Messages in this thread | | | Date | Mon, 28 Jun 1999 15:43:21 -0400 (EDT) | From | Chuck Lever <> | Subject | Re: Perforamnce comparison between 2.3.8 and 2.2.10 |
| |
On Sun, 27 Jun 1999, Rudolf Leitgeb wrote: > On Sun, 27 Jun 1999, you wrote: > > Ehh.. Wasn't the main performance gain for the multi-processor systems? I > > wasn't really awake when I read the thread, but that was something that > > I remembered. > > Well, I run SMP and got the same results that Marc got. Evidently none > of us was able to verify the supposedly massinve speed improvements > with the 2.3.8 kernel. I personally rather find it out here on this > mailing list than learn about it from Mindcraft ;-) > > One interesting point though: When I did the kernel compile tests under > 2.3.8, I would not only get higher sys times, but also higher user > times, which makes me wonder ...
i've also noticed that 2.3.8 is slower. but i think the problem is that swapping hasn't been tuned properly for the new page cache logic -- this results in more iterations through shrink_mmap(), where cached state gets dropped. this will slow things down.
i also believe that a multithreaded page cache for file data won't help much if the file system is still serialized :)
- Chuck Lever -- corporate: <chuckl@netscape.com> personal: <chucklever@netscape.net> or <cel@monkey.org>
The Linux Scalability project: http://www.citi.umich.edu/projects/linux-scalability/
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |