On Fri, May 02, 2025 at 04:08:24PM +0300, Kirill A. Shutemov wrote:
> The functions kvm_x86_ops::link_external_spt() and
> kvm_x86_ops::set_external_spte() are used to assign new memory to a VM.
> When using TDX with Dynamic PAMT enabled, the assigned memory must be
> covered by PAMT.
>
> The new function kvm_x86_ops::phys_prepare() is called before
> link_external_spt() and set_external_spte() to ensure that the memory is
> ready to be assigned to the virtual machine. In the case of TDX, it
> makes sure that the memory is covered by PAMT.
>
> kvm_x86_ops::phys_prepare() is called in a context where struct kvm_vcpu
> is available, allowing the implementation to allocate memory from a
> per-VCPU pool.
>
Why not invoke phys_prepare() and phys_cleanup() in set_external_spte_present()?
Or in tdx_sept_set_private_spte()/tdx_sept_link_private_spt()?
> The function kvm_x86_ops::phys_cleanup() frees PAMT memory in case of
> failure.
>
> Signed-off-by: Kirill A. Shutemov <kirill.shutemov@xxxxxxxxxxxxxxx>
> ---
> arch/x86/include/asm/kvm-x86-ops.h | 2 ++
> arch/x86/include/asm/kvm_host.h | 3 ++
> arch/x86/kvm/mmu/tdp_mmu.c | 47 +++++++++++++++++++++++++++---
> 3 files changed, 48 insertions(+), 4 deletions(-)
>
> diff --git a/arch/x86/include/asm/kvm-x86-ops.h b/arch/x86/include/asm/kvm-x86-ops.h
> index 79406bf07a1c..37081d04e82f 100644
> --- a/arch/x86/include/asm/kvm-x86-ops.h
> +++ b/arch/x86/include/asm/kvm-x86-ops.h
> @@ -99,6 +99,8 @@ KVM_X86_OP_OPTIONAL(link_external_spt)
> KVM_X86_OP_OPTIONAL(set_external_spte)
> KVM_X86_OP_OPTIONAL(free_external_spt)
> KVM_X86_OP_OPTIONAL(remove_external_spte)
> +KVM_X86_OP_OPTIONAL(phys_prepare)
> +KVM_X86_OP_OPTIONAL(phys_cleanup)
> KVM_X86_OP(has_wbinvd_exit)
> KVM_X86_OP(get_l2_tsc_offset)
> KVM_X86_OP(get_l2_tsc_multiplier)
> diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
> index 6c06f3d6e081..91958c55f918 100644
> --- a/arch/x86/include/asm/kvm_host.h
> +++ b/arch/x86/include/asm/kvm_host.h
> @@ -1813,6 +1813,9 @@ struct kvm_x86_ops {
> int (*remove_external_spte)(struct kvm *kvm, gfn_t gfn, enum pg_level level,
> kvm_pfn_t pfn_for_gfn);
>
> + int (*phys_prepare)(struct kvm_vcpu *vcpu, kvm_pfn_t pfn);
> + void (*phys_cleanup)(kvm_pfn_t pfn);
> +
> bool (*has_wbinvd_exit)(void);
>
> u64 (*get_l2_tsc_offset)(struct kvm_vcpu *vcpu);
> diff --git a/arch/x86/kvm/mmu/tdp_mmu.c b/arch/x86/kvm/mmu/tdp_mmu.c
> index 405874f4d088..f6c836b2e6fc 100644
> --- a/arch/x86/kvm/mmu/tdp_mmu.c
> +++ b/arch/x86/kvm/mmu/tdp_mmu.c
> @@ -1137,6 +1137,26 @@ void kvm_tdp_mmu_invalidate_roots(struct kvm *kvm,
> }
> }
>
> +static int tdp_mmu_install_spte(struct kvm_vcpu *vcpu,
> + struct tdp_iter *iter,
> + u64 spte)
> +{
> + kvm_pfn_t pfn = 0;
> + int ret = 0;
> +
> + if (is_mirror_sptep(iter->sptep) && !is_frozen_spte(spte)) {
> + pfn = spte_to_pfn(spte);
> + ret = static_call(kvm_x86_phys_prepare)(vcpu, pfn);
> + }
> + if (ret)
> + return ret;
> + ret = tdp_mmu_set_spte_atomic(vcpu->kvm, iter, spte);
> + if (pfn && ret)
> + static_call(kvm_x86_phys_cleanup)(pfn);
> +
> + return ret;
> +}
> +
> /*
> * Installs a last-level SPTE to handle a TDP page fault.
> * (NPT/EPT violation/misconfiguration)
> @@ -1170,7 +1190,7 @@ static int tdp_mmu_map_handle_target_level(struct kvm_vcpu *vcpu,
>
> if (new_spte == iter->old_spte)
> ret = RET_PF_SPURIOUS;
> - else if (tdp_mmu_set_spte_atomic(vcpu->kvm, iter, new_spte))
> + else if (tdp_mmu_install_spte(vcpu, iter, new_spte))
> return RET_PF_RETRY;
> else if (is_shadow_present_pte(iter->old_spte) &&
> (!is_last_spte(iter->old_spte, iter->level) ||
> @@ -1211,7 +1231,7 @@ static int tdp_mmu_map_handle_target_level(struct kvm_vcpu *vcpu,
> * Returns: 0 if the new page table was installed. Non-0 if the page table
> * could not be installed (e.g. the atomic compare-exchange failed).
> */
> -static int tdp_mmu_link_sp(struct kvm *kvm, struct tdp_iter *iter,
> +static int __tdp_mmu_link_sp(struct kvm *kvm, struct tdp_iter *iter,
> struct kvm_mmu_page *sp, bool shared)
> {
> u64 spte = make_nonleaf_spte(sp->spt, !kvm_ad_enabled);
> @@ -1230,6 +1250,25 @@ static int tdp_mmu_link_sp(struct kvm *kvm, struct tdp_iter *iter,
> return 0;
> }
>
> +static int tdp_mmu_link_sp(struct kvm_vcpu *vcpu, struct tdp_iter *iter,
> + struct kvm_mmu_page *sp, bool shared)
> +{
> + kvm_pfn_t pfn = 0;
> + int ret = 0;
> +
> + if (sp->external_spt) {
> + pfn = __pa(sp->external_spt) >> PAGE_SHIFT;
> + ret = static_call(kvm_x86_phys_prepare)(vcpu, pfn);
> + if (ret)
> + return ret;
> + }
> + ret = __tdp_mmu_link_sp(vcpu->kvm, iter, sp, shared);
> + if (pfn && ret)
> + static_call(kvm_x86_phys_cleanup)(pfn);
> +
> + return ret;
> +}
> +
> static int tdp_mmu_split_huge_page(struct kvm *kvm, struct tdp_iter *iter,
> struct kvm_mmu_page *sp, bool shared);
>
> @@ -1288,7 +1327,7 @@ int kvm_tdp_mmu_map(struct kvm_vcpu *vcpu, struct kvm_page_fault *fault)
> KVM_BUG_ON(is_mirror_sptep(iter.sptep), vcpu->kvm);
> r = tdp_mmu_split_huge_page(kvm, &iter, sp, true);
> } else {
> - r = tdp_mmu_link_sp(kvm, &iter, sp, true);
> + r = tdp_mmu_link_sp(vcpu, &iter, sp, true);
> }
>
> /*
> @@ -1514,7 +1553,7 @@ static int tdp_mmu_split_huge_page(struct kvm *kvm, struct tdp_iter *iter,
> * correctness standpoint since the translation will be the same either
> * way.
> */
> - ret = tdp_mmu_link_sp(kvm, iter, sp, shared);
> + ret = __tdp_mmu_link_sp(kvm, iter, sp, shared);
> if (ret)
> goto out;
>
> --
> 2.47.2
>
