Re: [PATCH v3 6/6] KVM: guest_memfd: GUP source pages prior to populating guest memory

[Sean Christopherson <seanjc@google.com>]

On Tue, Jan 13, 2026, Michael Roth wrote:
> On Tue, Jan 13, 2026 at 11:21:29AM -0800, Sean Christopherson wrote:
> > On Thu, Jan 08, 2026, Michael Roth wrote:
> > > @@ -842,47 +881,38 @@ long kvm_gmem_populate(struct kvm *kvm, gfn_t start_gfn, void __user *src, long
> > >  	if (!file)
> > >  		return -EFAULT;
> > >  
> > > -	filemap_invalidate_lock(file->f_mapping);
> > > -
> > >  	npages = min_t(ulong, slot->npages - (start_gfn - slot->base_gfn), npages);
> > >  	for (i = 0; i < npages; i++) {
> > > -		struct folio *folio;
> > > -		gfn_t gfn = start_gfn + i;
> > > -		pgoff_t index = kvm_gmem_get_index(slot, gfn);
> > > -		kvm_pfn_t pfn;
> > > +		struct page *src_page = NULL;
> > > +		void __user *p;
> > >  
> > >  		if (signal_pending(current)) {
> > >  			ret = -EINTR;
> > >  			break;
> > >  		}
> > >  
> > > -		folio = __kvm_gmem_get_pfn(file, slot, index, &pfn, NULL);
> > > -		if (IS_ERR(folio)) {
> > > -			ret = PTR_ERR(folio);
> > > -			break;
> > > -		}
> > > +		p = src ? src + i * PAGE_SIZE : NULL;
> > >  
> > > -		folio_unlock(folio);
> > > +		if (p) {
> > 
> > Computing 'p' when src==NULL is unnecessary and makes it hard to see that gup()
> > is done if and only if src!=NULL.
> > 
> > Anyone object to this fixup?
> 
> No objections here, and it does seem a bit more readable. Will include
> this if a new version is needed.

No need, I'll fixup when applying (already have, actually).  If you want to double
check that I didn't fat finger anything, the patches are in kvm-x86/gmem.

Thanks!