flat assembler
Message board for the users of flat assembler.
Index
> Main > reversing a modolus? 
Author 

zhak
I'm not sure that I understood you correctly...
Do you mean that you have numbers 9 and 19 and you want to find X, where X mod 19 = 9? If so then there is not only B8 which will give you 9 when moded with 19. What operand sizes are used? One byte only? 

03 May 2006, 07:53 

Madis731
Simple answer: you can't!
You only know B8%19 and 19, but you don't know B8 neither 7. You have to know at least one of them. When you've got 9 and 19 then the B8 can be anything from 9, 28, ... to 19*n+9. For the solution  if you have the neccessary variables  yours is correct and not very costy in cycles. 

03 May 2006, 08:26 

shism2
Ok in decimal not in hex my solution works.
a mod b = c a = [(b * (a/b)] + (c) " without the remainder of a/b Ex : 99 mod 19 = 4 99 = (19*(a/b)) +4 Hwoever I dont have a 

03 May 2006, 18:49 

shism2
Im using an encryption for a program based on modolus .. And I need a way to find a..... This is the only way I have .... But with a... But in the program I won't be able to have a


03 May 2006, 19:17 

shism2
Code: 19 mod 19 = 0 20 mod 19 = 1 21 mod 19 = 2 22 mod 19 = 3 23 mod 19 = 4 24 mod 19 = 5 25 mod 19 = 6 26 mod 19 = 7 27 mod 19 = 8 28 mod 19 = 9 29 mod 19 = 10 30 mod 19 = 11 31 mod 19 = 12 32 mod 19 = 13 33 mod 19 = 14 34 mod 19 = 15 35 mod 19 = 16 36 mod 19 = 17 37 mod 19 = 18 38 mod 19 = 0 39 mod 19 = 1 40 mod 19 = 2 etc etc..... It will keep going like this..... 

03 May 2006, 19:43 

zhak
One of the properties of congruences is symmetry, that's:
if a = b (mod m), then b = a (mod m). Things are very simple when m = 2. It's just XOR, you know. But if modulo != 2 than you cannot do it in just a few cycles. The simplest solution that comes to my mind right now is: 1. convert numbers to base m (19 in your example) 2. apply XOR equivalent to them. 3. conevrt back to base 2 (or 16  it doesn't matter). or you can work with "base m" numbers all the time. some math functions should be implemented then. But I cannot give the optimal solution at this time. The proplem is interesting and i'll think of it for sure... when I have time. visit http://mathworld.wolfram.com/. 

04 May 2006, 07:13 

zhak
BTW, you'll have more than one solution of this problem. Maybe selecting the minimum one (or, for example, the second from the mininum) for the chosen operand size is the way to avoid errors when decipher.


04 May 2006, 07:23 

zhak
I was trying to solve the problem, but was confused a bit. But now I think I've got the idea of what you wanna do. Is it a multiplicative cipher you're trying to implement? Please, provide the community with your algorithm
If it is so, then the simplest form of a multiplicative cipher is: encipher: ciphertext=msg*key (mod m) decipher: msg=ciphertext*(1/key) (mod m) to find 1/key (mod m) you should use Eucledian algo. I can provide you with formulas to implement it... if you need. 

04 May 2006, 09:48 

Vasilev Vjacheslav
zhak, we will be very appreciate for this formulas, i am also interested in this problem


04 May 2006, 10:07 

shism2
(input mod key)0C4h)
This is what Im using I really can't solve it sigh... Last edited by shism2 on 04 May 2006, 23:23; edited 1 time in total 

04 May 2006, 23:07 

shism2
I believe it's only possible by a bruteforce of such


04 May 2006, 23:10 

zhak
Vasilev Vjacheslav, here's how to find 1/a (mod b)


05 May 2006, 18:30 

shism2
So do you have a solution for my problem?


05 May 2006, 19:03 

< Last Thread  Next Thread > 
Forum Rules:

Copyright © 19992020, Tomasz Grysztar. Also on YouTube, Twitter.
Website powered by rwasa.