On Thu, Aug 13, 2026, Binbin Wu wrote:
> On 8/8/2026 5:52 AM, Ackerley Tng via B4 Relay wrote:
> > +#ifdef CONFIG_HAVE_KVM_ARCH_GMEM_CONVERT
> > +static void kvm_gmem_make_shared(struct inode *inode, pgoff_t start,
> > pgoff_t end)
> > +{
> > + struct folio_batch fbatch;
> > + pgoff_t next = start;
> > + int i;
> > +
> > + folio_batch_init(&fbatch);
> > + while (filemap_get_folios(inode->i_mapping, &next, end - 1, &fbatch)) {
> > + for (i = 0; i < folio_batch_count(&fbatch); ++i) {
> > + struct folio *folio = fbatch.folios[i];
> > + pgoff_t start_index, end_index;
> > + kvm_pfn_t start_pfn;
> > + kvm_pfn_t nr_pages;
> > +
> > + start_index = max(start, folio->index);
> > + end_index = min(end, folio_next_index(folio));
> > + /*
> > + * end_index is either in folio or points to
> > + * the first page of the next folio. Hence,
> > + * all pages in range [start_index, end_index)
> > + * are contiguous.
> > + */
> > + start_pfn = folio_file_pfn(folio, start_index);
> > + nr_pages = end_index - start_index;
> > +
> > + kvm_arch_gmem_make_shared(start_pfn, nr_pages);
> > + }
> > +
> > + folio_batch_release(&fbatch);
> > + cond_resched();
> > + }
> > +}
> > +#else
> > +static void kvm_gmem_make_shared(struct inode *inode, pgoff_t start,
> > pgoff_t end) {}
> > +#endif
> > +
> > static int __kvm_gmem_set_attributes(struct inode *inode, pgoff_t start,
> > size_t nr_pages, uint64_t attrs,
> > pgoff_t *err_index)
> > @@ -599,7 +636,12 @@ static int __kvm_gmem_set_attributes(struct inode
> > *inode, pgoff_t start,
> >
> > filter = to_private ? KVM_FILTER_SHARED : KVM_FILTER_PRIVATE;
> > kvm_gmem_invalidate_start(inode, start, end, filter);
> > +
> > + if (!to_private)
> > + kvm_gmem_make_shared(inode, start, end);
>
> If both KVM_AMD_SEV and KVM_INTEL_TDX are enabled, HAVE_KVM_ARCH_GMEM_CONVERT
> will be enabled and the logic in kvm_gmem_make_shared() introduces unnecessary
> overhead for TDX.
> Not sure about CSPs, but in a standard distribution kernel, it's very likely
> that both are enabled, right？
Yep.
> Should kvm_gmem_make_shared() do some optimization or the overhead is relative
> small in the conversion to shared path so that the optimization is not worth
> it?
I assume it's relatively small, but it's probably a good idea to gets numbers so
that we can make a semi-informed decision. E.g. if processing 1GiB work of 4KiB
folios is <100 cycles, I think we generally don't care. But if it's more like
tens of microseconds, then I think we do care enough to optimize the walking for
TDX.
