On 1/5/23 11:08, Richard W.M. Jones wrote:
On Sun, Jan 01, 2023 at 03:09:26PM +0100, Laszlo Ersek wrote:
> Reflect commit 38b35f3b7e5c ("rhv: Use osinfo to distinguish Windows >= 10
> variants", 2022-12-02) to the "ovirt:id" field in the OVF as well.
Take
> the values from "packaging/conf/osinfo-defaults.properties" in the
> ovirt-engine tree, as the comment on "get_ovirt_osid" explains.
>
> Bugzilla:
https://bugzilla.redhat.com/show_bug.cgi?id=2152465
> Signed-off-by: Laszlo Ersek <lersek(a)redhat.com>
> ---
> lib/create_ovf.ml | 21 +++++++++++++++------
> 1 file changed, 15 insertions(+), 6 deletions(-)
>
> diff --git a/lib/create_ovf.ml b/lib/create_ovf.ml
> index 8aff3d8f0b53..79b3285766c9 100644
> --- a/lib/create_ovf.ml
> +++ b/lib/create_ovf.ml
> @@ -437,13 +437,22 @@ and get_ovirt_osid = function
> i_arch = "i386" } ->
> 26
>
> + (* For Windows NT 10.0 always use the <osinfo> field since the
> + * other fields will not accurately reflect the version.
> + *)
> | { i_type = "windows"; i_major_version = 10; i_minor_version = 0;
> - i_arch = "x86_64"; i_product_variant = "Client" } ->
> - 27
> -
> - | { i_type = "windows"; i_major_version = 10; i_minor_version = 0;
> - i_arch = "x86_64" } ->
> - 29
> + i_arch = "x86_64"; i_osinfo = osinfo; i_product_name = product }
->
> + (match osinfo with
> + | "win10" -> (* windows_10x64 *) 27
> + | "win11" -> (* windows_11 *) 36
> + | "win2k16" -> (* windows_2016x64 *) 29
> + | "win2k19" -> (* windows_2019x64 *) 31
> + | "win2k22" -> (* windows_2022 *) 37
> + | _ ->
> + warning (f_"unknown Windows 10 variant: %s (%s)")
> + osinfo product;
> + (* windows_2022 *) 37
> + )
>
> | { i_type = typ; i_distro = distro;
> i_major_version = major; i_minor_version = minor; i_arch = arch;
ACK
... although you might extend the outer match so it matches on
the extra field.
I'll stick with it as posted, for two reasons:
- this change mirrors your change in 38b35f3b7e5c verbatim,
- Vera Wu tested this version, which took a few days (personally I had
no means to do it at all), and now Virt-Tools QE are in a hurry to get
the rebase, and we should include this in the rebase.
So, I've pushed it as commit fcd787573bc7.
I'm going to try and write up the release notes for the new
stable
branch of virt-v2v this week, with the aim of releasing it on Monday
next week. And I'll attempt to review everything outstanding today.
Thanks!
Laszlo