ESP-WROOM-32D

[peter--s]

Hello,

I wonder why all configurations use
sdkconfig_options: CONFIG_FREERTOS_UNICORE: y
although the lamps come with ESPWROOM32D, where D stands for dual (core) (cf. e.g. https://github.com/currantlabs/esp-idf-1/blob/master/docs/en/hw-reference/modules-and-boards.rst).

Apparently esphome flashes the unicore bootloader for reasons that I do not grasp, which leads to
Running on single core variant of a chip, but app is built with multi-core support
(together with the core dump message quoted below) unless
CONFIG_FREERTOS_UNICORE: y
is used.

HomeSpan/HomeSpan#490 points towards a similar issue caused by Arduino IDE (and the discussion drifts away over partly related
Core dump flash config is corrupted!
and its mitigation using appropriate flashing methods.

Maybe this has been fixed in Arduino IDE in the meantime and matching bootloaders are flashed, but unfortunately I had conducted my experiments with a very old version.

Anyway, I think multi-core builds could be used on these lamps if the matching bootloader gets flashed, too.

Thoughts?

[syssi]

As far as I know most Yeelight devices are equipped with single core ESPs.

[peter--s]

What is the base of this "knowledge"?

Most photos of Yeelight SOCs I have seen (including those that I have taken) feature the tag "ESPWROOM32D".

[syssi]

Maybe I'm really barking up the wrong tree. I only concluded that a cost-optimized ESP was being used based on the experience that dual-core builds crash and Tasmota & ESPHome exclusively offer solo-core builds for Yeelights. But I can confirm that the chips are labeled with WROOM32D. I even went to the trouble in the past of desoldering the chips and replacing them with full-fledged chips. The realization that this might have been completely unnecessary and that a new bootloader would have been sufficient hurts a little. ;-)

[peter--s]

Well, I am running very short on spare time at the moment without improvement in sight and I have not managed to dig into this further, but I do not see why the documentation should be wrong.
I hope it is really sufficient to Flash a different BL, at least I could not think of anything else that could cause what we observed.