weak checksum question

[Jeff Abrahamson]

The weak checksum in checksum.c (see snippet below) differs
substantially from the one discussed in Andrew Tridgell's doctoral
thesis on rsync and elsewhere that I've been able to find. I didn't
find discussion of the change in the mailing list archives. Well, so
I'm curious what the benefit of the change is.

(I'm using a rolling checksum in another app, thus the reason for this
nitty inspection.)

Can anyone provide a pointer to help me understand why this changed?

Thanks much.


In checksum.c:

    uint32 get_checksum1(char *buf1,int len)
    {
        int i;
        uint32 s1, s2;
        schar *buf = (schar *)buf1;

        s1 = s2 = 0;
        for (i = 0; i < (len-4); i+=4) {
            s2 += 4*(s1 + buf[i]) + 3*buf[i+1] + 2*buf[i+2] + buf[i+3] + 
              10*CHAR_OFFSET;
            s1 += (buf[i+0] + buf[i+1] + buf[i+2] + buf[i+3] + 4*CHAR_OFFSET); 
        }
        for (; i < len; i++) {
            s1 += (buf[i]+CHAR_OFFSET); s2 += s1;
        }
        return (s1 & 0xffff) + (s2 << 16);
    }


What I understood from documentation:

    /* Rolling checksum from Tridgell's PhD thesis. */
    int get_weak_checksum(signed char *buf, int len)
    {
            int s;
            int s1, s2;
            int len4;

            s1 = s2 = 0;
            len4 = len / 4;
            for(i = 0; i < len - 4; i += 4) {
                    s = (buf[i] << 24) + (buf[i+1] << 16)
                            + (buf[i+2] << 8) + (buf[i+3]);
                    s1 += s;
                    s2 += (len4 - 4 * i) * s;
            }
            return (s2 << 16) + (s1 & 0xffff);
    }



-- 
 Jeff

 Jeff Abrahamson  <http://www.purple.com/jeff/>