Re: [Qemu-devel] [PATCH 2/8] rdma: core rdma logic

[Michael R. Hines]

On 04/12/2013 09:41 AM, Paolo Bonzini wrote:
> Il 12/04/2013 15:29, Michael R. Hines ha scritto:
> > Oh, thank you. Huge bug  - I need to introduce a "COMPRESS"
> > message for this page into the protocol just like we already have
> > for TCP and bypass the current chunk.
> It's okay to send the page uncompressed if it's been registered already.
>   This way you skip most is_zero_page checks after the bulk phase.  I
> just care that a page is never registered if it is zero and never changes.
>
> Paolo

Acknowledged.