04-05-2015, 09:00 PM
That's an interessting question. Thing is, we dropped native permutation attack from oclHashcat a while back. But even then it was not really an option becuase there was no way to add a separator character like you would need.
But then, you don't really need to write the 166gb to file. You can simply pipe the output from your scripts (make them print to stdout) and oclHashcat can read it from stdin. That is of course slower than a native kernel doing it, but if the hash youwant to crack is a slow one that doesnt matter anyway.
But then, if your hash is a fast one you can simply mask attack this thing. It would be so fast that you can simply ignore the permutation effect, check the invalid but therefore you get the full gpu acceleration.
But then, you don't really need to write the 166gb to file. You can simply pipe the output from your scripts (make them print to stdout) and oclHashcat can read it from stdin. That is of course slower than a native kernel doing it, but if the hash youwant to crack is a slow one that doesnt matter anyway.
But then, if your hash is a fast one you can simply mask attack this thing. It would be so fast that you can simply ignore the permutation effect, check the invalid but therefore you get the full gpu acceleration.