Re: [PATCH v11 28/35] KVM: SEV: Implement gmem hook for initializing private pages

[Binbin Wu <binbin.wu@linux.intel.com>]

On 12/31/2023 1:23 AM, Michael Roth wrote:
> This will handle RMP table updates and direct map changes needed to put
> a page into a private state before mapping it into an SEV-SNP guest.
>
> Signed-off-by: Michael Roth <michael.roth@amd.com>
> ---
>   arch/x86/kvm/Kconfig   |  1 +
>   arch/x86/kvm/svm/sev.c | 98 ++++++++++++++++++++++++++++++++++++++++++
>   arch/x86/kvm/svm/svm.c |  2 +
>   arch/x86/kvm/svm/svm.h |  1 +
>   virt/kvm/guest_memfd.c |  4 +-
>   5 files changed, 104 insertions(+), 2 deletions(-)
>
> diff --git a/arch/x86/kvm/Kconfig b/arch/x86/kvm/Kconfig
> index 4ec53d6d5773..79c002e1bb5c 100644
> --- a/arch/x86/kvm/Kconfig
> +++ b/arch/x86/kvm/Kconfig
> @@ -125,6 +125,7 @@ config KVM_AMD_SEV
>   	depends on KVM_AMD && X86_64
>   	depends on CRYPTO_DEV_SP_PSP && !(KVM_AMD=y && CRYPTO_DEV_CCP_DD=m)
>   	select KVM_GENERIC_PRIVATE_MEM
> +	select HAVE_KVM_GMEM_PREPARE
>   	help
>   	  Provides support for launching Encrypted VMs (SEV) and Encrypted VMs
>   	  with Encrypted State (SEV-ES) on AMD processors.
> diff --git a/arch/x86/kvm/svm/sev.c b/arch/x86/kvm/svm/sev.c
> index b2ac696c436a..91f53f4a6059 100644
> --- a/arch/x86/kvm/svm/sev.c
> +++ b/arch/x86/kvm/svm/sev.c
> @@ -4154,3 +4154,101 @@ void handle_rmp_page_fault(struct kvm_vcpu *vcpu, gpa_t gpa, u64 error_code)
>   out:
>   	put_page(pfn_to_page(pfn));
>   }
> +
> +static bool is_pfn_range_shared(kvm_pfn_t start, kvm_pfn_t end)
> +{
> +	kvm_pfn_t pfn = start;
> +
> +	while (pfn < end) {
> +		int ret, rmp_level;
> +		bool assigned;
> +
> +		ret = snp_lookup_rmpentry(pfn, &assigned, &rmp_level);
> +		if (ret) {
> +			pr_warn_ratelimited("SEV: Failed to retrieve RMP entry: PFN 0x%llx GFN start 0x%llx GFN end 0x%llx RMP level %d error %d\n",
> +					    pfn, start, end, rmp_level, ret);
> +			return false;
> +		}
> +
> +		if (assigned) {
> +			pr_debug("%s: overlap detected, PFN 0x%llx start 0x%llx end 0x%llx RMP level %d\n",
> +				 __func__, pfn, start, end, rmp_level);
> +			return false;
> +		}
> +
> +		pfn++;

rmp_level can be got from snp_lookup_rmpentry().
I think the pfn can be updated according to rmp_level to avoid unnecessary
loops for 2MB large page, right?

> +	}
> +
> +	return true;
> +}
> +
> +static u8 max_level_for_order(int order)
> +{
> +	if (order >= KVM_HPAGE_GFN_SHIFT(PG_LEVEL_2M))
> +		return PG_LEVEL_2M;
> +
> +	return PG_LEVEL_4K;
> +}
> +
> +static bool is_large_rmp_possible(struct kvm *kvm, kvm_pfn_t pfn, int order)
> +{
> +	kvm_pfn_t pfn_aligned = ALIGN_DOWN(pfn, PTRS_PER_PMD);
> +
> +	/*
> +	 * If this is a large folio, and the entire 2M range containing the
> +	 * PFN is currently shared, then the entire 2M-aligned range can be
> +	 * set to private via a single 2M RMP entry.
> +	 */
> +	if (max_level_for_order(order) > PG_LEVEL_4K &&
> +	    is_pfn_range_shared(pfn_aligned, pfn_aligned + PTRS_PER_PMD))
> +		return true;
> +
> +	return false;
> +}
> +
> +int sev_gmem_prepare(struct kvm *kvm, kvm_pfn_t pfn, gfn_t gfn, int max_order)
> +{
> +	struct kvm_sev_info *sev = &to_kvm_svm(kvm)->sev_info;
> +	kvm_pfn_t pfn_aligned;
> +	gfn_t gfn_aligned;
> +	int level, rc;
> +	bool assigned;
> +
> +	if (!sev_snp_guest(kvm))
> +		return 0;
> +
> +	rc = snp_lookup_rmpentry(pfn, &assigned, &level);
> +	if (rc) {
> +		pr_err_ratelimited("SEV: Failed to look up RMP entry: GFN %llx PFN %llx error %d\n",
> +				   gfn, pfn, rc);
> +		return -ENOENT;
> +	}
> +
> +	if (assigned) {
> +		pr_debug("%s: already assigned: gfn %llx pfn %llx max_order %d level %d\n",
> +			 __func__, gfn, pfn, max_order, level);
> +		return 0;
> +	}
> +
> +	if (is_large_rmp_possible(kvm, pfn, max_order)) {
> +		level = PG_LEVEL_2M;
> +		pfn_aligned = ALIGN_DOWN(pfn, PTRS_PER_PMD);
> +		gfn_aligned = ALIGN_DOWN(gfn, PTRS_PER_PMD);
> +	} else {
> +		level = PG_LEVEL_4K;
> +		pfn_aligned = pfn;
> +		gfn_aligned = gfn;
> +	}
> +
> +	rc = rmp_make_private(pfn_aligned, gfn_to_gpa(gfn_aligned), level, sev->asid, false);
> +	if (rc) {
> +		pr_err_ratelimited("SEV: Failed to update RMP entry: GFN %llx PFN %llx level %d error %d\n",
> +				   gfn, pfn, level, rc);
> +		return -EINVAL;
> +	}
> +
> +	pr_debug("%s: updated: gfn %llx pfn %llx pfn_aligned %llx max_order %d level %d\n",
> +		 __func__, gfn, pfn, pfn_aligned, max_order, level);
> +
> +	return 0;
> +}
> diff --git a/arch/x86/kvm/svm/svm.c b/arch/x86/kvm/svm/svm.c
> index 240518f8d6c7..32cef8626b57 100644
> --- a/arch/x86/kvm/svm/svm.c
> +++ b/arch/x86/kvm/svm/svm.c
> @@ -5065,6 +5065,8 @@ static struct kvm_x86_ops svm_x86_ops __initdata = {
>   	.vcpu_deliver_sipi_vector = svm_vcpu_deliver_sipi_vector,
>   	.vcpu_get_apicv_inhibit_reasons = avic_vcpu_get_apicv_inhibit_reasons,
>   	.alloc_apic_backing_page = svm_alloc_apic_backing_page,
> +
> +	.gmem_prepare = sev_gmem_prepare,
>   };
>   
>   /*
> diff --git a/arch/x86/kvm/svm/svm.h b/arch/x86/kvm/svm/svm.h
> index d953ae41c619..9ece9612dbb9 100644
> --- a/arch/x86/kvm/svm/svm.h
> +++ b/arch/x86/kvm/svm/svm.h
> @@ -725,6 +725,7 @@ void sev_es_unmap_ghcb(struct vcpu_svm *svm);
>   struct page *snp_safe_alloc_page(struct kvm_vcpu *vcpu);
>   void handle_rmp_page_fault(struct kvm_vcpu *vcpu, gpa_t gpa, u64 error_code);
>   void sev_snp_init_protected_guest_state(struct kvm_vcpu *vcpu);
> +int sev_gmem_prepare(struct kvm *kvm, kvm_pfn_t pfn, gfn_t gfn, int max_order);
>   
>   /* vmenter.S */
>   
> diff --git a/virt/kvm/guest_memfd.c b/virt/kvm/guest_memfd.c
> index feec0da93d98..ddea45279fef 100644
> --- a/virt/kvm/guest_memfd.c
> +++ b/virt/kvm/guest_memfd.c
> @@ -66,8 +66,8 @@ static int kvm_gmem_prepare_folio(struct inode *inode, pgoff_t index, struct fol
>   		gfn = slot->base_gfn + index - slot->gmem.pgoff;
>   		rc = kvm_arch_gmem_prepare(kvm, gfn, pfn, compound_order(compound_head(page)));
>   		if (rc) {
> -			pr_warn_ratelimited("gmem: Failed to prepare folio for index %lx, error %d.\n",
> -					    index, rc);
> +			pr_warn_ratelimited("gmem: Failed to prepare folio for index %lx GFN %llx PFN %llx error %d.\n",
> +					    index, gfn, pfn, rc);
>   			return rc;
>   		}
>   	}