Arduino compilation about min HOT 4 CLOSED

That's me being sloppy: there's an implicit assumption integers are 32 bits. Bad boy!

Needs to be cast to a 32-bit integer first before shifting.

You can't change the shifts: it will all explode. Basically this is assembling a 32-bit number byte-by-byte. The compiler is normally smart enough to work out what's going on. Let me fix it.

from min.

0a20297

from min.

I would rewrite this as follows using almost identical code for each additional byte (is there a better way?)

        case RECEIVING_CHECKSUM_3:
            self->rx_frame_checksum = byte;
            self->rx_frame_checksum << 8;
            self->rx_frame_state = RECEIVING_CHECKSUM_2;
            break;
        case RECEIVING_CHECKSUM_2:
            self->rx_frame_checksum |= byte;
            self->rx_frame_checksum << 8;
            self->rx_frame_state = RECEIVING_CHECKSUM_1;
            break;
        case RECEIVING_CHECKSUM_1:
            self->rx_frame_checksum |= byte;
            self->rx_frame_checksum << 8;
             self->rx_frame_state = RECEIVING_CHECKSUM_0;
            break;
        case RECEIVING_CHECKSUM_0:
            self->rx_frame_checksum |= byte;

from min.

You could do that but on an 8-bit device it means shifting a 32-bit word, which is fairly costly (it might even try and do something like four individual rotates in a loop executed 8 times). Good compiler optimizers are generally coded to see things like "<< 16" and "<< 24" and see that it's really just writing a byte to the nth byte of a word and then optimize the code to do that directly.

Need to have a look at the assembly language to see what the Arduino compiler makes of it. GCC on the ARM is very slick at this kind of thing.

from min.