[PATCH 0/3] mm: fix races between huge and non-huge faults in PFN mappings
From: Lorenzo Stoakes (ARM)
Date: Sun Nov 15 2026 - 15:59:56 EST
A driver that implements both ->fault and ->huge_fault must anticipate
concurrent use of both.
This can happen, for example, if a user calls prctl(PR_SET_THP_DISABLE)
concurrent with faulting, as that only holds the mmap write lock and thus
does not exclude faulting threads stabilised by VMA lock.
A number of places in the code that implements huge PFN support fails to
take this into account correctly, which this series aims to address.
Generally the shape of the issue is that faulting the same empty page table
entry can result in the ->fault and ->huge_fault handlers racing and the
loser finding something they didn't expect.
The fault code must tolerate this, but fails to in 3 places:
1. walk_to_pmd() asserts, without the PMD lock, that the PMD entry is not
huge, so a VM_BUG_ON() triggers. [ vfio-pci, nvgrace-gpu ]
2. insert_pfn() treats a failure to map the PTE table as an allocation
failure and reports OOM, causing kernel accesses to the range to
spuriously fail with -EFAULT. [ vfio-pci, nvgrace-gpu ]
3. insert_pmd() and insert_pud() assume a populated entry is huge and
compare its PFN on write fault, warning when that of a page table does
not match. [ mshv_vtl, drm_gem_syhmem, panthor ]
This was originally reported and fixed by Eliav Farber in [0], this series
reworks patches 1-3 of that one.
Patch 4 of that series (x86 PAT encoding of huge PFN mappings) is dropped -
it is unrelated to the above, no in-tree driver is affected by it, and it
needs a different approach which is better pursued separately with the x86
maintainers.
All three issues were reproduced on x86-64 under QEMU using vfio-pci in
no-IOMMU mode on an ivshmem BAR, with 6 threads faulting the same address
in each PMD or PUD aligned range of the mapping while another toggles
PR_SET_THP_DISABLE (note that vfio-pci had to be modified to pass the write
flag to reproduce the third issue.)
Without this series the BUG is hit immediately, with only patch 1 applied
tens of thousands of -EFAULTs are observed within seconds, and with patches
1 and 2 (and the vfio-pci modification) both warnings fire immediately.
With the series applied 30 second runs at each of PMD and PUD size complete
with none of the above, with or without the vfio-pci modification.
[0]: https://lore.kernel.org/all/20261112051440.16140-1-farbere@xxxxxxxxxx/
---
To: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
To: David Hildenbrand <david@xxxxxxxxxx>
To: "Liam R. Howlett" <liam@xxxxxxxxxxxxx>
To: Vlastimil Babka <vbabka@xxxxxxxxxx>
To: Mike Rapoport <rppt@xxxxxxxxxx>
To: Suren Baghdasaryan <surenb@xxxxxxxxxx>
To: Michal Hocko <mhocko@xxxxxxxx>
To: Alex Williamson <alex@xxxxxxxxxxx>
To: Peter Xu <peterx@xxxxxxxxxx>
To: Zi Yan <ziy@xxxxxxxxxx>
To: Baolin Wang <baolin.wang@xxxxxxxxxxxxxxxxx>
To: Nico Pache <nico.pache@xxxxxxxxx>
To: Ryan Roberts <ryan.roberts@xxxxxxx>
To: Dev Jain <dev.jain@xxxxxxx>
To: Barry Song <baohua@xxxxxxxxxx>
To: Lance Yang <lance.yang@xxxxxxxxx>
To: Usama Arif <usama.arif@xxxxxxxxx>
To: Kiryl Shutsemau <kas@xxxxxxxxxx>
To: Wei Liu <wei.liu@xxxxxxxxxx>
To: Roman Kisel <romank@xxxxxxxxxxxxxxxxxxx>
To: Saurabh Sengar <ssengar@xxxxxxxxxxxxxxxxxxx>
To: Naman Jain <namjain@xxxxxxxxxxxxxxxxxxx>
To: Michael Kelley <mhklinux@xxxxxxxxxxx>
Cc: ljs@xxxxxxxxxx
Cc: linux-mm@xxxxxxxxx
Cc: linux-kernel@xxxxxxxxxxxxxxx
---
Eliav Farber (2):
mm/memory: remove racy huge PMD assertion from walk_to_pmd()
mm/memory: do not report OOM if insert_pfn() races a huge fault
Lorenzo Stoakes (ARM) (1):
mm/huge_memory: only compare PFNs of huge entries on huge fault insert
mm/huge_memory.c | 8 ++++++++
mm/memory.c | 12 +++++++++---
2 files changed, 17 insertions(+), 3 deletions(-)
---
base-commit: 5b99c3309de70337f38bd6e6998ab8335c2758e4
change-id: 20261115-huge-pfnmap-fault-races-534e626de5cc
Cheers,
--
Lorenzo Stoakes (ARM) <ljs@xxxxxxxxxx>