> "Brian F. G. Bidulock" <[EMAIL PROTECTED]> wrote: > >I think you are probably testing the latency of the round robin scheduler >rather than the performance of LiS.
Brian: I'm not sure about this. >From my prospective I measure my driver performance compiled/linked/run with different versions of LiS and HPUX streams environments. An outside box sends messages to my LiS-based driver(which just bounces them back) and outside box measures response time then. As I said, CPU load of machine running LiS/driver decreased significantly on LiS-2.16.18 vs LiS-2.16.16 . But response time measured by the outside box significantly increased. That's not a good thing from my prospective. I want my driver respond fast and use minimum CPU power for that. At this point I feel it's 50/50 chances of problem being in my driver or in LiS. I looked into Configure and see the following vars being set for production mode: CONFIG_DEV=n USE_KMEM_CACHE=y USE_LINUX_KMEM_TIMER=y I'm thinking of flipping them one by one and re-running my tests, in hope that that would expose something. I would appreciate any other advices on how to approach this. thanks, -- Eugene __________________________________________________________________ New! Unlimited Netscape Internet Service. Only $9.95 a month -- Sign up today at http://isp.netscape.com/register Act now to get a personalized email address! Netscape. Just the Net You Need. _______________________________________________ Linux-streams mailing list [EMAIL PROTECTED] http://gsyc.escet.urjc.es/mailman/listinfo/linux-streams
