Discussion:
spurious warning from generic crypt(3) format
(too old to reply)
Solar Designer
2016-02-29 19:48:13 UTC
Permalink
magnum -

It was reported to us that jumbo prints a spurious warning before
proceeding to crack SHA-crypt hashes on systems that lack support for
those hashes in crypt(3), like this:

xxxxx-MacBook-Pro:run xxxxxxxx$ ./john passwd.txt
Warning: detected hash type "sha512crypt", but the string is also recognized as "sha512crypt-opencl"
Use the "--format=sha512crypt-opencl" option to force loading these as that type instead
Warning: hash encoding string length 106, type id $6
appears to be unsupported on this system; will not load such hashes.
Using default input encoding: UTF-8
Loaded 1 password hash (sha512crypt, crypt(3) $6$ [SHA512 128/128 AVX 2x])
Press 'q' or Ctrl-C to abort, almost any other key for status

Perhaps the code in loader.c that tries to delegate these hashes to the
generic crypt(3) format should be disabled in jumbo when its own
SHA-crypt formats are compiled in. Would you implement this logic?

Thanks,

Alexander
magnum
2016-02-29 20:53:51 UTC
Permalink
Post by Solar Designer
Perhaps the code in loader.c that tries to delegate these hashes to the
generic crypt(3) format should be disabled in jumbo when its own
SHA-crypt formats are compiled in. Would you implement this logic?
https://github.com/magnumripper/JohnTheRipper/issues/2069

Actually I can't think of any crypt(3) variant that we lack specific
support for now. Maybe Jumbo should simply never use the generic
crypt(3) (or even list it as an alternative) unless specifically requested?

magnum
Solar Designer
2016-02-29 21:08:47 UTC
Permalink
Post by magnum
Post by Solar Designer
Perhaps the code in loader.c that tries to delegate these hashes to the
generic crypt(3) format should be disabled in jumbo when its own
SHA-crypt formats are compiled in. Would you implement this logic?
https://github.com/magnumripper/JohnTheRipper/issues/2069
Actually I can't think of any crypt(3) variant that we lack specific
support for now. Maybe Jumbo should simply never use the generic
crypt(3) (or even list it as an alternative) unless specifically requested?
Yes, this is how it should be.

Alexander

Loading...