Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Mon, 17 Mar 2014 21:34:36 +0000
From: "Joe ." <dy3afg9@...il.com>
To: john-users@...ts.openwall.com
Subject: Re: Re: Re: Re: pbkdf2-hmac-sha512 on GPU

Thanks Lukas,

My salt is 64 characters long, and the hash 256.
However, when using ml2john.py it cuts the hash to 128. I assume this is
deliberate... due to impossible hash collisions? (It's no faster like this,
but it gives me a different error)
With my hash cut down to 128, I get the error "Assertion failed: (strlen(c)
== 128)" as before with the shorter hash/salt I found in that file.

GPUs:
2x D300 (it's a new Mac Pro)

This is the output from --list=opencl-devices (minus the CPU and one GPU,
as they are both the same)

 Device #1 (1) name: AMD Radeon HD - FirePro D300 Compute Engine
Device vendor: AMD
Device type: GPU (LE)
Device version: OpenCL 1.2
Driver version: 1.2 (Jan 28 2014 20:11:48)
Native vector widths: char 4, short 2, int 1, long 1
Preferred vector width: char 4, short 2, int 1, long 1
Global Memory: 2.0 GB
Local Memory: 32.0 KB (Local)
Max memory alloc. size: 512.5 MB
Max clock (MHz): 850
Profiling timer res.: 37 ns
Max Work Group Size: 256
Parallel compute cores: 20
Stream processors: 1280  (20 x 64)

I would run ubuntu from a live disk for testing out, but I don't think
that's possible with Macs like with PCs.

Thanks,
Joe.







On Mon, Mar 17, 2014 at 1:23 AM, Lukas Odzioba <lukas.odzioba@...il.com>wrote:

> 2014-03-16 0:19 GMT+01:00 Joe . <dy3afg9@...il.com>:
> > Thanks, I'm able to build it successfully now with line number 68 looking
> > like this.
> > /* #define DEBUG */
>
> Ok.
>
> > my /path/to/hash.txt is formatted like this:
> > grub.pbkdf2.sha512.iterations.salt.hash (where 'iterations' is ~40,000)
> >
> > However, when I run:
> > ./john --format=grub-opencl /path/to/hash.txt
> > It gives me this error:
> > Assertion failed: (strlen(c) == 128), function binary, file
> > opencl_pbkdf2_hmac_sha512_fmt.c, line 184.
> > Abort trap: 6
> >
> > From what I can understand above, 'strlen' looks like 'string length',
> > perhaps this is the fault as my salt+hash is quite long at 320
> characters?
> > Just a guess.
>
> Yes that's the issue. That's why this format is called grub-opencl
> instead of more generic pbkdf2-sha512-opencl.
> If you can tell me exact lengths of your hash and salt I'll see what I
> can do with it.
>
> > When I run it against one of the tests I found in the file
> > opencl_pbkdf2_hmac_sha512_fmt.c (the first one of line 63) and copy it
> to a
> > .txt file, I get this:
> > Loaded 1 password hash (grub-opencl, grub-opencl [PBKDF2-SHA512])
> > Self test failed (get_hash[0](0))
> >
> > When I try a test:
> > ./john --format=grub-opencl --test
> > I get:
> > Benchmarking: grub-opencl, grub-opencl [PBKDF2-SHA512]... FAILED
> > (get_hash[0](0))
>
> That's a bit more problematic, what gpu do you have?
> Can you try to test it under some linux distribution?
>
> Lukas
>

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.