Re: [PATCH 11/11] KVM: MMU: improve write flooding detected

On 08/26/2011 06:53 PM, Marcelo Tosatti wrote:

On Fri, Aug 26, 2011 at 11:18:01AM +0800, Xiao Guangrong wrote:

quoted

On 08/25/2011 09:47 PM, Marcelo Tosatti wrote:

quoted

I guess it is OK to be more trigger happy with zapping by ignoring
the accessed bit, clearing the flood counter on page fault.

Yeah, i like this way, is this patch good for you?

Looks fine, can you rerun kernbench?

Sure, i tested the performance of this way, there is the result:

The origin way:
2m56.561
2m50.651
2m51.220
2m52.199
2m48.066

The way of using accessed bit:
2m51.194
2m55.980
2m50.755
2m47.396
2m46.807

The way of ignoring accessed bit:
2m45.547
2m44.551
2m55.840
2m56.333
2m45.534

I think the result is not bad.