[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20111218164027.GA1322@openwall.com>
Date: Sun, 18 Dec 2011 20:40:27 +0400
From: Solar Designer <solar@...nwall.com>
To: john-dev@...ts.openwall.com
Subject: Re: 1.7.9-jumbo
On Sun, Dec 18, 2011 at 04:48:48PM +0100, magnum wrote:
> I added functionality to relbench so it shows which formats has the
> worst/best performance gain (corresponding to min and max), maybe that
> is just what you did too. My patch is not good code but I suggest you
> add such functionality, it's very very handy.
Agreed.
> >Can someone look into the NTLM performance regression? And maybe into
> >others as well, but NTLM is the important one.
>
> Reproducible or not, I enclose the complete diff between the two
> versions. It should really be faster now and anything else is just
> "compiler randomness" that we can't do much about. In all versions of
> set_key() (they are different for normal/utf-8/codepage), it should
> really be faster code.
Maybe, or maybe not. Is it faster on your machine?
> Unless this performance drop come from another part of John?
I've just tried building 1.7.9-jumbo-5 with NT_fmt_plug.c taken from
1.7.8-jumbo-8. This has restored the original performance (same as
1.7.8-jumbo-8).
So I think that either the changes in NT_fmt_plug.c are actually
problematic, or they trigger some other problem elsewhere (e.g., maybe
we need to group more variables into structs and more related functions
into same source files to avoid cache associativity issues).
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.
