| From: | Oliver Roberts |
| Date: | 9 Aug 2000 at 12:23:12 |
| Subject: | Re: OGR/RC5 and MorphOS |
On 09-Aug-00 10:56:49 BST, Michael.J.Every wrote:
>> Core #0 or #1 may be faster under 68k emulation.
> Hmm, I suppose its worth a try. But since its supposed to be emulating an
> 060 I assumed that the 040/060 core would be the best. I'll try the other
> cores later.
Although it emulates the 060 instruction set, I very much doubt it emulates
exactly the way the hardware, caches and pipelines work. The RC5 cores
are written in assembly, and highly tuned to these kind of factors. Of
course, it's academic anyway, since it's pretty pointless using the 68k
client with MorphOS :)
>> BTW, does the PowerUp client run ok under MorphOS? If so, I'd be
>> interested in any benchmarks.
> Strangely, the newer PowerUP client doent work! It say something allong
> the lines of needing ppc.library v 46.something or higher... but MorphOS
> claims to have something crazy like ohh, I think V47+! But the older
Congrats - you've found a bug :) I've now fixed this for the next release.
> (v2.7100.413) version works. -benchmark gives 1173802 keys/sec!!!!!!!
> Wether that is accurate or not I dont know (Somehow I doubt it) But if it
> is... Wheeeeeeoooohhhh!!!!!! :-)
I'm sure that's totatally inaccurate. Even a native MorphOS client would
have trouble cracking keys any higher than 812kkeys/sec on a 240MHz 603e.
It's interesting that the old client runs without crashing though - I was
under the impression that there would be problems running the old and new
clients under MorphOS (for techies: since the PPC cores use gpr2, but
MorphOS reserves this for it's own exclusive use - 68k emulation related).
But, maybe that has something to do with the incorrect speed read out -
try -test to make sure the cores are functioning correctly.
Quote carefully and read all ADMIN:README mails