Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Tue, 24 Jul 2012 05:20:52 +0200
From: magnum <john.magnum@...hmail.com>
To: john-dev@...ts.openwall.com
Subject: Re: Current -fixes GPU formats vs TS

On 2012-07-24 05:05, jfoug wrote:
> From: magnum [mailto:john.magnum@...hmail.com]
>> pwsafe-cuda should find 1500, finds none.
>> md5crypt-opencl should find 1500, finds none.
>> phpass-opencl should find 1470, finds none.
>> nt-opencl should find 1478, finds none.
>> 
>> Some of the last four had wrong figures in TS too, I will fix that.
> 
> There are other things than just pw length to take into account.  But
> the tool 'is' a help.

I'm starting to wonder if this is just a strange Test Suite problem. For
example, I do this:

magnum@...l:test [master]$ ./jtrts.pl -noprelim nt-opencl
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.12.11, July 20, 2012.  By, Jim
Fougeron & others
- Testing:  John the Ripper password cracker, ver: 1.7.9-jumbo-7RC
[linux-x86-64-gpu]
--------------------------------------------------------------------------------

John Jumbo build detected.

form=nt-opencl                    guesses:    0 unk unk : Expected
count(s) (1478)  [!!!FAILED!!!]
.pot CHK:nt-opencl                guesses:    0 unk unk  [PASSED]

Some tests had Errors. Performed 1 tests.1 errors
Time used was 5 seconds

...so then I try this:

magnum@...l:test [master]$ ./jtrts.pl -noprelim nt-opencl -stoponerror
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.12.11, July 20, 2012.  By, Jim
Fougeron & others
- Testing:  John the Ripper password cracker, ver: 1.7.9-jumbo-7RC
[linux-x86-64-gpu]
--------------------------------------------------------------------------------

John Jumbo build detected.

form=nt-opencl                    guesses:    0 unk unk : Expected
count(s) (1478)  [!!!FAILED!!!]
Exiting on error.  The pot file ./tst.pot contains the found data
The command used to run this test was:

../run/john -ses=./tst -nolog -pot=./tst.pot NT_tst.in --wordlist=pw.dic
'-form=nt-opencl' 2>&1 >/dev/null


But here is what beats me. Right after the above -stoponerror, I do this:

magnum@...l:test [master]$ wc -l tst.pot
1478 tst.pot

So why does it say zero guesses? OK so I rm tst.pot and run the test as
printed:

magnum@...l:test [master]$ rm tst.pot && ../run/john -ses=./tst -nolog
-pot=./tst.pot NT_tst.in --wordlist=pw.dic '-form=nt-opencl' 2>&1 >/dev/null
OpenCL platform 0: NVIDIA CUDA, 1 device(s).
Using device 0: GeForce GTX 570
guesses: 1478  time: 0:00:00:00 DONE (Tue Jul 24 05:06:31 2012)  c/s:
21817K  trying: Skippin� an�* - ERS__17
Use the "--show" option to display all of the cracked passwords reliably

WTF? First, why did I get output despite sending it to /dev/null?
Reversing the order of the redirects fixes this:

magnum@...l:test [master]$ rm tst.pot && ../run/john -ses=./tst -nolog
-pot=./tst.pot NT_tst.in --wordlist=pw.dic '-form=nt-opencl' >/dev/null
2>&1
magnum@...l:test [master]$ wc -l tst.pot
1478 tst.pot

But is that the problem? No, I did the same within jtrts.pl and it does
not change a thing. I'm out of clues. Can you spot something in the above?

Whatever is the problem, it looks like nt-opencl works just fine, and
hopefully the other three do too.

magnum

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.