Persist accelerometer deadzone in NVS across reboots.

Each node saves its local deadzone on UART or ESP-NOW set; pod_settings loads and applies it after BMA456 init.

Co-authored-by: Cursor <cursoragent@cursor.com>

main/pod_settings.h

@@ -0,0 +1,19 @@
+#ifndef POD_SETTINGS_H
+#define POD_SETTINGS_H
+
+#include "esp_err.h"
+#include <stdint.h>
+
+/** Initialize NVS (idempotent) and log stored settings. Call once early in app_main. */
+esp_err_t pod_settings_init(void);
+
+/** Persist local accelerometer deadzone (LSB per axis). */
+esp_err_t pod_settings_save_accel_deadzone(uint32_t deadzone_lsb);
+
+/** Load deadzone from NVS, or BMA456_DEFAULT_ACCEL_DEADZONE if unset. */
+uint32_t pod_settings_load_accel_deadzone(void);
+
+/** Apply NVS deadzone to BMA456 when the sensor is present. */
+void pod_settings_apply_accel_deadzone(void);
+
+#endif