[PATCH 3/5] KVM: guest_memfd: bind backing memory to a NUMA node at creation

[Gregory Price <gourry@gourry.net>]

guest_memfd presently allocates its page-cache folios through a
per-inode shared mempolicy (kvm_gmem_get_folio()).

Today that policy can only be set after the fact, via mbind() on
a host mmap of the fd. This requires the fd to be mmap-able and
cannot reach folios that are only ever guest-faulted (no host VMA).

Neither holds for a non-mappable (confidential) guest_memfd.

Add GUEST_MEMFD_FLAG_BIND_NODE.

When set, KVM builds an MPOL_BIND policy for the requested node and
installs it over the whole inode's shared policy, so every folio
is allocated on the requested node with no userspace mbind().

The flag is advertised through KVM_CAP_GUEST_MEMFD_FLAGS only when
CONFIG_NUMA is enabled.

Two user-visible behaviors worth noting:

- mempolicy_create() constrains the request against the calling task's
  cpuset. Requesting a node outside the cpuset mems_allowed results
  in the ioctl failing with -EINVAL. Binding is essentially subject
  to the same cpuset constraint as mbind().

  This behavior is correct - a task cannot grant a guest_memfd access
  to a node it cannot access itself.

- The policy hangs off the inode, and nothing rebinds an inode's
  shared policy on a later cpuset change.

    mpol_rebind_task() walks tsk->mempolicy
    mpol_rebind_mm() walks vma->vm_policy

  Neither walker reaches a struct shared_policy.

  The binding is therefore fixed for the life of the fd. That matches
  shmem, whose inode policy behaves the same way, and is the intent
  here: the node is a property of the guest's backing memory, not of
  whoever happens to hold the fd.

Suggested-by: Dave Jiang <dave.jiang@intel.com>
Co-developed-by: Dave Jiang <dave.jiang@intel.com>
Signed-off-by: Dave Jiang <dave.jiang@intel.com>
Signed-off-by: Gregory Price <gourry@gourry.net>
Assisted-by: Claude:claude-opus-4-8
---
 include/linux/kvm_host.h |  3 +++
 include/uapi/linux/kvm.h |  5 ++++-
 virt/kvm/guest_memfd.c   | 44 ++++++++++++++++++++++++++++++++++++++--
 3 files changed, 49 insertions(+), 3 deletions(-)

diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h
index 03bfc92864b6e..738e276633c1e 100644
--- a/include/linux/kvm_host.h
+++ b/include/linux/kvm_host.h
@@ -739,6 +739,9 @@ static inline u64 kvm_gmem_get_supported_flags(struct kvm *kvm)
 	if (!kvm || kvm_arch_supports_gmem_init_shared(kvm))
 		flags |= GUEST_MEMFD_FLAG_INIT_SHARED;
 
+	if (IS_ENABLED(CONFIG_NUMA))
+		flags |= GUEST_MEMFD_FLAG_BIND_NODE;
+
 	return flags;
 }
 #endif
diff --git a/include/uapi/linux/kvm.h b/include/uapi/linux/kvm.h
index ac2d77d149635..8d3ae7e2ead8e 100644
--- a/include/uapi/linux/kvm.h
+++ b/include/uapi/linux/kvm.h
@@ -1658,11 +1658,14 @@ struct kvm_memory_attributes {
 #define KVM_CREATE_GUEST_MEMFD	_IOWR(KVMIO,  0xd4, struct kvm_create_guest_memfd)
 #define GUEST_MEMFD_FLAG_MMAP		(1ULL << 0)
 #define GUEST_MEMFD_FLAG_INIT_SHARED	(1ULL << 1)
+#define GUEST_MEMFD_FLAG_BIND_NODE	(1ULL << 2)
 
 struct kvm_create_guest_memfd {
 	__u64 size;
 	__u64 flags;
-	__u64 reserved[6];
+	__u32 node;
+	__u32 pad;
+	__u64 reserved[5];
 };
 
 #define KVM_PRE_FAULT_MEMORY	_IOWR(KVMIO, 0xd5, struct kvm_pre_fault_memory)
diff --git a/virt/kvm/guest_memfd.c b/virt/kvm/guest_memfd.c
index 625e62e1a0318..dc9f071dd969b 100644
--- a/virt/kvm/guest_memfd.c
+++ b/virt/kvm/guest_memfd.c
@@ -423,6 +423,31 @@ static struct mempolicy *kvm_gmem_get_policy(struct vm_area_struct *vma,
 	 */
 	return mpol_shared_policy_lookup(&GMEM_I(inode)->policy, pgoff);
 }
+
+static int kvm_gmem_bind_node(struct inode *inode, int node)
+{
+	struct mempolicy *pol;
+	nodemask_t nodes;
+	int err;
+
+	if ((unsigned int)node >= MAX_NUMNODES)
+		return -EINVAL;
+
+	init_nodemask_of_node(&nodes, node);
+	pol = mempolicy_create(MPOL_BIND, 0, &nodes);
+	if (IS_ERR(pol))
+		return PTR_ERR(pol);
+
+	err = mpol_set_shared_policy_range(&GMEM_I(inode)->policy, 0,
+					  MAX_LFS_FILESIZE >> PAGE_SHIFT, pol);
+	mpol_put(pol);
+	return err;
+}
+#else
+static int kvm_gmem_bind_node(struct inode *inode, int node)
+{
+	return -EINVAL;
+}
 #endif /* CONFIG_NUMA */
 
 static const struct vm_operations_struct kvm_gmem_vm_ops = {
@@ -520,7 +545,7 @@ bool __weak kvm_arch_supports_gmem_init_shared(struct kvm *kvm)
 	return true;
 }
 
-static int __kvm_gmem_create(struct kvm *kvm, loff_t size, u64 flags)
+static int __kvm_gmem_create(struct kvm *kvm, loff_t size, u64 flags, int node)
 {
 	static const char *name = "[kvm-gmem]";
 	struct gmem_file *f;
@@ -561,6 +586,12 @@ static int __kvm_gmem_create(struct kvm *kvm, loff_t size, u64 flags)
 
 	GMEM_I(inode)->flags = flags;
 
+	if (flags & GUEST_MEMFD_FLAG_BIND_NODE) {
+		err = kvm_gmem_bind_node(inode, node);
+		if (err)
+			goto err_inode;
+	}
+
 	file = alloc_file_pseudo(inode, kvm_gmem_mnt, name, O_RDWR, &kvm_gmem_fops);
 	if (IS_ERR(file)) {
 		err = PTR_ERR(file);
@@ -593,6 +624,7 @@ int kvm_gmem_create(struct kvm *kvm, struct kvm_create_guest_memfd *args)
 {
 	loff_t size = args->size;
 	u64 flags = args->flags;
+	int node = NUMA_NO_NODE;
 
 	if (flags & ~kvm_gmem_get_supported_flags(kvm))
 		return -EINVAL;
@@ -600,7 +632,15 @@ int kvm_gmem_create(struct kvm *kvm, struct kvm_create_guest_memfd *args)
 	if (size <= 0 || !PAGE_ALIGNED(size))
 		return -EINVAL;
 
-	return __kvm_gmem_create(kvm, size, flags);
+	if (flags & GUEST_MEMFD_FLAG_BIND_NODE) {
+		if (args->pad || args->node >= MAX_NUMNODES)
+			return -EINVAL;
+		node = args->node;
+	} else if (args->node || args->pad) {
+		return -EINVAL;
+	}
+
+	return __kvm_gmem_create(kvm, size, flags, node);
 }
 
 int kvm_gmem_bind(struct kvm *kvm, struct kvm_memory_slot *slot,
-- 
2.53.0-Meta