Re: [PATCH v1 11/11] KVM: arm64: Implement lazy vCPU state sync for non-protected guests

[Marc Zyngier <maz@kernel.org>]

On Fri, 12 Jun 2026 07:59:25 +0100,
tabba@google.com wrote:
> 
> pKVM copies a non-protected guest's register context between the host
> and the hypervisor on every world switch, even when the host never
> inspects it. Defer the copy: on entry, flush the host context into the
> hyp vCPU only when the host marked it dirty (PKVM_HOST_STATE_DIRTY); on
> exit, leave it in the hyp vCPU and copy it back only when the host needs
> it, via a __pkvm_vcpu_sync_state hypercall on trap handling or at vcpu
> put. A protected guest's context is copied as before, since lazy sync
> only helps where the host is trusted to see the guest's registers.
> 
> The PC is the exception: it is copied back on every exit so the
> kvm_exit tracepoint reports the guest's real exit PC rather than the
> value left by the previous sync.
> 
> Signed-off-by: Fuad Tabba <tabba@google.com>
> ---
>  arch/arm64/include/asm/kvm_asm.h   |  1 +
>  arch/arm64/include/asm/kvm_host.h  |  2 +
>  arch/arm64/kvm/arm.c               |  7 +++
>  arch/arm64/kvm/handle_exit.c       | 22 ++++++++
>  arch/arm64/kvm/hyp/nvhe/hyp-main.c | 88 ++++++++++++++++++++++++++++--
>  5 files changed, 115 insertions(+), 5 deletions(-)
> 
> diff --git a/arch/arm64/include/asm/kvm_asm.h b/arch/arm64/include/asm/kvm_asm.h
> index 043495f7fc78..6e1135b3ded4 100644
> --- a/arch/arm64/include/asm/kvm_asm.h
> +++ b/arch/arm64/include/asm/kvm_asm.h
> @@ -113,6 +113,7 @@ enum __kvm_host_smccc_func {
>  	__KVM_HOST_SMCCC_FUNC___pkvm_finalize_teardown_vm,
>  	__KVM_HOST_SMCCC_FUNC___pkvm_vcpu_load,
>  	__KVM_HOST_SMCCC_FUNC___pkvm_vcpu_put,
> +	__KVM_HOST_SMCCC_FUNC___pkvm_vcpu_sync_state,
>  	__KVM_HOST_SMCCC_FUNC___pkvm_tlb_flush_vmid,
>  
>  	MARKER(__KVM_HOST_SMCCC_FUNC_MAX)
> diff --git a/arch/arm64/include/asm/kvm_host.h b/arch/arm64/include/asm/kvm_host.h
> index a49042bfa801..1ef660774adc 100644
> --- a/arch/arm64/include/asm/kvm_host.h
> +++ b/arch/arm64/include/asm/kvm_host.h
> @@ -1113,6 +1113,8 @@ struct kvm_vcpu_arch {
>  /* SError pending for nested guest */
>  #define NESTED_SERROR_PENDING	__vcpu_single_flag(sflags, BIT(8))
>  
> +/* pKVM host vcpu state is dirty, needs resync (nVHE-only) */
> +#define PKVM_HOST_STATE_DIRTY	__vcpu_single_flag(iflags, BIT(4))

Please don't add flags in random positions. Keep them grouped by
type.

Also, it is unclear to me whether this is really an input to the
hypervisor (which is what iflags indicates), but rather a new type of
flag that indicate some hypervisor-private state. May not be that
important right now, but worth keeping in mind.

	M.

-- 
Without deviation from the norm, progress is not possible.