[discuss] rep; nop question

Jan Hubicka jh at suse.cz
Fri Mar 11 11:54:51 CET 2005


> Hi!
> 
> > > > Could you please tell me what is the effect of the rep; nop sequence on AMD64?
> > > > It does not seem to be documented anywhere in the AMD docs ...
> > > 
> > > Probably same is intel. Basically "try not to burn too much power".
> > 
> > No, it is equivalent to "nop" just with 2 byte encoding.  Since decoding
> > speed is bound by number of instruction, sequence of nops with reps
> > decode faster than sequence of nops.
> 
> I don't think so. IIRC it is magic on intel. Its like "nop but let the
> other HT sibling run", or something like that.

Yes, it is pause on intel and fast padding on Athlon (at least according
to my recollection).

Honza
> 								Pavel
> -- 
> People were complaining that M$ turns users into beta-testers...
> ...jr ghea gurz vagb qrirybcref, naq gurl frrz gb yvxr vg gung jnl!



More information about the discuss mailing list