There is a race between isolate_huge_page() and __free_huge_page().

CPU0: CPU1:

if (PageHuge(page)) put_page(page) __free_huge_page(page) spin_lock(&hugetlb_lock) update_and_free_page(page) set_compound_page_dtor(page, NULL_COMPOUND_DTOR) spin_unlock(&hugetlb_lock) isolate_huge_page(page) // trigger BUG_ON VM_BUG_ON_PAGE(!PageHead(page), page) spin_lock(&hugetlb_lock) page_huge_active(page) // trigger BUG_ON VM_BUG_ON_PAGE(!PageHuge(page), page) spin_unlock(&hugetlb_lock)

When we isolate a HugeTLB page on CPU0. Meanwhile, we free it to the buddy allocator on CPU1. Then, we can trigger a BUG_ON on CPU0. Because it is already freed to the buddy allocator.

Fixes: c8721bbbdd36 ("mm: memory-hotplug: enable memory hotplug to handle hugepage") Signed-off-by: Muchun Song songmuchun@bytedance.com Reviewed-by: Mike Kravetz mike.kravetz@oracle.com Acked-by: Michal Hocko mhocko@suse.com Cc: stable@vger.kernel.org --- mm/hugetlb.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/mm/hugetlb.c b/mm/hugetlb.c index a176ceed55f1..e7ed30afbb8f 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -5575,9 +5575,9 @@ bool isolate_huge_page(struct page *page, struct list_head *list) { bool ret = true;

- VM_BUG_ON_PAGE(!PageHead(page), page); spin_lock(&hugetlb_lock); - if (!page_huge_active(page) || !get_page_unless_zero(page)) { + if (!PageHeadHuge(page) || !page_huge_active(page) || + !get_page_unless_zero(page)) { ret = false; goto unlock; }