[PATCH 00/11] mm: Unify memmap initialization for compound pages

[Muchun Song]

Hi,

This series is based on the mm-new branch at commit b95f8bb86713.

This series is split out from the earlier, larger series
"mm: Generalize HVO for HugeTLB and device DAX" [1]. That series
generalized vmemmap optimization across HugeTLB and device DAX. One
remaining piece is the separate compound struct page initialization used
by boot-time gigantic HugeTLB pages and compound ZONE_DEVICE ranges.

The prerequisite changes already let both users derive their HVO layout
from sparsemem section metadata and use the common sparse-vmemmap
population path. Their post-population initialization still differs,
however. memmap_init_zone_device() builds compound pages from reusable
templates, whereas boot-time HugeTLB retains separate HVO fallback,
compound-page, and pageblock migratetype handling.

This series consolidates those paths in three stages.

The first four patches make the generic memmap initialization code
suitable for sharing. They add zone-consistency verification to template
copies, factor the PFN-independent page state, simplify the ZONE_DEVICE
template setup, and introduce a helper for initializing frozen compound
pages.

The next four patches simplify boot-time gigantic HugeTLB initialization.
They make it rely solely on section-based HVO, mark the entire gigantic
allocation as noinit so HugeTLB owns its memmap setup, and switch the
retained struct pages to the common frozen compound-page helper. With the
last external user gone, __init_single_page() becomes local to mm_init.c.

The final three patches move initialization of early-reserved CMA
pageblocks into cma_activate_area() and localize the CMA helper. This
allows MIGRATE_MOVABLE initialization to move into the shared memmap
helpers. CMA later overrides those pageblocks to MIGRATE_CMA during
activation, while non-CMA HugeTLB and ZONE_DEVICE users retain the
default migratetype.

After this series, boot-time HugeTLB and compound ZONE_DEVICE ranges,
including device DAX, use the same helpers for frozen compound-page and
pageblock initialization.

[1] https://lore.kernel.org/20260513130542.35604-1-songmuchun@xxxxxxxxxxxxx/

Muchun Song (11):
  mm/mm_init: verify zones for template-initialized pages
  mm/mm_init: factor out common struct page initialization
  mm/mm_init: simplify zone device page initialization
  mm/mm_init: factor out frozen compound page initialization
  mm/hugetlb: drop boot-time HVO handling for gigantic folios
  mm/hugetlb: initialize all struct pages for bootmem hugepages
  mm/hugetlb: use common compound page initialization helper
  mm/mm_init: make __init_single_page() static
  mm/cma: centralize CMA pageblock initialization
  mm/cma: localize init_cma_pageblock() to cma.c
  mm/mm_init: move pageblock migratetype init to memmap helpers

 include/linux/mm.h   |   8 --
 mm/cma.c             |  24 +++--
 mm/hugetlb.c         | 111 ++-------------------
 mm/hugetlb_vmemmap.c |  17 +---
 mm/mm_init.c         | 232 ++++++++++++++++---------------------------
 mm/mm_init.h         |  14 +--
 mm/sparse-vmemmap.c  |   7 +-
 7 files changed, 120 insertions(+), 293 deletions(-)


base-commit: b95f8bb86713c860ce8536d22a988b949b884d5f
-- 
2.54.0