Re: [PATCH v5] x86/vmx: add hvm functions to get/set non-register state

[Tamas K Lengyel <tamas@xxxxxxxxxxxxx>]

On Wed, Apr 20, 2022 at 2:50 AM Jan Beulich <jbeulich@xxxxxxxx> wrote:
>
> On 20.04.2022 08:39, Tian, Kevin wrote:
> >> From: Tamas K Lengyel <tamas@xxxxxxxxxxxxx>
> >> Sent: Tuesday, April 19, 2022 2:43 AM
> >>
> >> On Fri, Mar 25, 2022 at 9:34 AM Tamas K Lengyel <tamas.lengyel@xxxxxxxxx>
> >> wrote:
> >>>
> >>> During VM forking and resetting a failed vmentry has been observed due
> >>> to the guest non-register state going out-of-sync with the guest register
> >>> state. For example, a VM fork reset right after a STI instruction can 
> >>> trigger
> >>> the failed entry. This is due to the guest non-register state not being 
> >>> saved
> >>> from the parent VM, thus the reset operation only copies the register 
> >>> state.
> >>>
> >>> Fix this by adding a new pair of hvm functions to get/set the guest
> >>> non-register state so that the overall vCPU state remains in sync.
> >>>
> >>> Signed-off-by: Tamas K Lengyel <tamas.lengyel@xxxxxxxxx>
> >>> ---
> >>> v5: Switch to internal-only hvm funcs instead of adding to hvm_hw_cpu
> >>
> >> Patch ping.
> >
> > I'd like to hear opinions from Andrew/Jan first. Obviously they have
> > different thoughts when reviewing the earlier versions.
>
> Well, I certainly would prefer if migration was taken care of at the same
> time, but I can understand if Tamas doesn't want to put more time into
> getting that case working. Plus, aiui, this solution to his problem won't
> prevent the eventual wider scope change to be used also for the specific
> purpose here, perhaps by simply fully replacing what is being done now.

Can we move forward with this patch then? As Jan points out, it
doesn't prevent anyone coming up with a fix to the migration case,
whatever shape that might take.

Thanks,
Tamas