On Mon, 9 Sep 2002, James Morris wrote:
>
> I guess it could be an existing hardware problem which has been triggered
> by some innocuous change.
Does HZ make any difference to you? That's easily tested by just editing
include/asm-i386/param.h and changing HZ to 100 to get the old behaviour..
I think that change happened earlier than your performance problem, but
since lmbench variations can be noticeable and you only had one run per
kernel, it might be one of those "comes and goes" things..
Another thing that you might check out is to disable USB if you have it
on. The memory throughput issue makes me wonder whether there might be a
lot of DMA going on, and USB (by "design", and I use the term losely) has
this silly "DMA always" approach, even if nothing really is happening on
the bus.
I don't really see anythign very suspicious in the 31->32 changelogs.
Somebody else already mentioned TLS etc, but that shouldn't really be
noticeable...
Linus
-
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 : Sun Sep 15 2002 - 22:00:14 EST