On Mon, Feb 26, 2024 at 12:25:12AM -0800, isaku.yamahata@xxxxxxxxx wrote:
> From: Isaku Yamahata <isaku.yamahata@xxxxxxxxx>
>
> TDX wants to know the faulting address is shared or private so that the max
> level is limited by Secure-EPT or not. Because fault->gfn doesn't include
> shared bit, gfn doesn't tell if the faulting address is shared or not.
> Pass is_private for TDX case.
>
> TDX logic will be if (!is_private) return 0; else return PG_LEVEL_4K.
>
> Signed-off-by: Isaku Yamahata <isaku.yamahata@xxxxxxxxx>
> ---
> arch/x86/include/asm/kvm_host.h | 3 ++-
> arch/x86/kvm/mmu/mmu.c | 3 ++-
> 2 files changed, 4 insertions(+), 2 deletions(-)
>
> diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
> index d15f5b4b1656..57ce89fc2740 100644
> --- a/arch/x86/include/asm/kvm_host.h
> +++ b/arch/x86/include/asm/kvm_host.h
> @@ -1797,7 +1797,8 @@ struct kvm_x86_ops {
>
> gva_t (*get_untagged_addr)(struct kvm_vcpu *vcpu, gva_t gva, unsigned int flags);
>
> - int (*gmem_max_level)(struct kvm *kvm, kvm_pfn_t pfn, gfn_t gfn, u8 *max_level);
> + int (*gmem_max_level)(struct kvm *kvm, kvm_pfn_t pfn, gfn_t gfn,
> + bool is_private, u8 *max_level);
> };
>
> struct kvm_x86_nested_ops {
> diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c
> index 1e5e12d2707d..22db1a9f528a 100644
> --- a/arch/x86/kvm/mmu/mmu.c
> +++ b/arch/x86/kvm/mmu/mmu.c
> @@ -4324,7 +4324,8 @@ static int kvm_faultin_pfn_private(struct kvm_vcpu *vcpu,
>
> max_level = kvm_max_level_for_order(max_order);
> r = static_call(kvm_x86_gmem_max_level)(vcpu->kvm, fault->pfn,
> - fault->gfn, &max_level);
> + fault->gfn, fault->is_private,
> + &max_level);
fault->is_private is always true in kvm_faultin_pfn_private().
Besides, as shared page allocation will not go to kvm_faultin_pfn_private(),
why do we need to add the "is_private" parameter ?
> if (r) {
> kvm_release_pfn_clean(fault->pfn);
> return r;
> --
> 2.25.1
>
>
