On Sat, Apr 27, 2024 at 04:26:44AM +0100, Matthew Wilcox wrote: > There's a series of commits in linux-mm with the titles: > > sparc: use is_huge_zero_pmd() > mm: add is_huge_zero_folio() > mm: add pmd_folio() > mm: convert migrate_vma_collect_pmd to use a folio > mm: convert huge_zero_page to huge_zero_folio > mm: convert do_huge_pmd_anonymous_page to huge_zero_folio > dax: use huge_zero_folio > mm: rename mm_put_huge_zero_page to mm_put_huge_zero_folio > > > it available for non-hugetlb setups? Not only would this be cleaner > > and more efficient, but it would actually work for the case where you'd > > have to zero more than 1MB on a 4k PAGE_SIZE system, which doesn't > > seem impossible with 2MB folios. > > It is available for non-hugetlb setups. It is however allocated on > demand, so it might not be available. We could just export get_huge_zero_page/put_huge_zero_page and make sure it is is available for block sizse > PAGE_SIZE file systems, or is there a good argument against that?
