Re: [PATCH 0/2] consistency between extensions
On Fri, Nov 10, 2017 at 09:58:16AM -0600, Eric Blake wrote:
> Since Vladimir is working on adding an implementation of BLOCK_STATUS
> support to qemu, I want to make sure the spec extensions make sense
> together, and noticed some minor differences between the various
> branches. This series is the result.
>
> Patch 1 is for the master branch, patch 2 is for the
> extension-structured-reply branch. Then there is an obvious followup
> to the extension-resize branch (use bit 9 instead of 10 for
> NBD_FLAG_SEND_RESIZE), and to the extension-block-status branch
> (use 3 instead of 5 as the value of NBD_REPLY_TYPE_BLOCK_STATUS).
>
> Qemu 2.11 is about a month away from release, and it now includes
> the STRUCTURED_REPLY extension, so it is time to start thinking about
> promoting that branch from experimental to mainline. I'm hoping to
> also have time in the next month to implement STRUCTURED_REPLY on
> the nbdkit implementation, and prove that it interacts with qemu's
> implementation, as an extra reason to call the extension stable.
> RESIZE and BLOCK_STATUS will not be implemented any sooner than
> qemu 2.12, so those two branches still remain experimental.
>
> Eric Blake (2):
> doc: Update reserved values for pending extensions
> doc: Update reservation for NBD_REPLY_TYPE_ bits
These look good, thanks.
--
Could you people please use IRC like normal people?!?
-- Amaya Rodrigo Sastre, trying to quiet down the buzz in the DebConf 2008
Hacklab
Reply to: