05-15-2012, 10:15 AM
(05-15-2012, 07:16 AM)halfie Wrote:(05-15-2012, 04:04 AM)perlish Wrote:(03-18-2012, 07:23 PM)atom Wrote: I am not planning it. Try JtR. Its the king of DES @ CPU. Any reason why you want it in hashcat?
JTR can only run in one CPU core,but hashcat can run in all CPU core.If in a 16 CPU cores system,hashcat will be 16 times faster than JTR.
Not true. You can compile JtR with OpenMP to enable usage of all cores. Edit src/Makefile to do this.
(03-19-2012, 08:02 AM)Kakoose Wrote: Yeah, I use JTR for all my des(unix) hashes but nearly all hash types I've tried with hashcat are [much] faster than most other programs, I thought it would be great if hashcat could add it, ergo cracking des faster.
p.s. little off-topic, since i'm not really one with the patches or technical files related to jtr, would you mind telling me which community patch would be best/fastest for cracking des unix hashes using JTR (there are patches e.g. jumbo-5, omp-des-4, omp-des-7 or even jumbo-2-omp-des-4-mscash2)? if you know, could you point out which would be the most suitable? thanks!
Try jumbo-5 patch and enable OpenMP / MPI in Makefile.
Latest code is at https://github.com/magnumripper/magnum-jumbo
I've tried jumbo5-JTR with MPI,but it make me sad,it's not REALLY MPI,it only run server process to do the same thing.
you can do a test for it.
1.add a user test123 with password test123(or sth others in the dictionary) in a linux.
2.use JTR with MPI to crack the shadow file. mpirun -n 8 ./john /etc/shadow .
3.you will see the john will crack the password "test123" for 8 times.
so I thought JTR+MPI did not speed up the crack.