Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Fri, 16 Dec 2011 18:58:38 +0100
From: magnum <>
Subject: Re: External mode (lack of) truncation

On 12/15/2011 10:53 PM, magnum wrote:
> Solar,
> I just noticed external.c does not truncate at format's max length. I
> thought it did, because I believe all other modes do. Some formats (like
> nt2) die from, for example, --ext:repeats because of this and I am
> guilty of some (or all) of them doing so :-/
> Should/could this be fixed in external.c or should I produce a patch
> that fixes the affected formats? I believe doing it in external.c might
> be better for performance because we can do it blindly vs having a test
> inside the key setup loops.

Patch 0001 fixes external mode. I also uploaded four other patches, see 
wiki for descriptions. The PARA patch should give you a boost for MD5 if 
using gcc 4.6 and not using pre-built intrinsics. And there are now 
clauses for clang (which otherwise hit the #ifdef __GCC__ clauses)


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.