Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Date: Mon, 19 Sep 2005 22:33:57 +0400
From: Solar Designer <>
Subject: Re: Suggestions for XEON performance?

I wrote:
> > (I am planning to add "linux-x86-sse-elf" which would be the right
> > choice for P4s.  With that, they would become only a little bit slower
> > than P3s per-MHz.  Stay tuned.)

On Mon, Sep 19, 2005 at 08:19:00PM +0200, dvorak wrote:
> Is there a place where such future features are listed,


> since i started to
> look into this as well since i didn't find a reference, but maybe i should
> just ask the list.

Yes, if you want to avoid duplicate work on an enhancement to John, just
ask on this list before you start.

> > > Also should I use gcc 3.4 or 4.0?
> > 
> > It shouldn't matter.
> For the assembly stuff that is, on a P4 i noticed that gcc 3.4 uses very slow
> memset implementations (rep stosd iirc). But i guess most (if not all)
> performance critical loops are in assembly anyways.

Yes - most are in assembly.  And memset() performance is irrelevant for
John anyway.

(Offtopic - that "very slow memset" you saw might not be as slow in
practice as you think it is.  You may want to benchmark it rather than
merely look at the code and speculate.  "rep stosl" was slow on the
original Intel Pentium.  It is not that slow on P4.)

Alexander Peslyak <solar at>
GPG key ID: B35D3598  fp: 6429 0D7E F130 C13E C929  6447 73C3 A290 B35D 3598 - bringing security into open computing environments

Was I helpful?  Please give your feedback here:

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.