[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: /etc/passwd migration between systems



As Jim Worthington said:
> 
> Also I get the following message when trying to run crack ver 4.1:
> 
>  Crack v4.1f: The Password Cracker, (c) Alec D.E. Muffett, 1992
> 
>  Version of crypt() being used internally is not compatible with
>  standard.
> 
> Has anybody successfully been able to run crack on Debian Linux 1.1?
> 
> Jim
> 

Randy Gobbel and I are looking at Crack right now on Linux (kernel 2.0.12).

If you make FCRYPT equal to null in Sources/Makefile, and possibly remove
references to FCRYPT, you can successfully compile and run Crack. To me it
seems like the Makefile is configure to either use fcrypt or UFC, but not a
standard crypt. This is because most standard crypts are SLOW. But the
Linux crypt is very fast. Change the Makefile like so:


FCRYPT=


However.....

Using the default -O optimization makes Crack stop when it begins applying
the rules that append numeric digits to the end of a word (near the end of
dict.rules).  Randy's mahcine (586) freezes, while mine (486) merely dumps
core.

If you replace -O with -g, in an attempt to run gdb on it, Crack runs fine
all the way through the end of the rules. But, when Crack is about to write
the final information to file, it messes up again. On Randy's machine,
inodes get messed up. On mine, the file is written to disk, but under a
garbage name.... 250+ characters of random punctuation and control
characters. I did not count the number of characters... 256 would be an
interesting number.

Randy & I aren't sure what's causing the problem. Any help at all would be
appreciated. There are 3 places I see that could contian a bug:

1) Crack
2) libc
3) the kernel

This is my first time running Crack, so I don't know if it ran well before
on 1.2.x kernels, or on other Unices.

libc might be at fault, because of the weird disk events that happen when
Crack finishes.

Or the kernel might be at fault... my Crack (!) dumps core, but Randy's
machine totally freezes.

If anyone would like to compile Crack and help in our effort to find the
bug, you can get the sources at http://www.deter.com , in one of the
sub-pages.

Thanks!

--gilbert

______________________________________________________________________
Gilbert Ramirez Jr.                     gram@merece.uthscsa.edu
University of Texas                     http://merece.uthscsa.edu/gram
Health Science Center at San Antonio    University Health System



Reply to: