Roman Bogorodskiy wrote:
Roman Bogorodskiy wrote:
> This series reworks SATA address allocation in the bhyve driver.
>
> While commit messages provide enough details (I hope), there
> are some general important notes:
>
> - currently, sata devices get PCI addresses and this no
> longer works, so right now bhyve driver fails on any
> SATA device
> - While this series fixes SATA devices' addresses (I hope *again*),
> old Domain XMLs with already generated PCI addresses will
> not work. Also, it will not work for XMLs where user
> manually specified PCI address for a disk, though it
> worked before. This is not good and I'm open for suggestions
> how to handle that. I'm thinking about writing a tiny
> Python script that will drop incorrect addresses from
> domain XMLs.
>
>
> Fabian Freyer (1):
> bhyve: detect 32 SATA devices per controller support
>
> Roman Bogorodskiy (3):
> bhyve: add virBhyveDriverCreateXMLConf
> bhyve: fix SATA address allocation
> bhyve: add tests for SATA address allocation
Ping?
While it doesn't seem likely that there's a chance to get that in the
upcoming release because even if this is fine as it as, I still need to
properly test at least basic domain XML migration across releases, but
it'd be cool to have this landed in the beginning of the next release
cycle.
Ping?
Roman Bogorodskiy