mm-userfaultfd-detect-vma-type-change-after-copy-retry-in-mfill_copy_folio_retry.patch removed from -mm tree
The quilt patch titled
Subject: mm/userfaultfd: detect VMA type change after copy retry in mfill_copy_folio_retry()
has been removed from the -mm tree. Its filename was
mm-userfaultfd-detect-vma-type-change-after-copy-retry-in-mfill_copy_folio_retry.patch
This patch was dropped because it was merged into the mm-hotfixes-stable branch of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
------------------------------------------------------
From: David Carlier <[email protected]>
Subject: mm/userfaultfd: detect VMA type change after copy retry in mfill_copy_folio_retry()
Date: Fri, 24 Apr 2026 19:36:38 +0100
mfill_copy_folio_retry() drops mmap_lock for the copy_from_user() call. During this window, the VMA can be replaced with a different type (e.g. hugetlb), making the caller's ops pointer stale. Subsequent use of the stale ops would dispatch into the wrong per-vma handlers.
Capture the VMA's ops via vma_uffd_ops() before dropping the lock and compare against the current vma_uffd_ops() after re-acquiring it. Return -EAGAIN if they differ so the operation can be retried. This avoids comparing against the caller's ops which may have been overridden to anon_uffd_ops for MAP_PRIVATE file-backed mappings.
---
mm/userfaultfd.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-)
mm/userfaultfd.c renamed from mm/userfaultfd.c~mm-userfaultfd-detect-vma-type-change-after-copy-retry-in-mfill_copy_folio_retry +11 -1
@@ -443,8 +443,10 @@ static int mfill_copy_folio_locked(strucreturn ret; } -static int mfill_copy_folio_retry (struct mfill_state *state,struct folio *folio) +static int mfill_copy_folio_retry (struct mfill_state *state, +struct folio *folio) { +const struct vm_uffd_ops *orig_ops =vma_uffd_ops (state->vma);unsigned long src_addr = state->src_addr;void *kaddr;int err;
@@ -465,6 +467,14 @@ static int mfill_copy_folio_retry(structif (err)return err; +/* +* The VMA type may have changed while the lock was dropped +* (e.g. replaced with a hugetlb mapping), making the caller's +* ops pointer stale. +*/ +if (vma_uffd_ops (state->vma) != orig_ops) +return -EAGAIN ; + err =mfill_establish_pmd (state);if (err)return err;
_
Patches currently in -mm which might be from [email protected] are
mm-hugetlb-restore-reservation-on-error-in-hugetlb_mfill_atomic_pte-resubmission-path.patch mm-page_io-rename-swap_iocb-fields-for-clarity.patch