On Wednesday, 3 May 2017 at 4:34 PM, Daniel P. Berrange wrote:

On Wed, May 03, 2017 at 08:30:31AM +0800, Eli Qiao wrote:
hi all,

Thanks for helping reviewing for CAT support in the past days.

I writing this email to ask for the plan in libvirt support.

I think we’v discussed this early this year, and I’v proposed a patch set [1].
But don’t get merged because of some performance reason ?

Then I proposed a redesign RFC[2] based on Martin’s cache branch,
thans Martin for the reviewing, I can address them but the question here
is it depends on Martin’s `cache` branch, which if for exposing host’s `cache`
information in capabilities xml, and it doesn’t get merged ether, I feel helpless.

Martin posted review comments on your latest patches just a couple of days
ago, and you've not posted any newer version of the patches since then
that address those comments.

Yes, I see that and response, I will refine my RFC patch.

I just worry about the dependencies. It’s long time and don’t get reviewed after the last post.
https://www.redhat.com/archives/libvir-list/2017-April/msg00541.html

And Martin’s `cache` branch was not merged yet, that’s not under my control, this
is my concern. 
As CAT is a key feature which would be required by many customers and
especially the OpenStack integration.

Regardless of what/who needs a feature, we're not going to rush to
merge patches if there are still outstanding issues that need fixing.

Agree. 
I would like to get some advice on which direction should go, with the new implementation, 
the dependency is a problem.
Would like to know the plan and get the some suggestions.

Carry on addressing the feedback provided & posting new versions of the
patches for review.
Sure, I will continue working on that. 

BR,
Eli Qiao