On Sat, Apr 20, 2002 at 03:43:17PM -0700, H. Peter Anvin wrote:
> Andrea Arcangeli wrote:
> >
> >As said in an earlier email it is a matter of memory bandwith, 59 bytes
> >of icache and zero data, against 7 bytes of icache and 512bytes of data.
> >the 512bytes of data are visibly slower, period. Saving mem bandwith is
> >much more important than reducing the number of instructions, even more
> >on SMP!
> >
>
> It's not 512 bytes of data -- only the part that's actually used is
> accessed.
On current x86 yes (so far), but the x86-64 the whole 512bytes will have
to be read from ram instead.
Andrea
-
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 : Tue Apr 23 2002 - 22:00:28 EST