Re: [PATCH v2] mm/mglru: fix a no-op lru_gen_clear_refs() call in lru_lazyfree()

[Kairui Song]

On Mon, Oct 26, 2026 at 4:41 PM Hui Zhu <hui.zhu@xxxxxxxxx> wrote:
>
> From: Hui Zhu <zhuhui@xxxxxxxxxx>
>
> lru_lazyfree() calls lru_gen_clear_refs() after lruvec_del_folio(), but
> lru_gen_del_folio() has already cleared LRU_GEN_MASK, so folio_lru_gen()
> returns -1 and lru_gen_clear_refs() exits at its "gen < 0" check without
> clearing the refs or PG_workingset.
>
> A MADV_FREE folio therefore keeps the state it had before madvise(2),
> even though madvise_free_pte_range() has already cleared the accessed and
> dirty bits of its PTEs.  lru_gen_folio_seq() places it one generation
> younger, and PG_workingset puts it in the last tier, so sort_folio()
> promotes and protects it instead of evicting it and charges it to
> lrugen->protected[], which read_ctrl_pos() uses to balance anon against
> file reclaim.  __folio_inc_gen() clears the refs but not PG_workingset,
> so this repeats every time the folio reaches the oldest generation.  If
> the folio is redirtied first, ttu_anon_lazyfree_folio() keeps it and
> lru_gen_eviction() packs the stale state into the shadow entry, so a
> later refault is promoted to the youngest generation.
>
> Call lru_gen_clear_refs() while the folio is still on the generation
> list, as deactivate_file_folio() and folio_deactivate() already do.  Its
> return value only says whether the folio is in the oldest generation,
> which lru_lazyfree() does not need because it re-adds the folio below.
> Keep folio_clear_referenced() unconditional: the classic LRU needs it,
> and it also covers the folios that lru_gen_clear_refs() skips.
>
> Fixes: cc8ec7be78ff ("mm/mglru: optimize deactivation")
> Suggested-by: Barry Song <baohua@xxxxxxxxxx>
> Signed-off-by: Hui Zhu <zhuhui@xxxxxxxxxx>
> ---
> Changelog:
> v2:
> Call lru_gen_clear_refs() before lruvec_del_folio() instead of
> open-coding the flag clears; drop the lru_gen_enabled() branch and
> the comment (Barry Song).
>
>  mm/folio.c | 6 ++----
>  1 file changed, 2 insertions(+), 4 deletions(-)
>
> diff --git a/mm/folio.c b/mm/folio.c
> index 35e242b48870..ea3adb6a89d7 100644
> --- a/mm/folio.c
> +++ b/mm/folio.c
> @@ -597,12 +597,10 @@ static void lru_lazyfree(struct lruvec *lruvec, struct folio *folio)
>             folio_test_swapcache(folio) || folio_test_unevictable(folio))
>                 return;
>
> +       lru_gen_clear_refs(folio);
>         lruvec_del_folio(lruvec, folio);
>         folio_clear_active(folio);
> -       if (lru_gen_enabled())
> -               lru_gen_clear_refs(folio);
> -       else
> -               folio_clear_referenced(folio);
> +       folio_clear_referenced(folio);

Why can't we just call folio_set_lru_refs(.., 0), and should we clear
PG_workingset? Keeping it helps PSI on refualt / redirty tracking (not
sure if meaningful but CLRU does this), clearing it makes eviction
better, but the eviction issue is MGLRU-only, CLRU keeps it.

And folio_set_lru_refs(.., 0) should be a little bit faster and unify
CLRU / MGLRU. And we can detect lazy freed folios in sort_folios to
avoid PID form touching them?

Something like:
diff --git a/mm/folio.c b/mm/folio.c
index 35e242b48870..416cd4beb182 100644
--- a/mm/folio.c
+++ b/mm/folio.c
@@ -599,10 +599,7 @@ static void lru_lazyfree(struct lruvec *lruvec,
struct folio *folio)

        lruvec_del_folio(lruvec, folio);
        folio_clear_active(folio);
-       if (lru_gen_enabled())
-               lru_gen_clear_refs(folio);
-       else
-               folio_clear_referenced(folio);
+       folio_set_lru_refs(folio, 0);
        /*
         * Lazyfree folios are clean anonymous folios.  They have
         * the swapbacked flag cleared, to distinguish them from normal
diff --git a/mm/vmscan.c b/mm/vmscan.c
index 6a1af9b92349..e64b7f544d15 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -4835,6 +4835,10 @@ static bool sort_folio(struct lruvec *lruvec,
struct folio *folio, struct scan_c
        if (!folio_evictable(folio))
                return false;

+       /* No need to protect lazyfreed folios */
+       if (folio_test_lazyfree(folio) && !folio_test_dirty(folio))
+               return false;
+
        /* promoted */
        if (gen != lru_gen_from_seq(lrugen->min_seq[type])) {
                list_move(&folio->lru, &lrugen->folios[gen][type][zone]);

---

BTW I also tried to cover this in MGLRU-FG series (not including the
sort_folio part):
https://lore.kernel.org/linux-mm/20261016-mglru-fg-v4-13-2106750c0bfa@xxxxxxxxxxx/

I tested keeping min(refs, LRU_REFS_WORKINGSET) / just set to 0 as in
v4, nothing different from the workload (testing using Android apps as
workload), so I think it's fine either way but the above patch I
posted might help to impove it by a little?

And if CLRU didn't actually expect PG_workingset to be kept but that's
just a leftover or trivial, maybe we can always do both
folio_clear_workingset and folio_set_lru_refs(0) here?