Openwall GNU/*/Linux - a small security-enhanced Linux distro for servers
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Tue, 23 Jul 2013 03:59:59 +0200
From: magnum <john.magnum@...hmail.com>
To: john-dev@...ts.openwall.com
Subject: Re: PBKDF2-HMAC-SHA256

On 22 Jul, 2013, at 21:53 , Dhiru Kholia <dhiru.kholia@...il.com> wrote:
> On 07/22/13 at 08:36pm, Lukas Odzioba wrote:
>> 2013/7/16 Lukas Odzioba <lukas.odzioba@...il.com>:
>>> Benchmarking: PBKDF2-HMAC-SHA256, rounds=12000 [PBKDF2-SHA256]... (4xOMP) DONE
>>> Raw:    704 c/s real, 176 c/s virtual
>> 
>> Attached OpenCL version.
> 
> Committed "pbkdf2-hmac-sha256-opencl.diff" to bleeding-jumbo.
> 
> There is *no* need to use the "any_cracked" approach in this format. Let
> us try not to repeat my past "sins" ;)

Right. This is better, although it doesn't affect speed for a slow format like this:

static int cmp_all(void *binary, int count)
{
	int i;

	for (i = 0; i < count; i++)
		if (host_crack[i].hash[0] == ((uint32_t *) binary)[0])
			return 1;
	return 0;
}

The occasional false positives will be rejected by cmp_one(). This will happen once in about 70 days at this speed...

magnum

Powered by blists - more mailing lists

Your e-mail address:

Powered by Openwall GNU/*/Linux - Powered by OpenVZ