[PATCH v2 8/8] mm: memory-failure: fix race window when trying to get hugetlb folio

From: Miaohe Lin
Date: Tue Jul 11 2023 - 01:50:31 EST


page_folio() is fetched before calling get_hwpoison_hugetlb_folio()
without hugetlb_lock being held. So hugetlb page could be demoted
before get_hwpoison_hugetlb_folio() holding hugetlb_lock but after
page_folio() is fetched. So get_hwpoison_hugetlb_folio() will hold
unexpected extra refcnt of hugetlb folio while leaving demoted page
un-refcnted.

Fixes: 25182f05ffed ("mm,hwpoison: fix race with hugetlb page allocation")
Signed-off-by: Miaohe Lin <linmiaohe@xxxxxxxxxx>
Acked-by: Naoya Horiguchi <naoya.horiguchi@xxxxxxx>
---
mm/memory-failure.c | 20 ++++++++++++++++----
1 file changed, 16 insertions(+), 4 deletions(-)

diff --git a/mm/memory-failure.c b/mm/memory-failure.c
index af34fd4669d3..9ab97016877e 100644
--- a/mm/memory-failure.c
+++ b/mm/memory-failure.c
@@ -1383,8 +1383,15 @@ static int __get_hwpoison_page(struct page *page, unsigned long flags)
bool hugetlb = false;

ret = get_hwpoison_hugetlb_folio(folio, &hugetlb, false);
- if (hugetlb)
- return ret;
+ if (hugetlb) {
+ /* Make sure hugetlb demotion did not happen from under us. */
+ if (folio == page_folio(page))
+ return ret;
+ if (ret > 0) {
+ folio_put(folio);
+ folio = page_folio(page);
+ }
+ }

/*
* This check prevents from calling folio_try_get() for any
@@ -1473,8 +1480,13 @@ static int __get_unpoison_page(struct page *page)
bool hugetlb = false;

ret = get_hwpoison_hugetlb_folio(folio, &hugetlb, true);
- if (hugetlb)
- return ret;
+ if (hugetlb) {
+ /* Make sure hugetlb demotion did not happen from under us. */
+ if (folio == page_folio(page))
+ return ret;
+ if (ret > 0)
+ folio_put(folio);
+ }

/*
* PageHWPoisonTakenOff pages are not only marked as PG_hwpoison,
--
2.33.0