Skip to contentSkip to conversation

mmhugetlb-take-hugetlb_lock-before-decrementing-h-resv_huge_pages.patch removed from -mm tree

merged mm-hotfixes-stable Andrew Morton 1 participant 1 message Oct 21, 2022 Stable branch

lore
Share & tools
lore

Andrew Morton

author tree notice Applied
#0 lore raw

The quilt patch titled

     Subject: mm,hugetlb: take hugetlb_lock before decrementing h->resv_huge_pages

has been removed from the -mm tree. Its filename was
mmhugetlb-take-hugetlb_lock-before-decrementing-h-resv_huge_pages.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: Rik van Riel <[email protected]>
Subject: mm,hugetlb: take hugetlb_lock before decrementing h->resv_huge_pages
Date: Mon, 17 Oct 2022 20:25:05 -0400

The h->*_huge_pages counters are protected by the hugetlb_lock, but alloc_huge_page has a corner case where it can decrement the counter outside of the lock.

This could lead to a corrupted value of h->resv_huge_pages, which we have observed on our systems.

Take the hugetlb_lock before decrementing h->resv_huge_pages to avoid a potential race.

Fixes a88c76954804 ("mm: hugetlb: fix hugepage memory leak caused by wrong reserve count")
Cc Naoya Horiguchi <[email protected]>
Cc Glen McCready <[email protected]>
Cc Mike Kravetz <[email protected]>
Cc Muchun Song <[email protected]>

---

 mm/hugetlb.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
mm/hugetlb.c renamed from mm/hugetlb.c~mmhugetlb-take-hugetlb_lock-before-decrementing-h-resv_huge_pages +1 -1
@@ -2924,11 +2924,11 @@ struct page *alloc_huge_page(struct vm_a
 		page = alloc_buddy_huge_page_with_mpol(h, vma, addr);
 		if (!page)
 			goto out_uncharge_cgroup;
+		spin_lock_irq(&hugetlb_lock);
 		if (!avoid_reserve && vma_has_reserves(vma, gbl_chg)) {
 			SetHPageRestoreReserve(page);
 			h->resv_huge_pages--;
 		}
-		spin_lock_irq(&hugetlb_lock);
 		list_add(&page->lru, &h->hugepage_activelist);
 		set_page_refcounted(page);
 		/* Fall through */

_

Patches currently in -mm which might be from [email protected] are