Andrew Warren wrote:
> > For a quick check: enciphering an all zero's data block using an
> > all zero's key gives a 0xed285da1 0x455b33c1 cipher block.
>
>     Hmm... My code returns 0x66750736 0x5CF66BDD.
>
>     There's a bug in either your code or mine (or in both)... Later
>     today, when Dr. Shepherd gives me his results, I'll know
>     whether mine is correct.

I couldn't wait so I took the 'C' reference implementation I've cited
some days ago (at http://vader.brad.ac.uk/tea/tea.shtml) and compiled it
on a powerPC UNIX machine I have. Then I ran the encipher routine on
"all zero" plaintext and key using different number of rounds. Following
there are some results:

rounds   ciphertext
01       0x9e3779b9 0xdbe8d32f
06       0xa5b716e2 0x42a412de
08       0xed285da1 0x455b33c1   <-- rounds I originally implemented
16       0xa889f798 0x182d8083
24       0x2ab23274 0xf046037b
32       0x41ea3a0a 0x94baa940

My PIC routines give the same results (whew...) so I can imagine my code
should be correct.

> > BTW: if there is someone willing to have my code I can post it here
> > (it is a good practice?) or I can send individual copy.
>
>     Programs as large as these shouldn't be sent to the PICLIST.
>     After my code is verified by Dr. Shepherd, I'll post it on my
>     web page.

Ok, I agree. I don't have a web page at this moment so, if someone wants
to take a look at my routines (I'll be very happy to hear some critics),
send me a note and I will send back the code.

Be warned that the code is not finished yet. I have to speed up things
because I currently use some math routines that are quite slow (they are
very generalized in that they use FSR/INDF for accessing the data). Also
if someone improves the routines I will be glad to hear some feedback.

Ciao
  Marco

----
Marco DI LEO                  email: m.dileo@sistinf.it
Sistemi Informativi S.p.A.    tel:   +39 6 50292 300
V. Elio Vittorini, 129        fax:   +39 6 5015991
I-00144 Roma
Italy