Francesco Pace

Thanks for the feedback @Josverl and for offering to test on additional devices! **Re: Module architecture** - I created a separate module following similar patterns in the codebase, but integrating...

@Josverl , I've updated the PR description to reflect the current API implementation. The CSI functionality is now exposed as direct methods on the WLAN object rather than a singleton...

> Interesting feature! > > I've had a quick look at the code, and there are a few things that you may want to look into if you want to...

> Thanks for following up with this! > > > Regarding the footprint concern, I should clarify that CSI must be compiled into the firmware - it cannot be enabled...

@agatti Instead of adding `CONFIG_ESP_WIFI_CSI_ENABLED=y` to `sdkconfig.base`, I introduced dedicated boards `ESP32_CSI_C6` and `ESP32_CSI_S3` with CSI pre-configured and optimized. **Rationale for dedicated boards:** 1. **Isolation**: CSI-specific optimizations (power management disabled,...

> Is there a reason why variants can't be used ? That is the way that the same board can be used with different software options, which is what you...

> Thank you for making those changes to the code! I've looked into this a bit further and I believe this can be shrunken down a little bit more in...

> > I had assumed that was the correct approach, but it seems I was mistaken. > > I personally don't think there is a Right or Wrong here. With...

> Then, just run the modified firmware a few times to experience different kinds of fatal error messages, whilst recording the serial output from boot until the crash dump including...

@agatti you were right about the `bytearray_by_ref` issue. However, that alone didn't solve the crash. Added `MP_REGISTER_ROOT_POINTER(void *csi_ringbuf_root)` to prevent the garbage collector from freeing the buffer while the ISR...