On Fri, Sep 21, 2001 at 03:15:17PM +0200, Juergen Doelle wrote:
> 2.4.10 2.4.10 + improvement
> spinlock patch by patch
>
> U 103.77 102.14 - 1.6%
> 1 96.82 96.77 - 0.1%
> 2 155.32 155.62 0.2%
> 4 209.45 222.11 6.0%
> 8 208.06 234.82 12.9%
>
> The improvement is less than in previous posted results, because the
> pagemap_lru_lock and the lru_list_lock are already cacheline aligned
> in 2.4.10 (2.4.9).
Do you have CPU utilisation differences for these cases, as well as
pure IO bandwidth differences? It would be interesting to see just
how much the IO code's internal latency impacts on the final dbench
numbers.
--Stephen
-
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/