Re: [RFC PATCH 3/7] target/i386: SEV: Add support for enabling debug-swap SEV feature

[Markus Armbruster <armbru@redhat.com>]

Naveen N Rao <naveen@kernel.org> writes:

> Hi Markus,
>
> On Fri, Sep 12, 2025 at 01:20:43PM +0200, Markus Armbruster wrote:
>> "Naveen N Rao (AMD)" <naveen@kernel.org> writes:
>> 
>> > Add support for enabling debug-swap VMSA SEV feature in SEV-ES and
>> > SEV-SNP guests through a new "debug-swap" boolean property on SEV guest
>> > objects. Though the boolean property is available for plain SEV guests,
>> > check_sev_features() will reject setting this for plain SEV guests.
>> 
>> Let's see whether I understand...
>> 
>> It's a property of sev-guest and sev-snp-guest objects.  These are the
>> "SEV guest objects".
>> 
>> I guess a sev-snp-guest object implies it's a SEV-SNP guest, and setting
>> @debug-swap on such an object just works.
>> 
>> With a sev-guest object, it's either a "plain SEV guest" or a "SEV-ES"
>> guest.
>> 
>> If it's the latter, setting @debug-swap just works.
>> 
>> If it's the former, and you set @debug-swap to true, then KVM
>> accelerator initialization will fail later on.  This might trigger
>> fallback to TCG.
>> 
>> Am I confused?
>
> You're spot on, except that in the last case above (plain old SEV 
> guest), qemu throws an error:
> 	qemu-system-x86_64: check_sev_features: SEV features require either SEV-ES or SEV-SNP to be enabled

Okay.

Can you (or anyone) explain to me why SEV-SNP gets its own object type,
but SEV-ES does not?

[...]