Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [day] [month] [year] [list]
Message-ID: <20110921203558.GA17209@openwall.com>
Date: Thu, 22 Sep 2011 00:35:58 +0400
From: Solar Designer <solar@...nwall.com>
To: announce@...ts.openwall.com, john-users@...ts.openwall.com
Subject: [openwall-announce] John the Ripper 1.7.8-jumbo-7

Hi,

I've just released John the Ripper 1.7.8-jumbo-7:

http://www.openwall.com/john/

As is usual with community-enhanced releases like this, my role was
mostly limited to making the release itself, whereas most of the code
changes are by others in the community (thanks!)

There have been lots of changes since 1.7.8-jumbo-5.  Here are some that
we managed to recall:

* Support for encrypted pkzip archives has been added, testing millions
of candidate passwords per second. (JimF)
(This is in addition to WinZip/AES archives, support for which was added
in prior -jumbo updates.)
* Support for Mac OS X 10.7 Lion salted SHA-512 hashes has been added
(enabled when building against OpenSSL 0.9.8+ only), with optional OpenMP
parallelization. (Solar)
* Optional OpenMP parallelization has been added for salted SHA-1 hashes
used by Mac OS X 10.4 to 10.6. (Solar)
* PoC support for DES-based 10-character tripcodes has been added (does
not use the bitslice DES implementation yet, hence is slow). (Solar)
* The DIGEST-MD5 authentication cracker has been revised to be usable
without requiring source code customizations. (magnum)
* Highly experimental support for dynamically loaded plugins (adding new
formats) has been added (currently only enabled on Linux). (David Jones)
* Added the ability for the john.conf file to ".include" other .conf
files and/or individual sections (e.g., a wordlist rules section may
include more rules from elsewhere). (JimF)
* John now makes an attempt to suppress duplicate rules (handy along
with the ".include" feature). (JimF)
* More character encodings are now supported.  The full list is: raw,
utf-8, iso-8859-1, iso-8859-7, iso-8859-15, koi8-r, cp437, cp737, cp850,
cp858, cp866, cp1251, cp1252, cp1253.
* Full encoding support for "single crack" mode, rules, and character
classes has been implemented. (JimF/magnum)
* Full encoding support for all formats, including md5_gen. (JimF/magnum)
* Some new character classes have been added. (JimF/magnum)
* Support for user-defined character classes has been added, along with
some samples in the default john.conf. (magnum)
* New rule reject flag "->N" (reject unless length N is supported by the
current hash/cipher type). (magnum)
* New Boolean options in john.conf: LogCrackedPasswords, AlwaysReportUTF8,
UnicodeStoreUTF8, CPstoreUTF8. (magnum)
* raw-md5-unicode has been replaced with faster 'thin' raw-md5u. (magnum)
* The generic crypt(3) module now recognizes some "subformats"
(md5/sha256/sha512), solely for benchmarking. (magnum)
* Wordlist mode is now much faster when using memory buffer and running
against a fast hash/cipher type. (JimF/magnum)
* Unicode DumbForce-like external mode samples have been added (Dumb16
and Dumb32). (magnum)
* Numerous bug, performance, and portability fixes (JimF/magnum/Solar)

Here are some Mac OS X hash benchmarks, on 2xE5420 2.5 GHz using OpenSSL
crypto code:

Benchmarking: Mac OS X 10.7+ salted SHA-512 [64/64]... (8xOMP) DONE
Many salts:     9501K c/s real, 1186K c/s virtual
Only one salt:  5489K c/s real, 684496 c/s virtual

Benchmarking: Mac OS X 10.4 - 10.6 salted SHA-1 [32/64]... (8xOMP) DONE
Many salts:     22629K c/s real, 2825K c/s virtual
Only one salt:  8967K c/s real, 1119K c/s virtual

Non-OpenMP build, same machine (using one CPU core):

Benchmarking: Mac OS X 10.7+ salted SHA-512 [64/64]... DONE
Many salts:     1457K c/s real, 1457K c/s virtual
Only one salt:  1362K c/s real, 1362K c/s virtual

Benchmarking: Mac OS X 10.4 - 10.6 salted SHA-1 [32/64]... DONE
Many salts:     3676K c/s real, 3676K c/s virtual
Only one salt:  3124K c/s real, 3124K c/s virtual

There's a lot of room for improvement (by implementing own SHA-512 and
SHA-1 code rather than using OpenSSL's), but this is representative of
the relative speed of these hashes (both are fast).

Enjoy, and please be sure to provide your feedback on john-users.

Alexander

Powered by blists - more mailing lists

Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.