Re: [PATCH v2 5/5] KVM: SNP: Mark source page dirty in sev_gmem_post_populate

[Sean Christopherson <seanjc@google.com>]

On Fri, May 22, 2026, Ackerley Tng wrote:
> Mark the folio as dirty after copying data into the source page in
> sev_gmem_post_populate. After the memcpy, failing to mark the page dirty
> can lead to the memory management subsystem discarding the changes if the
> page is reclaimed or otherwise processed by the swap subsystem.
> 
> Fixes: 2a62345b3052 ("KVM: guest_memfd: GUP source pages prior to populating guest memory")
> Signed-off-by: Ackerley Tng <ackerleytng@google.com>
> ---
>  arch/x86/kvm/svm/sev.c | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/arch/x86/kvm/svm/sev.c b/arch/x86/kvm/svm/sev.c
> index dbf75326a40f4..1a361f08c7a3d 100644
> --- a/arch/x86/kvm/svm/sev.c
> +++ b/arch/x86/kvm/svm/sev.c
> @@ -2395,6 +2395,7 @@ static int sev_gmem_post_populate(struct kvm *kvm, gfn_t gfn, kvm_pfn_t pfn,
>  		void *dst_vaddr = kmap_local_pfn(pfn);
>  
>  		memcpy(src_vaddr, dst_vaddr, PAGE_SIZE);
> +		folio_mark_dirty(page_folio(src_page));

I'd rather use set_page_dirty().  I'll fixup when applying, unless someon objects.

>  		kunmap_local(dst_vaddr);
>  		kunmap_local(src_vaddr);
> 
> -- 
> 2.54.0.794.g4f17f83d09-goog
>