3DES (hc 3.40): K1 == K3 and K1 != K2 possible?

[al1984]

Hi,
Thanks for the new version. Mode 14000 (des) works well. For mode 14100 (3des), is it possible to implement K1=K3, and K1 is independent to k2?
Thanks.

[atom]

Sure it is, but you need to patch the kernel yourself. That's why it's open source

[al1984]

I have edited the 3 m14100 files in order to try and get k1=k3 and k1 independent of k2, and compiled hashcat. Now I get integer overflow.
Any help on how to set k1=k3 and k1 independent of k2 properly would be appreciated.
Sorry if this is the wrong thread to post in.
Thanks.

[philsmd]

What do you mean by integer overflow? Did you not see that error message before. Please be more specific.
Most of the time when users report an integer overflow is when the job would take kinda millions of years with the largest custer of processors/GPUs. Are you sure your keyspace is feasible?
Has it to do with your changes?

BTW: I've splitted the thread.

[al1984]

Yes, it has to do with the changes that I made.
Prior to the changes, my charset and mask was: -1 X:\xxxx\xxxx\xxxx\hashcat-3.30\charsets\DES_full.charset -2 ?1?s --hex-charset ?2?2?2?2?2?2?2?2?2?2?2?2?2?2?2?2?2?2?2?2?2?2?2?2
and that worked, but I believe that it was looking for k1, k2, k3 all different. I added the ?s to charset 2 to avoid getting "mask is too small".

My question is, in order to get k1==k3, and k1!=k2, do I have to make changes to the 3 m14100 files, or do I have to setup the mask correctly, or both?
Thanks.

[atom]

it should be enough to set the mask correctly, but I think you will still run into integer overflow because of a too large keyspace