2010-04-28 03:57:43

by Xiao Guangrong

[permalink] [raw]
Subject: [PATCH v3 0/10] KVM MMU: allow more shadow pages become asynchronous

Changlog v3:

Those changes all form Avi's suggestion, thanks.

- use smart way to fix the bug in patch 1
- remove duplicates code in patch 5
- check error code and fix forgot release page in patch 9
- sync shadow pages in a batch instead of one by one

And, there is one TODO thing:
Marker shadow page as unsync at create time avoid write-protect,
this idea is from Avi:

|Another interesting case is to create new shadow pages in the unsync state.
|That can help when the guest starts a short lived process: we can avoid write
|protecting its pagetables completely

I'll send the patch out after this patchset applied.

Changlog v2:

- when level is PT_DIRECTORY_LEVEL, the 'offset' should be
'role.quadrant << 8', thanks Avi for point it out

- keep invlpg code in paging_tmpl.h address Avi's suggestion

- split kvm_sync_page() into kvm_sync_page() and kvm_sync_page_transient()
to clarify the code address Avi's suggestion


2010-04-28 10:37:14

by Avi Kivity

[permalink] [raw]
Subject: Re: [PATCH v3 0/10] KVM MMU: allow more shadow pages become asynchronous

On 04/28/2010 06:54 AM, Xiao Guangrong wrote:
> Changlog v3:
>

Reviewed-by: Avi Kivity <[email protected]>

--
error compiling committee.c: too many arguments to function

2010-04-28 17:14:15

by Marcelo Tosatti

[permalink] [raw]
Subject: Re: [PATCH v3 0/10] KVM MMU: allow more shadow pages become asynchronous

On Wed, Apr 28, 2010 at 11:54:31AM +0800, Xiao Guangrong wrote:
> Changlog v3:
>
> Those changes all form Avi's suggestion, thanks.
>
> - use smart way to fix the bug in patch 1
> - remove duplicates code in patch 5
> - check error code and fix forgot release page in patch 9
> - sync shadow pages in a batch instead of one by one
>
> And, there is one TODO thing:
> Marker shadow page as unsync at create time avoid write-protect,
> this idea is from Avi:
>
> |Another interesting case is to create new shadow pages in the unsync state.
> |That can help when the guest starts a short lived process: we can avoid write
> |protecting its pagetables completely
>
> I'll send the patch out after this patchset applied.
>
> Changlog v2:
>
> - when level is PT_DIRECTORY_LEVEL, the 'offset' should be
> 'role.quadrant << 8', thanks Avi for point it out
>
> - keep invlpg code in paging_tmpl.h address Avi's suggestion
>
> - split kvm_sync_page() into kvm_sync_page() and kvm_sync_page_transient()
> to clarify the code address Avi's suggestion

Applied, thanks.