> Introducing folio_dup_file_rmap_pud() is better. But please see how I move > the comment and how passing mm+vma to the helper is not really reasonable.
Thanks, agreed on both. v2 will pass only the VMAs, with the mms taken from them as in your patch, drop src_mm from copy_pud_range() and move the TODO next to the rmap duplication. folio_dup_file_rmap_pud() stays. Could v2 then replace your 01/36, so you can drop it from your series? I'd add your Suggested-by; if you prefer Co-developed-by with your Signed-off-by from the patch above, I'll use that instead. -- Best regards, Konstantin Khorenko
