Re: Optimized copy&move (was: Re: [PATCH] Mantaining turnstile aligne d to 128 bytes in i386 CPUs)

On Wed, 17 Jan 2007 14:41:44 -0500
Ivan Voras <ivoras_at_fer.hr> wrote:

> Bruce Evans wrote:
> 
> > And MMX/XMM registers ar not needed to get movnt on machines with
> SSE2,
> > since movnti is part of SSE2.  This reduces the advantages of using
> MMX/XMM
> > registers on P4's and A64's in 32-bit mode to the non-nt parts of the
> > above (fully cached case), which I think are less important than the
> nt
> > parts.
> 
> Hmm, I'm looking at i386/i386/support.s and there are several versions
> of bcopy and bmove functions, including some that optimize by using FPU
> registers (large_i586_bcopy_loop), and a version that uses movnti
> (sse2_pagezero), but I can't find the bit of magic which glues them to
> bzero() call.
> 
> Also, as as I can tell by the comments, the FPU version works by
> manually saving context... why is this possible (i.e. won't something
> preempt it?)
> 

Potentially stupid question but, is it not possible to benchmark these
variations at build or boot time and use the most appropriate method? Or
at least the one most appropriate 90% of the time?

Nick