On 13.09.23 14:51, David Hildenbrand wrote:
Let's sanity-check PageAnonExclusive vs. mapcount in page_add_anon_rmap()
and hugepage_add_anon_rmap() after setting PageAnonExclusive simply by
re-reading the mapcounts.
We can stop initializing the "first" variable in page_add_anon_rmap()
and no longer need an atomic_inc_and_test() in hugepage_add_anon_rmap().
While at it, switch to VM_WARN_ON_FOLIO().
Signed-off-by: David Hildenbrand <david@xxxxxxxxxx>
---
Hi Andrew, the following fixup on top:
From 4b945ca15817b491123cc922f022b253134075b7 Mon Sep 17 00:00:00 2001
From: David Hildenbrand <david@xxxxxxxxxx>
Date: Mon, 18 Sep 2023 11:16:57 +0200
Subject: [PATCH] fixup: mm/rmap: simplify PageAnonExclusive sanity checks when
adding anon rmap
While PTE-mapping a THP, we temporarily have an exclusive page of a THP
mapped by both, a PMD and a PTE at the same time. Update our check to
allow for that combination.
Reported-by: syzbot+6e4f59235036c3c2e296@xxxxxxxxxxxxxxxxxxxxxxxxx
Signed-off-by: David Hildenbrand <david@xxxxxxxxxx>
---
mm/rmap.c | 6 ++++--
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/mm/rmap.c b/mm/rmap.c
index 789a2beb8b3a..f13a2927163d 100644
--- a/mm/rmap.c
+++ b/mm/rmap.c
@@ -1249,8 +1249,10 @@ void page_add_anon_rmap(struct page *page, struct vm_area_struct *vma,
}
if (flags & RMAP_EXCLUSIVE)
SetPageAnonExclusive(page);
- VM_WARN_ON_FOLIO(page_mapcount(page) > 1 && PageAnonExclusive(page),
- folio);
+ /* While PTE-mapping a THP we have a PMD and a PTE mapping. */
+ VM_WARN_ON_FOLIO((atomic_read(&page->_mapcount) > 0 ||
+ (folio_test_large(folio) && folio_entire_mapcount(folio) > 1)) &&
+ PageAnonExclusive(page), folio);
mlock_vma_folio(folio, vma, compound);
}
--
2.41.0
--
Cheers,
David / dhildenb
