Re: [PATCH v9 00/15] Use generic_file_read_iter() in hugetlbfs

On Wed, Aug 05, 2026 at 10:05:40PM +0100, Matthew Wilcox (Oracle) wrote:
> This set of patches are separated out from Jane's earlier series.
> https://lore.kernel.org/linux-mm/[email protected]/
>
> We need three things to use filemap_read_iter() instead of a hugetlbfs
> specific read_iter: hwpoison handling, treating gaps in the pagecache as
> holes, and having hugetlb set the folio size in the mapping.  hwopison
> handling turned out to need a lot of prep work; the other two things
> are fairly straightforward.
>
> Sashiko complains that the implementation of AS_AUTHORITATIVE is
> incomplete, but hugetlbfs doesn't use any of that functionality.
> We'll add it when we need it, either when hugetlbfs uses it, or when we
> convert shmem to use filemap.
>
> v8 here:
> https://lore.kernel.org/linux-mm/[email protected]/
>
> Changes from v7:
>  - Fix the KPF_HWPOISON blooper
>  - Expand HVO in a comment
>  - Collect R-b tags from Gregory & Jane
>
> Jane Chu (2):
>   filemap: Add hwpoison handling to filemap_read()
>   hugetlb: replace hugetlbfs_read_iter() with generic_file_read_iter()
>
> Matthew Wilcox (Oracle) (13):
>   memory-failure: Fix hardware poison check in unpoison_memory() again
>   memory-failure: Prevent hugetlb freeing during unpoisoning
>   mm: Rename folio_contain_hwpoison_page() to folio_has_hwpoison_page()
>   hugetlb: Mark some function arguments as const
>   guest_memfd: Use folio_has_hwpoisoned_page()
>   kpageflags: Use is_page_hwpoison() to set KPF_HWPOISON
>   hugetlb: Move poison to pages before clearing hugetlb page type
>   hugetlb: Use the has_hwpoisoned flag
>   mm: Remove locking mf_mutex in is_raw_hwpoison_page_in_hugepage()
>   mm: Check individual hugetlb pages for poison
>   filemap: Remove checks in mapping_set_folio_order_range()
>   hugetlb: Set mapping folio order
>   filemap: Add support for authoritative mappings
>
>  fs/Kconfig                 |   2 +-
>  fs/hugetlbfs/inode.c       | 123 +++--------------------
>  fs/proc/page.c             |   8 +-
>  include/linux/hugetlb.h    |  43 +++++---
>  include/linux/mm_types.h   |   4 +-
>  include/linux/page-flags.h |  64 ++++++++++--
>  include/linux/pagemap.h    |  45 +++++----
>  mm/Kconfig                 |   6 +-
>  mm/filemap.c               |  51 +++++++++-
>  mm/huge_memory.c           |   2 +-
>  mm/hugetlb.c               |  41 ++++----
>  mm/memory-failure.c        | 197 +++++++++++++++++++++++++------------
>  mm/memory_hotplug.c        |   2 +-
>  mm/rmap.c                  |  43 +++++---

Much as I'd like to not look at this, nor think about it, nor even admit that
any of this actually exists, I do co-maintain the rmap with David so please do
cc- me on future revisions :)

>  mm/shmem.c                 |   2 +-
>  mm/vmscan.c                |   2 +-
>  virt/kvm/guest_memfd.c     |   2 +-
>  17 files changed, 368 insertions(+), 269 deletions(-)
>
> --
> 2.47.3
>
>

--
Cheers, Lorenzo