Wrong log levels in some drivers, again

[SteWers]

PROBLEM DESCRIPTION

The issue #23362 is back. I tried to solve it like in #23366, but I'd no success.

I hope someone can fix this easily.

[Staars]

Yes, we always get nice surprises from upstream libraries from time to time. The underlying problem is from the NimBLE stack itself and an open issue since 2022.

Anyway ...

#define XDRV_79                    79

#include <vector>
#include <deque>
#include <string.h>
#include <cstdarg>

#include <NimBLEDevice.h>
#include <NimBLEAdvertisedDevice.h>
// #include "NimBLEEddystoneURL.h"
#include "NimBLEEddystoneTLM.h"
#include "NimBLEBeacon.h"

#undef LOG_LEVEL_DEBUG
#undef LOG_LEVEL_INFO
#undef LOG_LEVEL_WARN
#undef LOG_LEVEL_ERROR
#undef LOG_LEVEL_CRITICAL
#undef LOG_LEVEL_NONE

Add the last #undefs to the xdrv_79_esp32_ble.ino after the last #include from the BLE library like you see above.
If this works for you, then you can add this to your currently open PR.

I will update the other BLE related stuff a bit later.

[SteWers]

Thanks for the hint. That solves it in #24188.
I became aware a of it, because I see RSL: BLE = {"blerestart":"no adverts in 120s"}. The Age at the BLE settings page is set to zero periodically. So I don't know why BLE restarts because of no adverts.