On Tue, Jan 12, 2021 at 5:42 PM Michal Hocko [off-list ref] wrote:

On Sun 10-01-21 20:40:12, Muchun Song wrote:

quoted

If the refcount is one when it is migrated, it means that the page
was freed from under us. So we are done and do not need to migrate.

I would consider the following easier to understand. Feel free to reuse.
"
All pages isolated for the migration have an elevated reference count
and therefore seeing a reference count equal to 1 means that the last
user of the page has dropped the reference and the page has became
unused and there doesn't make much sense to migrate it anymore. This has
been done for regular pages and this patch does the same for hugetlb
pages. Although the likelyhood of the race is rather small for hugetlb
pages it makes sense the two code paths in sync.
"

Thanks.

quoted

This optimization is consistent with the regular pages, just like
unmap_and_move() does.

Signed-off-by: Muchun Song <redacted>
Reviewed-by: Mike Kravetz <redacted>
Acked-by: Yang Shi <redacted>

Acked-by: Michal Hocko <mhocko@suse.com>

Thanks.

quoted

---
 mm/migrate.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/mm/migrate.c b/mm/migrate.c
index 4385f2fb5d18..a6631c4eb6a6 100644
--- a/mm/migrate.c
+++ b/mm/migrate.c

@@ -1279,6 +1279,12 @@ static int unmap_and_move_huge_page(new_page_t get_new_page,
              return -ENOSYS;
      }

+     if (page_count(hpage) == 1) {
+             /* page was freed from under us. So we are done. */
+             putback_active_hugepage(hpage);
+             return MIGRATEPAGE_SUCCESS;
+     }
+
      new_hpage = get_new_page(hpage, private);
      if (!new_hpage)
              return -ENOMEM;
--

2.11.0

--
Michal Hocko
SUSE Labs