Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Wed, 04 Jan 2012 21:38:02 +0100
From: magnum <john.magnum@...hmail.com>
To: john-dev@...ts.openwall.com
Subject: Re: MD5 intrinsics compile-time condition

On 12/26/2011 07:15 PM, Solar Designer wrote:
> On Sat, Dec 24, 2011 at 03:34:11AM +0100, magnum wrote:
>> If someone else can produce test results for
>> para 1, 2 and 3 for versions of gcc older than 4.4 and running on intel,
>> we can put additional clauses for them instead. Otherwise this change
>> may be detrimental for other intrinsics formats with some versions of
>> gcc. The optimal para's for MD4 and SHA1 should ideally also be tested.
>> Also, all tests should be separate for 32-bit and 64-bit...

I just checked in (to github) a patch that adds two make targets 
(testpara and testpara32) for determining the best SSE_PARA values for a 
particular system.

It's a hack but it seems pretty accurate. It averages five benchmarks of 
each and keeps increasing PARA's until no more gain found.

Example:

$ make testpara
(...lots snipped...)
======================================================
model name	: Intel(R) Core(TM)2 Duo CPU     P8600  @ 2.40GHz
gcc version: gcc (gcc version 4.6.1 (Ubuntu/Linaro 4.6.1-9ubuntu3) )
Best paras:
   raw-MD4: 3  (44108K c/s)
crypt-MD5: 3  (25634 c/s)
  raw-SHA1: 3  (12010K c/s)


Anyone running x86 or x86-64, and a gcc older than 4.4 or newer than 
4.6, please try it out and report the results.

magnum

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.