[Patch v4] mm/huge_memory: fix early failure try_to_migrate() when split huge pmd for shared THP

[Patch v4] mm/huge_memory: fix early failure try_to_migrate() when split huge pmd for shared THP

[Wei Yang]

Commit 60fbb14396d5 ("mm/huge_memory: adjust try_to_migrate_one() and
split_huge_pmd_locked()") return false unconditionally after
split_huge_pmd_locked(). This may fail try_to_migrate() early when
TTU_SPLIT_HUGE_PMD is specified.

The reason is the above commit adjusted try_to_migrate_one() to, when a
PMD-mapped THP entry is found, and TTU_SPLIT_HUGE_PMD is specified (for
example, via unmap_folio()), return false unconditionally. This breaks the
rmap walk and fail try_to_migrate() early, if this PMD-mapped THP is mapped
in multiple processes.

The user sensible impact of this bug could be:

  * On memory pressure, shrink_folio_list() may split partially mapped
    folio with split_folio_to_list(). Then free unmapped pages without IO.
    If failed, it may not be reclaimed.
  * On memory failure, memory_failure() would call try_to_split_thp_page()
    to split folio contains the bad page. If succeed, the PG_has_hwpoisoned
    bit is only set in the after-split folio contains @split_at. By doing
    so, we limit bad memory. If failed to split, the whole folios is not
    usable.

One way to reproduce:

    Create an anonymous THP range and fork 512 children, so we have a
    THP shared mapped in 513 processes. Then trigger folio split with
    /sys/kernel/debug/split_huge_pages debugfs to split the THP folio to
    order 0.

Without the above commit, we can successfully split to order 0.
With the above commit, the folio is still a large folio.

And currently there are two core users of TTU_SPLIT_HUGE_PMD:

  * try_to_unmap_one()
  * try_to_migrate_one()

try_to_unmap_one() would restart the rmap walk, so only
try_to_migrate_one() is affected.

We can't simply revert commit 60fbb14396d5 ("mm/huge_memory: adjust
try_to_migrate_one() and split_huge_pmd_locked()"), since it removed some
duplicated check covered by page_vma_mapped_walk().

This patch fixes this by restart page_vma_mapped_walk() after
split_huge_pmd_locked(). Since we cannot simply return "true" to fix the
problem, as that would affect another case:

    When invoking folio_try_share_anon_rmap_pmd() from
    split_huge_pmd_locked(), the latter can fail and leave a large folio
    mapped through PTEs, in which case we ought to return true from
    try_to_migrate_one(). This might result in unnecessary walking of the
    rmap but is relatively harmless.

Fixes: 60fbb14396d5 ("mm/huge_memory: adjust try_to_migrate_one() and split_huge_pmd_locked()")
Signed-off-by: Wei Yang <richard.weiyang@gmail.com>
Reviewed-by: Baolin Wang <baolin.wang@linux.alibaba.com>
Reviewed-by: Zi Yan <ziy@nvidia.com>
Tested-by: Lance Yang <lance.yang@linux.dev>
Reviewed-by: Lance Yang <lance.yang@linux.dev>
Reviewed-by: Gavin Guo <gavinguo@igalia.com>
Acked-by: David Hildenbrand (arm) <david@kernel.org>
Reviewed-by: Lorenzo Stoakes (Oracle) <ljs@kernel.org>
Cc: Gavin Guo <gavinguo@igalia.com>
Cc: "David Hildenbrand (Red Hat)" <david@kernel.org>
Cc: Zi Yan <ziy@nvidia.com>
Cc: Baolin Wang <baolin.wang@linux.alibaba.com>
Cc: Lance Yang <lance.yang@linux.dev>
Cc: <stable@vger.kernel.org>

---
v4:
  * only commit msg adjustment
    - rephrase the reason analysis
    - move reproduce method afterward
    - more explanation on user sensible effect of the bug, especially expand
      what "Limit bad page" means
    - remove the explanation on whey it need to fork 512 child for reproduce
    - explain why simply revert commit 60fbb14396d5 is not taken
    - mention TTU_SPLIT_HUGE_PMD users and confirm not affect others
    - rephrase the reason why can't simply return true
v3:
  * gather RB
  * adjust the commit log and comment per David
  * add userspace-visible runtime effect in change log
v2:
  * restart page_vma_mapped_walk() after split_huge_pmd_locked()
---
 mm/rmap.c | 12 +++++++++---
 1 file changed, 9 insertions(+), 3 deletions(-)

diff --git a/mm/rmap.c b/mm/rmap.c
index beb423f3e8ec..e609dd5b382f 100644
--- a/mm/rmap.c
+++ b/mm/rmap.c
@@ -2444,11 +2444,17 @@ static bool try_to_migrate_one(struct folio *folio, struct vm_area_struct *vma,
 			__maybe_unused pmd_t pmdval;
 
 			if (flags & TTU_SPLIT_HUGE_PMD) {
+				/*
+				 * split_huge_pmd_locked() might leave the
+				 * folio mapped through PTEs. Retry the walk
+				 * so we can detect this scenario and properly
+				 * abort the walk.
+				 */
 				split_huge_pmd_locked(vma, pvmw.address,
 						      pvmw.pmd, true);
-				ret = false;
-				page_vma_mapped_walk_done(&pvmw);
-				break;
+				flags &= ~TTU_SPLIT_HUGE_PMD;
+				page_vma_mapped_walk_restart(&pvmw);
+				continue;
 			}
 #ifdef CONFIG_ARCH_ENABLE_THP_MIGRATION
 			pmdval = pmdp_get(pvmw.pmd);
-- 
2.34.1