On Jun 20, 2016, at 13:25, Paul Mackerras wrote:On Mon, May 16, 2016 at 02:58:18PM +0800, Li Zhong wrote:This patch tries to use smaller locks for each irq in the ics, insteadof a lock at the ics level, to provide better scalability.This looks like a worth-while thing to do. Do you have anyperformance measurements to justify the change? This will increasethe size of struct kvmppc_ics by 4kB, so it would be useful to showthe performance increase that justifies it.Actually, I saw some “improvement” because of the vcpus were not binded, io jobs and irqs on the guest were not binded. After I fixed those random factors, the result became stable, but I couldn’t see any obvious improvements from the patches... Maybe I need find some other test cases that could support this change. Also, when you resend the patch, please make the patch descriptionmore definite - say "With this patch, we use" rather than "this patchtries to use", for instance.OK, I will change that when doing a resend, if I can find some workload that could benefit from this change. Thanks, ZhongRegards,Paul.___Linuxppc-dev mailing listLinuxppc-dev@lists.ozlabs.orghttps://lists.ozlabs.org/listinfo/linuxppc-dev
___
Linuxppc-dev mailing list
Linuxppc-dev@lists.ozlabs.org
https://lists.ozlabs.org/listinfo/linuxppc-dev
