Cracking MD5 of 270 length plain-text -partially know-
#1
Hello

First of all, thank you all for developing/supporting hashcat! 

I've a one question and couldn't find an answers so far. Would you like to help me ? Here is the issue.

MD5(KNOW_DATA + SECRET)

I know value of md5. Also I know KNOW_DATA which is 263 character length and contains colons, semi-colons, quotes or A-Za-z0-9. 

Also I know SECRET is not complicated it should be 6 or 7 length with only a-z-0-9 . 

I know time consuming of 270 length md5 calculation takes too much time which cause inefficient brute-forcing but How can I brute-force SECRET with hashcat ? I've tried several thing but it end up with Segfault error. 


UPDATE: Ofc. I've tried -m 20 but since KNOW_DATA contains junk of characters such as : it cause some errors..
Any suggest ?
#2
You're not going to be able to use any optimized hash cracking application for this (especially not hashcat), you'll need to use or write a cracker with a generic multiblock MD5 implementation from e.g. OpenSSL or something.
#3
(11-27-2015, 10:48 AM)epixoip Wrote: You're not going to be able to use any optimized hash cracking application for this (especially not hashcat), you'll need to use or write a cracker with a generic multiblock MD5 implementation from e.g. OpenSSL or something.

Thank you for your answer. I've build my own basic python script which is works without any issue but yea I would like to use optimized hash cracking app.
#4
You're not going to be able to use an optimized hash cracking app. Part of those optimizations is removing support for multiple input blocks, which limits the input to 55 bytes.
#5
You could easily optimize for a 500% boost yourself though: Calculate the first four limbs and save a copy of the context that far. Then in the inner loop you just copy that context and do the last (1) limb for each candidate.
#6
As magnum says, you can optimise quite a bit. I forget exactly how it works, but Python's md5 has an update() method, so you can create the digest of the first chunk and simply copy it every loop. I reccommend testing with some known hashes to make sure you're doing it right; things like newlines can mak everything go wrong.
#7
To be clear, I wasn't saying he couldn't optimize his own cracker for this task. I'm saying he can't use an existing optimized cracker like oclHashcat.