[PATCH v2 03/10] mm/thp: try_to_unmap() use TTU_SYNC for safe splitting (fwd)

[Hugh Dickins]

---------- Forwarded message ----------
Date: Tue, 8 Jun 2021 21:10:19 -0700 (PDT)
From: Hugh Dickins <hughd@xxxxxxxxxx>
To: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
Cc: Hugh Dickins <hughd@xxxxxxxxxx>,
    Kirill A. Shutemov <kirill.shutemov@xxxxxxxxxxxxxxx>,
    Yang Shi <shy828301@xxxxxxxxx>, Wang Yugui <wangyugui@xxxxxxxxxxxx>,
    Matthew Wilcox <willy@xxxxxxxxxxxxx>,
    Naoya Horiguchi <naoya.horiguchi@xxxxxxx>,
    Alistair Popple <apopple@xxxxxxxxxx>, Ralph Campbell <rcampbell@xxxxxxxxxx>,
    Zi Yan <ziy@xxxxxxxxxx>, Miaohe Lin <linmiaohe@xxxxxxxxxx>,
    Minchan Kim <minchan@xxxxxxxxxx>, Jue Wang <juew@xxxxxxxxxx>,
    Peter Xu <peterx@xxxxxxxxxx>, Jan Kara <jack@xxxxxxx>,
    Shakeel Butt <shakeelb@xxxxxxxxxx>, Oscar Salvador <osalvador@xxxxxxx>
Subject: [PATCH v2 03/10] mm/thp: try_to_unmap() use TTU_SYNC for safe splitting

Stressing huge tmpfs often crashed on unmap_page()'s VM_BUG_ON_PAGE
(!unmap_success): with dump_page() showing mapcount:1, but then its
raw struct page output showing _mapcount ffffffff i.e. mapcount 0.

And even if that particular VM_BUG_ON_PAGE(!unmap_success) is removed,
it is immediately followed by a VM_BUG_ON_PAGE(compound_mapcount(head)),
and further down an IS_ENABLED(CONFIG_DEBUG_VM) total_mapcount BUG():
all indicative of some mapcount difficulty in development here perhaps.
But the !CONFIG_DEBUG_VM path handles the failures correctly and silently.

I believe the problem is that once a racing unmap has cleared pte or pmd,
try_to_unmap_one() may skip taking the page table lock, and emerge from
try_to_unmap() before the racing task has reached decrementing mapcount.

Instead of abandoning the unsafe VM_BUG_ON_PAGE(), and the ones that
follow, use PVMW_SYNC in try_to_unmap_one() in this case: adding TTU_SYNC
to the options, and passing that from unmap_page().

When CONFIG_DEBUG_VM, or for non-debug too?  Consensus is to do the same
for both: the slight overhead added should rarely matter, except perhaps
if splitting sparsely-populated multiply-mapped shmem.  Once confident
that bugs are fixed, TTU_SYNC here can be removed, and the race tolerated.

Fixes: fec89c109f3a ("thp: rewrite freeze_page()/unfreeze_page() with generic rmap walkers")
Signed-off-by: Hugh Dickins <hughd@xxxxxxxxxx>
Cc: <stable@xxxxxxxxxxxxxxx>
---
v2: moved TTU_SYNC definition up, to avoid conflict with other patchset
    use TTU_SYNC even when non-debug, per Peter Xu and Yang Shi
    expanded PVMW_SYNC's spin_unlock(pmd_lock()), per Kirill and Peter

 include/linux/rmap.h |  1 +
 mm/huge_memory.c     |  2 +-
 mm/page_vma_mapped.c | 11 +++++++++++
 mm/rmap.c            | 17 ++++++++++++++++-
 4 files changed, 29 insertions(+), 2 deletions(-)

diff --git a/include/linux/rmap.h b/include/linux/rmap.h
index def5c62c93b3..8d04e7deedc6 100644
--- a/include/linux/rmap.h
+++ b/include/linux/rmap.h
@@ -91,6 +91,7 @@ enum ttu_flags {
 
 	TTU_SPLIT_HUGE_PMD	= 0x4,	/* split huge PMD if any */
 	TTU_IGNORE_MLOCK	= 0x8,	/* ignore mlock */
+	TTU_SYNC		= 0x10,	/* avoid racy checks with PVMW_SYNC */
 	TTU_IGNORE_HWPOISON	= 0x20,	/* corrupted page is recoverable */
 	TTU_BATCH_FLUSH		= 0x40,	/* Batch TLB flushes where possible
 					 * and caller guarantees they will
diff --git a/mm/huge_memory.c b/mm/huge_memory.c
index 5885c5f5836f..84ab735139dc 100644
--- a/mm/huge_memory.c
+++ b/mm/huge_memory.c
@@ -2350,7 +2350,7 @@ void vma_adjust_trans_huge(struct vm_area_struct *vma,
 
 static void unmap_page(struct page *page)
 {
-	enum ttu_flags ttu_flags = TTU_IGNORE_MLOCK |
+	enum ttu_flags ttu_flags = TTU_IGNORE_MLOCK | TTU_SYNC |
 		TTU_RMAP_LOCKED | TTU_SPLIT_HUGE_PMD;
 	bool unmap_success;
 
diff --git a/mm/page_vma_mapped.c b/mm/page_vma_mapped.c
index 2cf01d933f13..5b559967410e 100644
--- a/mm/page_vma_mapped.c
+++ b/mm/page_vma_mapped.c
@@ -212,6 +212,17 @@ bool page_vma_mapped_walk(struct page_vma_mapped_walk *pvmw)
 			pvmw->ptl = NULL;
 		}
 	} else if (!pmd_present(pmde)) {
+		/*
+		 * If PVMW_SYNC, take and drop THP pmd lock so that we
+		 * cannot return prematurely, while zap_huge_pmd() has
+		 * cleared *pmd but not decremented compound_mapcount().
+		 */
+		if ((pvmw->flags & PVMW_SYNC) &&
+		    PageTransCompound(pvmw->page)) {
+			spinlock_t *ptl = pmd_lock(mm, pvmw->pmd);
+
+			spin_unlock(ptl);
+		}
 		return false;
 	}
 	if (!map_pte(pvmw))
diff --git a/mm/rmap.c b/mm/rmap.c
index 693a610e181d..07811b4ae793 100644
--- a/mm/rmap.c
+++ b/mm/rmap.c
@@ -1405,6 +1405,15 @@ static bool try_to_unmap_one(struct page *page, struct vm_area_struct *vma,
 	struct mmu_notifier_range range;
 	enum ttu_flags flags = (enum ttu_flags)(long)arg;
 
+	/*
+	 * When racing against e.g. zap_pte_range() on another cpu,
+	 * in between its ptep_get_and_clear_full() and page_remove_rmap(),
+	 * try_to_unmap() may return false when it is about to become true,
+	 * if page table locking is skipped: use TTU_SYNC to wait for that.
+	 */
+	if (flags & TTU_SYNC)
+		pvmw.flags = PVMW_SYNC;
+
 	/* munlock has nothing to gain from examining un-locked vmas */
 	if ((flags & TTU_MUNLOCK) && !(vma->vm_flags & VM_LOCKED))
 		return true;
@@ -1777,7 +1786,13 @@ bool try_to_unmap(struct page *page, enum ttu_flags flags)
 	else
 		rmap_walk(page, &rwc);
 
-	return !page_mapcount(page) ? true : false;
+	/*
+	 * When racing against e.g. zap_pte_range() on another cpu,
+	 * in between its ptep_get_and_clear_full() and page_remove_rmap(),
+	 * try_to_unmap() may return false when it is about to become true,
+	 * if page table locking is skipped: use TTU_SYNC to wait for that.
+	 */
+	return !page_mapcount(page);
 }
 
 /**
-- 
2.26.2