Re: [RFC v2 PATCH 06/13] KVM: Register/unregister memfd backed memslot

[Steven Price]

On 19/11/2021 13:47, Chao Peng wrote:
> Signed-off-by: Yu Zhang <yu.c.zhang@linux.intel.com>
> Signed-off-by: Chao Peng <chao.p.peng@linux.intel.com>
> ---
>  virt/kvm/kvm_main.c | 23 +++++++++++++++++++----
>  1 file changed, 19 insertions(+), 4 deletions(-)
> 
> diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> index 271cef8d1cd0..b8673490d301 100644
> --- a/virt/kvm/kvm_main.c
> +++ b/virt/kvm/kvm_main.c
> @@ -1426,7 +1426,7 @@ static void update_memslots(struct kvm_memslots *slots,
>  static int check_memory_region_flags(struct kvm *kvm,
>                            const struct kvm_userspace_memory_region_ext *mem)
>  {
> -     u32 valid_flags = 0;
> +     u32 valid_flags = KVM_MEM_FD;
>  
>       if (!kvm->dirty_log_unsupported)
>               valid_flags |= KVM_MEM_LOG_DIRTY_PAGES;
> @@ -1604,10 +1604,20 @@ static int kvm_set_memslot(struct kvm *kvm,
>               kvm_copy_memslots(slots, __kvm_memslots(kvm, as_id));
>       }
>  
> +     if (mem->flags & KVM_MEM_FD && change == KVM_MR_CREATE) {
> +             r = kvm_memfd_register(kvm, mem, new);
> +             if (r)
> +                     goto out_slots;
> +     }
> +
>       r = kvm_arch_prepare_memory_region(kvm, new, mem, change);
>       if (r)
>               goto out_slots;
>  
> +     if (mem->flags & KVM_MEM_FD && (r || change == KVM_MR_DELETE)) {
                                        ^
r will never be non-zero as the 'if' above will catch that case and jump
to out_slots.

I *think* the intention was that the "if (r)" code should be after this
check to clean up in the case of error from
kvm_arch_prepare_memory_region() (as well as an explicit MR_DELETE).

Steve

> +             kvm_memfd_unregister(kvm, new);
> +     }
> +
>       update_memslots(slots, new, change);
>       slots = install_new_memslots(kvm, as_id, slots);
>  
> @@ -1683,10 +1693,12 @@ int __kvm_set_memory_region(struct kvm *kvm,
>               return -EINVAL;
>       if (mem->guest_phys_addr & (PAGE_SIZE - 1))
>               return -EINVAL;
> -     /* We can read the guest memory with __xxx_user() later on. */
>       if ((mem->userspace_addr & (PAGE_SIZE - 1)) ||
> -         (mem->userspace_addr != untagged_addr(mem->userspace_addr)) ||
> -          !access_ok((void __user *)(unsigned long)mem->userspace_addr,
> +         (mem->userspace_addr != untagged_addr(mem->userspace_addr)))
> +             return -EINVAL;
> +     /* We can read the guest memory with __xxx_user() later on. */
> +     if (!(mem->flags & KVM_MEM_FD) &&
> +         !access_ok((void __user *)(unsigned long)mem->userspace_addr,
>                       mem->memory_size))
>               return -EINVAL;
>       if (as_id >= KVM_ADDRESS_SPACE_NUM || id >= KVM_MEM_SLOTS_NUM)
> @@ -1727,6 +1739,9 @@ int __kvm_set_memory_region(struct kvm *kvm,
>               new.dirty_bitmap = NULL;
>               memset(&new.arch, 0, sizeof(new.arch));
>       } else { /* Modify an existing slot. */
> +             /* Private memslots are immutable, they can only be deleted. */
> +             if (mem->flags & KVM_MEM_FD && mem->private_fd >= 0)
> +                     return -EINVAL;
>               if ((new.userspace_addr != old.userspace_addr) ||
>                   (new.npages != old.npages) ||
>                   ((new.flags ^ old.flags) & KVM_MEM_READONLY))
>