Noob Question: What to use

[bozocv]

Originally Posted by LaurV

Interesting... What online game are you trying to break?

Not a game but something gamers hate. Crypto Locker.

But it failed....

I got this in log file:

Mon Dec 14 18:47:34 2015 polynomial 'A' values have 14 factors
Tue Dec 15 10:03:33 2015 137426 relations (32204 full + 105222 combined from 2062236 partial), need 137254
Tue Dec 15 10:03:33 2015 begin with 10678 relations
Tue Dec 15 10:03:33 2015 reduce to 1222 relations in 2 passes
Tue Dec 15 10:03:33 2015 attempting to read 1222 relations
Tue Dec 15 10:03:33 2015 failed to read relation 0
Tue Dec 15 10:03:33 2015 failed to read relation 1
Tue Dec 15 10:03:33 2015 failed to read relation 2
Tue Dec 15 10:03:33 2015 failed to read relation 3
Tue Dec 15 10:03:33 2015 failed to read relation 4
Tue Dec 15 10:03:33 2015 failed to read relation 5
Tue Dec 15 10:03:33 2015 failed to read relation 6
Tue Dec 15 10:03:33 2015 failed to read relation 7
Tue Dec 15 10:03:33 2015 failed to read relation 8
Tue Dec 15 10:03:33 2015 failed to read relation 9
Tue Dec 15 10:03:33 2015 failed to read relation 10
--------
Tue Dec 15 10:03:33 2015 recovered 0 relations
Tue Dec 15 10:03:33 2015 recovered 0 polynomials

What happened?

Using:
Msieve v. 1.52 (SVN 939)

[Dubslow]

Msieve is not the only tool you must use. Read the tutorial linked in post 6, or use Yafu instead (which is also not sufficient on its own).

Edit: And what makes you think CryptoLocker uses severely-outdated RSA-512 keys?

[LaurV]

Originally Posted by VBCurtis

I thought 512-bit keys were left in the PS2/TI89 era. Hrmph.
I, too, wonder what game this is!

Yeah, otoh, the number may be fake, because is even...
Anyhow, CL I know uses 2048 bit encryption... (except of a very old version from ~2008 called Gpcode which used 1024 bit, but which is cracked already, it had some vulnerability).

[bozocv]

It is for TeslaCrypt. They say files are ecrypted with RSA but in reality they are encrypted with AES256 CBC.

There is dh parameter transfered to attackers server and is = ECDH_shared_secret(T_publickey,PrivateKey)* PrivateKey

That is why I try factoring.

PrivateKey most probably is not Prime.

[Dubslow]

Wait, were you trying to use the MPQS implementation included in Msieve? That certainly won't work for a 512 bit number. Seriously, read up the links posted here about NFS and how to do it.

[xilman]

Originally Posted by bozocv

It is for TeslaCrypt. They say files are ecrypted with RSA but in reality they are encrypted with AES256 CBC.

There is dh parameter transfered to attackers server and is = ECDH_shared_secret(T_publickey,PrivateKey)* PrivateKey

That is why I try factoring.

PrivateKey most probably is not Prime.

The complete prime factorization is

2

3

31

132533

3226496593

12455594243

15799018209107267302187

717723943183117386533741146133

266576225412799739809351907249876804115685240872207255457726285116805202111


Good luck.

Paul

[Dubslow]

Originally Posted by xilman

The complete prime factorization is

2

3

31

132533

3226496593

12455594243

15799018209107267302187

717723943183117386533741146133

266576225412799739809351907249876804115685240872207255457726285116805202111


Good luck.

Paul

At least assuming that the hex posted earlier is in fact the number to be factored, in human-standard byte order. (In reverse byte order it still has many small factors.)

[bozocv]

Was following totaly wrong information from the guy who is nowhere to be seen now, on other forum. I used wrong HEX offset and everything.... and the guy also sugested to use msieve..... Guess I will have to inform myself in more detail first.

Will do this with Yafu now.

0x37D27D3791674222AC456F061F1F740504F1AA94DC6288D0D0CB8AB3A2478086D1C7
083E0BC102006D8B0A55D5363B56016E8621C659712057739CE3FD4A4DDC

if I call factor(x) will it return prime factors only?

[Dubslow]

Originally Posted by bozocv

if I call factor(x) will it return prime factors only?

An algorithm or function that claims to factor things isn't very good at what it does if it returns as output things that may be further factored.

[bozocv]

Originally Posted by Dubslow

An algorithm or function that claims to factor things isn't very good at what it does if it returns as output things that may be further factored.

I understand.... I will try to multiply different primes for potentialy key.

For the new one result can be found on www.factordb.com

Wish me luck.

[xilman]

Originally Posted by Dubslow

An algorithm or function that claims to factor things isn't very good at what it does if it returns as output things that may be further factored.

In which case GMP-ECM isn't very good. It returned

Step 2 took 844ms
********** Factor found in step 2: 40187932388829914099
Found composite factor of 20 digits: 40187932388829914099
Composite cofactor 3022796762436865758652479689958046977795867564387867433730923774249775659151058260449786348808439252499670414861993921076650681 has 127 digits

at one point during my computations, the final results of which are posted above.