In message <42543516.8095.63A48D@generic.my-fqdn.de>
          "Andrew Warren" <fastfwd@ix.netcom.com> wrote:

> Just wrote this for a friend who needed to do simple error-
> correction.  I've previously posted (4,7) Hamming encode/decode 
> routines in assembly language, but he needed it in C, so...

Nicely done.. Only one thing though:

>     const uint8 fixbits[8] = {0x00, 0x90, 0xA0, 0x81,
>                               0xC0, 0x82, 0x84, 0x88}
> 
>     uint8 receive(uint8 x)
>     {
>         return x ^ fixbits[(encode[x & 0x0F] ^ x) >> 4];
>     }

You're declaring fixbits as an 8-byte array. That means your array elements
are numbered 0 to 7. Now, if x = 10 (say), because you're ANDing by 0x0F (to
get the low nibble), this code will try and get the low 4 bits (range 0-15),
meaning it'll try and get entry 10 from fixbits[8]. Which doesn't exist.
End result? The program jumps off into never-never land.

If I'm right (not sure if I am or not...) then this line:
         return x ^ fixbits[(encode[x & 0x0F] ^ x) >> 4];
should be changed to
         return x ^ fixbits[(encode[x & 0x07] ^ x) >> 4];
   (change:                            ^^^^^^)

I'll have a quick look at the code later... it looks like it'll compile under
gcc on Linux pretty painlessly.

Later,
-- 
Phil.                              | Acorn Risc PC600 Mk3, SA202, 64MB, 6GB,
philpem@philpem.me.uk              | ViewFinder, 10BaseT Ethernet, 2-slice,
http://www.philpem.me.uk/          | 48xCD, ARCINv6c IDE, SCSI
... "Bother", said Pooh, as he accidentally deleted his message base.
-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist