Hi Andrew,
Thank you for revising this patch!
> From: "andrew.yang" <[email protected]>
>
> damon_get_folio() would always increase folio _refcount and
> folio_isolate_lru() would increase folio _refcount if the folio's lru
> flag is set.
>
> If an unevictable folio isolated successfully, there will be two more
> _refcount. The one from folio_isolate_lru() will be decreased in
> folio_puback_lru(), but the other one from damon_get_folio() will be
> left behind. This causes a pin page.
>
> Whatever the case, the _refcount from damon_get_folio() should be
> decreased.
>
> Signed-off-by: andrew.yang <[email protected]>
> Fixes: 57223ac29584 ("mm/damon/paddr: support the pageout scheme")
> Cc: <[email protected]> # 5.16.x
Reviewed-by: SeongJae Park <[email protected]>
This may not cleanly applicable on 6.1.y and 6.2.y. I would post backports
once this is merged in the mainline, unless others do earlier than me.
Thanks,
SJ
> ---
> v3:
> add fixes tag and cc stable
> v2:
> according to David's suggestion
> 1. revise subject
> according to SeongJae's suggestions
> 1. rebase to mm-unstable tree
> 2. remove braces for th single statements
> ---
> mm/damon/paddr.c | 7 +++----
> 1 file changed, 3 insertions(+), 4 deletions(-)
>
> diff --git a/mm/damon/paddr.c b/mm/damon/paddr.c
> index 607bb69e526c..6c655d9b5639 100644
> --- a/mm/damon/paddr.c
> +++ b/mm/damon/paddr.c
> @@ -250,12 +250,11 @@ static unsigned long damon_pa_pageout(struct damon_region *r, struct damos *s)
> folio_put(folio);
> continue;
> }
> - if (folio_test_unevictable(folio)) {
> + if (folio_test_unevictable(folio))
> folio_putback_lru(folio);
> - } else {
> + else
> list_add(&folio->lru, &folio_list);
> - folio_put(folio);
> - }
> + folio_put(folio);
> }
> applied = reclaim_pages(&folio_list);
> cond_resched();
> --
> 2.18.0
>
>