From: Isaku Yamahata <isaku.yamahata@linux.intel.com>
To: Paolo Bonzini <pbonzini@redhat.com>
Cc: linux-kernel@vger.kernel.org, kvm@vger.kernel.org,
seanjc@google.com, michael.roth@amd.com,
isaku.yamahata@intel.com, thomas.lendacky@amd.com,
isaku.yamahata@linux.intel.com
Subject: Re: [PATCH 18/21] KVM: x86: Add gmem hook for initializing memory
Date: Wed, 28 Feb 2024 12:29:06 -0800 [thread overview]
Message-ID: <20240228202906.GB10568@ls.amr.corp.intel.com> (raw)
In-Reply-To: <20240227232100.478238-19-pbonzini@redhat.com>
On Tue, Feb 27, 2024 at 06:20:57PM -0500,
Paolo Bonzini <pbonzini@redhat.com> wrote:
> guest_memfd pages are generally expected to be in some arch-defined
> initial state prior to using them for guest memory. For SEV-SNP this
> initial state is 'private', or 'guest-owned', and requires additional
> operations to move these pages into a 'private' state by updating the
> corresponding entries the RMP table.
>
> Allow for an arch-defined hook to handle updates of this sort, and go
> ahead and implement one for x86 so KVM implementations like AMD SVM can
> register a kvm_x86_ops callback to handle these updates for SEV-SNP
> guests.
>
> The preparation callback is always called when allocating/grabbing
> folios via gmem, and it is up to the architecture to keep track of
> whether or not the pages are already in the expected state (e.g. the RMP
> table in the case of SEV-SNP).
>
> In some cases, it is necessary to defer the preparation of the pages to
> handle things like in-place encryption of initial guest memory payloads
> before marking these pages as 'private'/'guest-owned', so also add a
> helper that performs the same function as kvm_gmem_get_pfn(), but allows
> for the preparation callback to be bypassed to allow for pages to be
> accessed beforehand.
>
> Link: https://lore.kernel.org/lkml/ZLqVdvsF11Ddo7Dq@google.com/
> Co-developed-by: Michael Roth <michael.roth@amd.com>
> Signed-off-by: Michael Roth <michael.roth@amd.com>
> Message-Id: <20231230172351.574091-5-michael.roth@amd.com>
> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
> ---
> arch/x86/include/asm/kvm-x86-ops.h | 1 +
> arch/x86/include/asm/kvm_host.h | 1 +
> arch/x86/kvm/x86.c | 6 +++
> include/linux/kvm_host.h | 14 ++++++
> virt/kvm/Kconfig | 4 ++
> virt/kvm/guest_memfd.c | 72 +++++++++++++++++++++++++++---
> 6 files changed, 92 insertions(+), 6 deletions(-)
>
> diff --git a/arch/x86/include/asm/kvm-x86-ops.h b/arch/x86/include/asm/kvm-x86-ops.h
> index ac8b7614e79d..adfaad15e7e6 100644
> --- a/arch/x86/include/asm/kvm-x86-ops.h
> +++ b/arch/x86/include/asm/kvm-x86-ops.h
> @@ -139,6 +139,7 @@ KVM_X86_OP(complete_emulated_msr)
> KVM_X86_OP(vcpu_deliver_sipi_vector)
> KVM_X86_OP_OPTIONAL_RET0(vcpu_get_apicv_inhibit_reasons);
> KVM_X86_OP_OPTIONAL(get_untagged_addr)
> +KVM_X86_OP_OPTIONAL_RET0(gmem_prepare)
>
> #undef KVM_X86_OP
> #undef KVM_X86_OP_OPTIONAL
> diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
> index 7de8a3f2a118..6d873d08f739 100644
> --- a/arch/x86/include/asm/kvm_host.h
> +++ b/arch/x86/include/asm/kvm_host.h
> @@ -1804,6 +1804,7 @@ struct kvm_x86_ops {
> unsigned long (*vcpu_get_apicv_inhibit_reasons)(struct kvm_vcpu *vcpu);
>
> gva_t (*get_untagged_addr)(struct kvm_vcpu *vcpu, gva_t gva, unsigned int flags);
> + int (*gmem_prepare)(struct kvm *kvm, kvm_pfn_t pfn, gfn_t gfn, int max_order);
> };
>
> struct kvm_x86_nested_ops {
> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> index f10a5a617120..eff532ea59c9 100644
> --- a/arch/x86/kvm/x86.c
> +++ b/arch/x86/kvm/x86.c
> @@ -13598,6 +13598,12 @@ bool kvm_arch_no_poll(struct kvm_vcpu *vcpu)
> }
> EXPORT_SYMBOL_GPL(kvm_arch_no_poll);
>
> +#ifdef CONFIG_HAVE_KVM_GMEM_PREPARE
> +int kvm_arch_gmem_prepare(struct kvm *kvm, gfn_t gfn, kvm_pfn_t pfn, int max_order)
> +{
> + return static_call(kvm_x86_gmem_prepare)(kvm, pfn, gfn, max_order);
> +}
> +#endif
>
> int kvm_spec_ctrl_test_value(u64 value)
> {
> diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h
> index 97afe4519772..03bf616b7308 100644
> --- a/include/linux/kvm_host.h
> +++ b/include/linux/kvm_host.h
> @@ -2434,6 +2434,8 @@ static inline bool kvm_mem_is_private(struct kvm *kvm, gfn_t gfn)
> #ifdef CONFIG_KVM_PRIVATE_MEM
> int kvm_gmem_get_pfn(struct kvm *kvm, struct kvm_memory_slot *slot,
> gfn_t gfn, kvm_pfn_t *pfn, int *max_order);
> +int kvm_gmem_get_uninit_pfn(struct kvm *kvm, struct kvm_memory_slot *slot,
> + gfn_t gfn, kvm_pfn_t *pfn, int *max_order);
> #else
> static inline int kvm_gmem_get_pfn(struct kvm *kvm,
> struct kvm_memory_slot *slot, gfn_t gfn,
> @@ -2442,6 +2444,18 @@ static inline int kvm_gmem_get_pfn(struct kvm *kvm,
> KVM_BUG_ON(1, kvm);
> return -EIO;
> }
> +
> +static inline int kvm_gmem_get_uninit_pfn(struct kvm *kvm,
> + struct kvm_memory_slot *slot, gfn_t gfn,
> + kvm_pfn_t *pfn, int *max_order)
> +{
> + KVM_BUG_ON(1, kvm);
> + return -EIO;
> +}
> #endif /* CONFIG_KVM_PRIVATE_MEM */
>
> +#ifdef CONFIG_HAVE_KVM_GMEM_PREPARE
> +int kvm_arch_gmem_prepare(struct kvm *kvm, gfn_t gfn, kvm_pfn_t pfn, int max_order);
> +#endif
> +
> #endif
> diff --git a/virt/kvm/Kconfig b/virt/kvm/Kconfig
> index a11e9c80fac9..dcce0c3b5b13 100644
> --- a/virt/kvm/Kconfig
> +++ b/virt/kvm/Kconfig
> @@ -111,3 +111,7 @@ config KVM_GENERIC_PRIVATE_MEM
> select KVM_GENERIC_MEMORY_ATTRIBUTES
> select KVM_PRIVATE_MEM
> bool
> +
> +config HAVE_KVM_GMEM_PREPARE
> + bool
> + depends on KVM_PRIVATE_MEM
> diff --git a/virt/kvm/guest_memfd.c b/virt/kvm/guest_memfd.c
> index de0d5a5c210c..7ec7afafc960 100644
> --- a/virt/kvm/guest_memfd.c
> +++ b/virt/kvm/guest_memfd.c
> @@ -13,12 +13,50 @@ struct kvm_gmem {
> struct list_head entry;
> };
>
> -static struct folio *kvm_gmem_get_folio(struct inode *inode, pgoff_t index)
> +static int kvm_gmem_prepare_folio(struct inode *inode, pgoff_t index, struct folio *folio)
> +{
> +#ifdef CONFIG_HAVE_KVM_GMEM_PREPARE
> + struct list_head *gmem_list = &inode->i_mapping->i_private_list;
> + struct kvm_gmem *gmem;
> +
> + list_for_each_entry(gmem, gmem_list, entry) {
> + struct kvm_memory_slot *slot;
> + struct kvm *kvm = gmem->kvm;
> + struct page *page;
> + kvm_pfn_t pfn;
> + gfn_t gfn;
> + int rc;
> +
> + slot = xa_load(&gmem->bindings, index);
> + if (!slot)
> + continue;
> +
> + page = folio_file_page(folio, index);
> + pfn = page_to_pfn(page);
> + 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);
> + return rc;
> + }
> + }
> +
> +#endif
> + return 0;
> +}
Can we make it conditional?
TDX doesn't need prepare hook to set gmem_parepare = NULL. With large memory
guest(several hundreds Gbyte) to lookup page cache, this loop slows down guest
startup. I think it would also applies to SW_PROTECTED_VM (and pKVM in future).
diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h
index 3835732491b9..cafb8d0997b5 100644
--- a/include/linux/kvm_host.h
+++ b/include/linux/kvm_host.h
@@ -842,6 +842,9 @@ struct kvm {
#ifdef CONFIG_KVM_GENERIC_MEMORY_ATTRIBUTES
/* Protected by slots_locks (for writes) and RCU (for reads) */
struct xarray mem_attr_array;
+#endif
+#ifdef CONFIG_HAVE_KVM_GMEM_PREPARE
+ bool gmem_need_prepare;
#endif
char stats_id[KVM_STATS_NAME_SIZE];
};
diff --git a/virt/kvm/guest_memfd.c b/virt/kvm/guest_memfd.c
index 74e19170af8a..ab7d0f7d3d38 100644
--- a/virt/kvm/guest_memfd.c
+++ b/virt/kvm/guest_memfd.c
@@ -16,6 +16,7 @@ struct kvm_gmem {
static int kvm_gmem_prepare_folio(struct inode *inode, pgoff_t index, struct folio *folio)
{
#ifdef CONFIG_HAVE_KVM_GMEM_PREPARE
+ rc = kvm_arch_gmem_prepare(inode, index, folio);
struct list_head *gmem_list = &inode->i_mapping->i_private_list;
struct kvm_gmem *gmem;
@@ -27,6 +28,9 @@ static int kvm_gmem_prepare_folio(struct inode *inode, pgoff_t index, struct fol
gfn_t gfn;
int rc;
+ if (!kvm->gmem_need_prepare)
+ continue;
+
slot = xa_load(&gmem->bindings, index);
if (!slot)
continue;
--
Isaku Yamahata <isaku.yamahata@linux.intel.com>
next prev parent reply other threads:[~2024-02-28 20:29 UTC|newest]
Thread overview: 76+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-02-27 23:20 [PATCH 00/21] TDX/SNP part 1 of n, for 6.9 Paolo Bonzini
2024-02-27 23:20 ` [PATCH 01/21] KVM: x86: Split core of hypercall emulation to helper function Paolo Bonzini
2024-02-28 2:09 ` Xiaoyao Li
2024-03-05 6:24 ` Binbin Wu
2024-02-27 23:20 ` [PATCH 02/21] KVM: Allow page-sized MMU caches to be initialized with custom 64-bit values Paolo Bonzini
2024-02-29 13:46 ` Xiaoyao Li
2024-03-05 6:55 ` Binbin Wu
2024-03-26 15:56 ` Binbin Wu
2024-05-13 20:38 ` Isaku Yamahata
2024-05-13 20:51 ` Isaku Yamahata
2024-05-13 20:56 ` Sean Christopherson
2024-02-27 23:20 ` [PATCH 03/21] KVM: x86/mmu: Replace hardcoded value 0 for the initial value for SPTE Paolo Bonzini
2024-02-29 13:50 ` Xiaoyao Li
2024-03-05 7:09 ` Binbin Wu
2024-02-27 23:20 ` [PATCH 04/21] KVM: x86/mmu: Allow non-zero value for non-present SPTE and removed SPTE Paolo Bonzini
2024-02-29 7:00 ` Xu Yilun
2024-02-29 13:55 ` Xiaoyao Li
2024-03-11 23:26 ` Huang, Kai
2024-02-27 23:20 ` [PATCH 05/21] KVM: x86/mmu: Add Suppress VE bit to EPT shadow_mmio_mask/shadow_present_mask Paolo Bonzini
2024-03-01 7:26 ` Xiaoyao Li
2024-03-05 13:17 ` Binbin Wu
2024-02-27 23:20 ` [PATCH 06/21] KVM: x86/mmu: Track shadow MMIO value on a per-VM basis Paolo Bonzini
2024-03-01 7:44 ` Xiaoyao Li
2024-03-05 8:35 ` Binbin Wu
2024-03-12 1:21 ` Huang, Kai
2024-02-27 23:20 ` [PATCH 07/21] KVM: VMX: Introduce test mode related to EPT violation VE Paolo Bonzini
2024-02-28 1:56 ` Sean Christopherson
2024-03-12 1:35 ` Huang, Kai
2024-03-12 16:54 ` Sean Christopherson
2024-03-12 21:03 ` Huang, Kai
2024-02-27 23:20 ` [PATCH 08/21] KVM: VMX: Move out vmx_x86_ops to 'main.c' to dispatch VMX and TDX Paolo Bonzini
2024-02-27 23:20 ` [PATCH 09/21] KVM: VMX: Modify NMI and INTR handlers to take intr_info as function argument Paolo Bonzini
2024-03-04 8:09 ` Xiaoyao Li
2024-03-05 13:42 ` Binbin Wu
2024-03-12 1:43 ` Huang, Kai
2024-02-27 23:20 ` [PATCH 10/21] KVM: SEV: Use a VMSA physical address variable for populating VMCB Paolo Bonzini
2024-02-28 2:00 ` Sean Christopherson
2024-02-28 17:32 ` Paolo Bonzini
2024-02-29 16:02 ` Sean Christopherson
2024-02-27 23:20 ` [PATCH 11/21] KVM: x86/tdp_mmu: Init role member of struct kvm_mmu_page at allocation Paolo Bonzini
2024-03-03 4:47 ` Xu Yilun
2024-03-25 23:32 ` Edgecombe, Rick P
2024-02-27 23:20 ` [PATCH 12/21] KVM: x86/tdp_mmu: Sprinkle __must_check Paolo Bonzini
2024-03-04 8:29 ` Xiaoyao Li
2024-02-27 23:20 ` [PATCH 13/21] KVM: x86/mmu: Pass around full 64-bit error code for KVM page faults Paolo Bonzini
2024-03-04 8:56 ` Xiaoyao Li
2024-03-04 15:39 ` Sean Christopherson
2024-04-05 17:57 ` Paolo Bonzini
2024-02-27 23:20 ` [PATCH 14/21] KVM: x86/mmu: pass error code back to MMU when async pf is ready Paolo Bonzini
2024-02-28 2:03 ` Sean Christopherson
2024-02-28 13:13 ` Paolo Bonzini
2024-02-27 23:20 ` [PATCH 15/21] KVM: x86/mmu: Use PFERR_GUEST_ENC_MASK to indicate fault is private Paolo Bonzini
2024-02-27 23:20 ` [PATCH 16/21] KVM: guest_memfd: pass error up from filemap_grab_folio Paolo Bonzini
2024-03-03 14:41 ` Xu Yilun
2024-02-27 23:20 ` [PATCH 17/21] filemap: add FGP_CREAT_ONLY Paolo Bonzini
2024-02-28 2:14 ` Sean Christopherson
2024-02-28 2:17 ` Yosry Ahmed
2024-02-28 13:15 ` Matthew Wilcox
2024-02-28 13:28 ` Paolo Bonzini
2024-02-28 19:24 ` Matthew Wilcox
2024-02-28 20:17 ` Paolo Bonzini
2024-03-04 2:55 ` Xu Yilun
2024-02-27 23:20 ` [PATCH 18/21] KVM: x86: Add gmem hook for initializing memory Paolo Bonzini
2024-02-28 20:29 ` Isaku Yamahata [this message]
2024-02-27 23:20 ` [PATCH 19/21] KVM: guest_memfd: add API to undo kvm_gmem_get_uninit_pfn Paolo Bonzini
2024-03-04 4:44 ` Xu Yilun
2024-02-27 23:20 ` [PATCH 20/21] KVM: x86: Add gmem hook for invalidating memory Paolo Bonzini
2024-02-27 23:21 ` [PATCH 21/21] KVM: x86: Add gmem hook for determining max NPT mapping level Paolo Bonzini
2024-03-12 0:39 ` Binbin Wu
2024-03-12 0:48 ` Binbin Wu
2024-02-28 1:24 ` [PATCH 00/21] TDX/SNP part 1 of n, for 6.9 Sean Christopherson
2024-02-28 13:29 ` Paolo Bonzini
2024-02-28 16:39 ` Sean Christopherson
2024-02-28 17:20 ` Paolo Bonzini
2024-02-28 18:04 ` Sean Christopherson
2024-02-28 2:11 ` Sean Christopherson
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20240228202906.GB10568@ls.amr.corp.intel.com \
--to=isaku.yamahata@linux.intel.com \
--cc=isaku.yamahata@intel.com \
--cc=kvm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=michael.roth@amd.com \
--cc=pbonzini@redhat.com \
--cc=seanjc@google.com \
--cc=thomas.lendacky@amd.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).