Re: [PATCH v9 01/41] KVM: guest_memfd: Introduce per-gmem attributes, use to guard user mappings

On 7/29/2026 8:35 AM, Ackerley Tng via B4 Relay wrote:
> From: Sean Christopherson<[email protected]>
> 
> Start plumbing in guest_memfd support for in-place private<=>shared
> conversions by tracking attributes via a maple tree.  KVM currently tracks
> private vs. shared attributes on a per-VM basis, which made sense when a
> guest_memfd_only_ supported private memory, but tracking per-VM simply
> can't work for in-place conversions as the shared/private status of a given
> page needs to be per-gmem_inode, not per-VM.
> 
> Use the filemap invalidation lock to protect the maple tree, as taking the
> lock for read when faulting in memory (for userspace or the guest) isn't
> expected to result in meaningful contention, and using a separate lock
> would add significant complexity (avoiding deadlock is quite difficult).
> 
> Co-developed-by: Vishal Annapurve<[email protected]>
> Signed-off-by: Vishal Annapurve<[email protected]>
> Co-developed-by: Fuad Tabba<[email protected]>
> Signed-off-by: Fuad Tabba<[email protected]>
> Signed-off-by: Sean Christopherson<[email protected]>
> Tested-by: Shivank Garg<[email protected]>
> Co-developed-by: Ackerley Tng<[email protected]>
> Signed-off-by: Ackerley Tng<[email protected]>

Reviewed-by: Xiaoyao Li <[email protected]>

<...>

>   static __always_inline struct gmem_inode *GMEM_I(struct inode *inode)
> @@ -60,9 +68,25 @@ static pgoff_t kvm_gmem_get_index(struct kvm_memory_slot *slot, gfn_t gfn)
>   	return gfn - slot->base_gfn + slot->gmem.pgoff;
>   }
>   
> +static u64 kvm_gmem_interpret_entry(struct inode *inode, void *entry)

Ummm, I'm not sure on the name. It doesn't look like KVM's style?

Maybe, just kvm_gmem_get_attributes()?

> +{
> +	if (WARN_ON_ONCE(!entry)) {
> +		bool initially_shared = GMEM_I(inode)->flags &
> +					GUEST_MEMFD_FLAG_INIT_SHARED;
> +
> +		return initially_shared ? 0 : KVM_MEMORY_ATTRIBUTE_PRIVATE;
> +	}
> +
> +	return xa_to_value(entry);
> +}