On Tue, 18 Sep 2007 19:41:14 +0900 Hisashi Hifumi <[EMAIL PROTECTED]> wrote:
> > > >So I do think that for safety and sanity's sake, we should be taking a ref > >on the pages when they are in a pagevec. That's going to hurt your nice > >performance numbers :( > > > > I did ping test again to observe performance deterioration caused by taking > a ref. > > -2.6.23-rc6-with-modifiedpatch > --- testmachine ping statistics --- > 3000 packets transmitted, 3000 received, 0% packet loss, time 53386ms > rtt min/avg/max/mdev = 0.074/0.110/4.716/0.147 ms, pipe 2, ipg/ewma > 17.801/0.129 ms > > The result for my original patch is as follows. > > -2.6.23-rc5-with-originalpatch > --- testmachine ping statistics --- > 3000 packets transmitted, 3000 received, 0% packet loss, time 51924ms > rtt min/avg/max/mdev = 0.072/0.108/3.884/0.114 ms, pipe 2, ipg/ewma > 17.314/0.091 ms > > > The influence to response was small. well.. that's not really the test which will show up any regressions. The extra get_page/put_page will affect things like kernel CPU utilisation on fast writeout on a single CPU. Say, run a huge write against a fast storage system on a single pentium 4 CPU and see how much the system CPU time is increased. The kernel's internal cpu accounting probably won't be accurate enough to get believeable numbers from a test like this - better to use the subtractive approach: see http://www.zipworld.com.au/~akpm/linux/#zc - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
