[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CA+TsHUBxcGwKy8tpp7sm2kWWWnxRg4_ekYQ-Zb9huQ7h=qVRkg@mail.gmail.com>
Date: Fri, 21 Sep 2012 09:04:16 +0530
From: Sayantan Datta <std2048@...il.com>
To: john-dev@...ts.openwall.com
Subject: Re: bitslice DES on GPU
Hi Alexander,
On Thu, Sep 20, 2012 at 10:20 AM, Solar Designer <solar@...nwall.com> wrote:
> I think you need to try
> DES_BS_EXPAND=0. The target without-overhead speed is ~300M.
>
I don't get the exact advantage of DES_BS_EXPAND=0;
We still need to allocate memory for the KS.p pointers. Only the expansion
of keys into KS.v doesn't needs to be done.
Or do you want me to remove KS.p too and access the Key[] array using the
index768[] in crypt_25() ?? This I think would save some memory.
Regards,
Sayantan
Content of type "text/html" skipped
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.
