Devicetree is processed before Kconfig, so an overlay cannot test
|
|
* CONFIG_ZEPHCORE_COMPANION_USB — the conf/overlay pairing is what makes this
|
|
* conditional.
|
|
*
|
|
* Boards whose console is already a physical uart0 (Heltec V3, ThinkNode M9) are
|
|
* unaffected: the include re-asserts the same uart0 + GPIO43/44 setup they
|
|
* already have. Heltec V4/V4.3 and Wireless Tracker V2 DO get moved — their
|
|
* board.overlay chooses usb_serial — which is the point, and only actually
|
|
* happens because board.overlay now goes into EXTRA_DTC_OVERLAY_FILE first
|
|
* (see the DTC overlay precedence block in CMakeLists.txt).
|
|
*/
|
|
#include "esp32s3_console_uart0.dtsi"
|