Skip to contentSkip to conversation

Core MM fixes for 7.4-rc4

GIT PULL Merged David Hildenbrand (Arm) 2 participants 2 messages Nov 17 Memory management

lore
Share & tools
lore
#0 lore raw

Hi Linus,

please pull these core-MM fixes for 7.4-rc4.

I wanted to have them ready for -rc3, but decided to give some of them a bit more soaking time in -next. These are all urgent fixes; non-urgent fixes are getting piled up for the next merge window.

The biggest piece is the folio refcount overflow prevention, which you should be aware of and which I want to have fixed ASAP. What's missing from that work are the net patches.

I'll try getting more attention from net maintainers on the net changes next, so they can go through the net tree.

Cheers,

David

----------------------------------------------------------------

The following changes since commit f66469d9905b39d2c260a15d7d6db450f4db67c1:

Merge tag 'net-7.4-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net (2026-11-05 18:00:16 -0800)

are available in the Git repository at:

git://git.kernel.org/pub/scm/linux/kernel/git/mm/core.git tags/mm-core-7.4-rc4

for you to fetch changes up to 25f5c92e1ac06496679ea229e5c825aa88968efd:

Merge patch series "mm: limit pagecache folio refcount" (2026-11-14 16:36:05 +0100)

----------------------------------------------------------------
Core memory management fixes for 7.4-rc4

This includes fixes to:

- THP (Transparent Huge Pages)
- GUP (Get User Pages)
- USERFAULTFD
- SHARED MEMORY (memfd)
- Core MM

The main changes are:

- Prevent folio refcount overflows with pagecache folios, mostly relevant for large folios today. The networking changes for this will go through the net tree.

- Fix out-of-bounds page table accesses with folded PMDs in 3 scenarios.

- Fix write-sealing of memfd to detect GUP pins with concurrent (un)mapping more reliably, preventing writes through GUP pins after write-sealing.

- Fix lost wakeups when splitting large folios.

- Fix exposing uninitialized data through MADV_COLLAPSE to user space.

The refcount changes also include a related iommufd fix for properly cleaning up when pinning memfd folios fails, which is now easier to trigger.

----------------------------------------------------------------
David Hildenbrand (Arm) (11):
mm/khugepaged: don't map folios that are not uptodate in MADV_COLLAPSE mm/memfd: handle GUP pins more reliably when write-sealing mm/userfaultfd: fix PTE table overruns in UFFDIO_MOVE mm/rmap: fix PTE table overruns during batched unmap mm/rmap: fix PTE table overruns during batched reference checks
mm: limit pagecache folio refcount when mapping pages
mm: limit pagecache folio refcount when splicing into pipes mm/mmu_gather: flush pending references early before the folio limit
iommufd: fix cleanup when adding memfd folio pins fails mm/gup: limit pagecache folio refcount when grabbing references Merge patch series "mm: limit pagecache folio refcount"

Zi Yan (1):
mm/huge_memory: prevent losing PG_waiters in __split_folio_to_order()

 drivers/iommu/iommufd/pages.c | 15 ++++++--
 fs/fuse/dev.c                 |  4 +-
 include/asm-generic/tlb.h     |  4 +-
 include/linux/mm.h            | 89 ++++++++++++++++++++++++++++++++++++-------
 include/linux/pgtable.h       | 24 ++++++++++++
 mm/filemap.c                  | 28 +++++++++-----
 mm/folio.c                    | 26 +++++++++++++
 mm/gup.c                      | 32 ++++++++++------
 mm/huge_memory.c              | 64 ++++++++++++++++++++++++++++---
 mm/internal.h                 |  6 ++-
 mm/khugepaged.c               |  6 +++
 mm/memfd.c                    |  7 ++++
 mm/memory.c                   | 48 +++++++++++++++++------
 mm/migrate.c                  | 11 ++++++
 mm/mmu_gather.c               | 10 +++--
 mm/page_alloc.c               | 17 ++++++++-
 mm/rmap.c                     |  4 +-
 mm/shmem.c                    |  7 +++-
 mm/userfaultfd.c              |  8 +++-
 19 files changed, 336 insertions(+), 74 deletions(-)