Switching from Uno to Mega: warning: left shift count >= width of type

[johnwasser]

@Mattuz

const byte data_R1=40;
const byte data_R2=41;

      PORTB &=~(3UL<<(data_R1-8));                                 // data_R2 is LOW; data_R1 is LOW;
      PORTB &=~(1UL<<(clockPin-8));

Yes.

40-8 is 32 and shifting a 32-bit number by 32 bits will clear it completely.

41-8 is 33 and shifting a 32-bit number by 33 bits will clear it completely.

You switched processors without understanding the hardware-specific parts of the code. You can't just copy Direct Port Access code to a different processor.

You will need to re-write all parts the reference PORTB or PORTD.

@Mattuz

      PORTB &=~(3UL<<(data_R1-8));                                 // data_R2 is LOW; data_R1 is LOW;
      PORTB &=~(1UL<<(clockPin-8));                                // digitalWrite(clockPin,LOW);
      PORTB |= !((buffer[index]>>(7-i)) & 0x01) << (data_R1-8);  // top set of rows
      PORTB |= !((buffer[index+128]>>(7-i)) & 0x01) << (data_R2-8); // bottom set of rows
      PORTB |= 1<<(clockPin-8); 

You need to pick three pins that are in the same PORT. You need to know the bit within the PORT (the '-8' trick only works on PORTB on an UNO or Nano for Pins 8 through Pin 13).