Thanks for the report.

Could you also enable the vMicro > Compiler > Show Build Properties, and upload another build log from vMicro, along with the Serial Monitor output?

If you also upload the log from the Arduino IDE, along with the Serial Monitor output as well (with the File > Preferences > Verbose output during Compilation and Upload checked) it will allow us to compare more easily.

On my environment both Arduino IDE and vMicro report the SHA256 mismatch, however the chip does not restart infinitely.

Is it the same in vMicro when using the SPIFFS partition layout from the board menu? (ArduinoIDE is showing SPIFFS, vMicro is currently using FFAT)

If you change the vMicro > Debugger > Optimization to "Default Optimization", then running a Build > Clean Solution in Visual Studio, then rebuilding and uploading make any difference?

Can you also confirm which pins (if any) you are using in your sketch, and that these are identical between the Arduino IDE and Visual Micro?

We cannot replicate this currently with our official ESP32C3 board, can you confirm which specific board you have, and if possible provide the sketch as well?

Thanks for the sketch and updated information.   

My board resets every 30s with this sketch due to the infinite looping if pin #1 isn't put high, can you confirm this isn't the issue?

Can you also confirm changing to SPIFFS is still making the board restart constantly? (as you said it isn't the same in the previous reply)

Can you run the Build > Clean Solution again, and then attach the full build output and serial output from the board from vMicro?
(This allows us to see the full compilation process for comparison without any caching)

Thanks for the detail.

If you upgrade your M5Stack board package to the latest version (2.0.2) does this resolve the issue?

We are working on a fix to the packages conflicting and we will update in the coming days when this is available.