On Thu, 5 Sep 2002, Jamie Lokier wrote:
> Gabriel Paubert wrote:
> > Now that is grossly inefficient ;-) since you can save one instruction by
> > moving roll after adcl (hand edited partial patch hunk, won't apply):
>
> Yes but is it _faster_? :-)
Hard to tell, with OOO engine and decoder constraints. But once again it
is in the out of mainline code path for odd buffer addresses, not in the
loop, so its performance is not critical. Actually code size may have more
impact it ends up spanning one more cache line (or even a 16 byte block
used as fetch unit by P6 cores).
>
> I've been doing some PPro assembly lately, and I'm reminded that
> sometimes inserting instructions can reduce the timing by up to 8 cycles
> or so.
The one instruction that you can still be moved around easily is the
pointer increment. But I would never try to improve code paths that I
consider non critical.
Gabriel.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Sat Sep 07 2002 - 22:00:24 EST