Re: [PATCH v6 1/3] mm: vmscan: ignore non-LRU-based reclaim in memcg reclaim

[Yosry Ahmed <yosryahmed@google.com>]

On Thu, Apr 13, 2023 at 3:40 AM Yosry Ahmed <yosryahmed@google.com> wrote:
>
> We keep track of different types of reclaimed pages through
> reclaim_state->reclaimed_slab, and we add them to the reported number
> of reclaimed pages.  For non-memcg reclaim, this makes sense. For memcg
> reclaim, we have no clue if those pages are charged to the memcg under
> reclaim.
>
> Slab pages are shared by different memcgs, so a freed slab page may have
> only been partially charged to the memcg under reclaim.  The same goes for
> clean file pages from pruned inodes (on highmem systems) or xfs buffer
> pages, there is no simple way to currently link them to the memcg under
> reclaim.
>
> Stop reporting those freed pages as reclaimed pages during memcg reclaim.
> This should make the return value of writing to memory.reclaim, and may
> help reduce unnecessary reclaim retries during memcg charging.  Writing to
> memory.reclaim on the root memcg is considered as cgroup_reclaim(), but
> for this case we want to include any freed pages, so use the
> global_reclaim() check instead of !cgroup_reclaim().
>
> Generally, this should make the return value of
> try_to_free_mem_cgroup_pages() more accurate. In some limited cases (e.g.
> freed a slab page that was mostly charged to the memcg under reclaim),
> the return value of try_to_free_mem_cgroup_pages() can be underestimated,
> but this should be fine. The freed pages will be uncharged anyway, and we
> can charge the memcg the next time around as we usually do memcg reclaim
> in a retry loop.
>
> Fixes: f2fe7b09a52b ("mm: memcg/slab: charge individual slab objects
> instead of pages")
>
> Signed-off-by: Yosry Ahmed <yosryahmed@google.com>
> ---
>  mm/vmscan.c | 49 ++++++++++++++++++++++++++++++++++++++++++-------
>  1 file changed, 42 insertions(+), 7 deletions(-)
>
> diff --git a/mm/vmscan.c b/mm/vmscan.c
> index 9c1c5e8b24b8..be657832be48 100644
> --- a/mm/vmscan.c
> +++ b/mm/vmscan.c
> @@ -511,6 +511,46 @@ static bool writeback_throttling_sane(struct scan_control *sc)
>  }
>  #endif
>
> +/*
> + * flush_reclaim_state(): add pages reclaimed outside of LRU-based reclaim to
> + * scan_control->nr_reclaimed.
> + */
> +static void flush_reclaim_state(struct scan_control *sc)
> +{
> +       /*
> +        * Currently, reclaim_state->reclaimed includes three types of pages
> +        * freed outside of vmscan:
> +        * (1) Slab pages.
> +        * (2) Clean file pages from pruned inodes (on highmem systems).
> +        * (3) XFS freed buffer pages.
> +        *
> +        * For all of these cases, we cannot universally link the pages to a
> +        * single memcg. For example, a memcg-aware shrinker can free one object
> +        * charged to the target memcg, causing an entire page to be freed.
> +        * If we count the entire page as reclaimed from the memcg, we end up
> +        * overestimating the reclaimed amount (potentially under-reclaiming).
> +        *
> +        * Only count such pages for global reclaim to prevent under-reclaiming
> +        * from the target memcg; preventing unnecessary retries during memcg
> +        * charging and false positives from proactive reclaim.
> +        *
> +        * For uncommon cases where the freed pages were actually mostly
> +        * charged to the target memcg, we end up underestimating the reclaimed
> +        * amount. This should be fine. The freed pages will be uncharged
> +        * anyway, even if they are not counted here properly, and we will be
> +        * able to make forward progress in charging (which is usually in a
> +        * retry loop).
> +        *
> +        * We can go one step further, and report the uncharged objcg pages in
> +        * memcg reclaim, to make reporting more accurate and reduce
> +        * underestimation, but it's probably not worth the complexity for now.
> +        */
> +       if (current->reclaim_state && global_reclaim(sc)) {
> +               sc->nr_reclaimed += current->reclaim_state->reclaimed;
> +               current->reclaim_state->reclaimed = 0;

Ugh.. this breaks the build. This should have been
current->reclaim_state->reclaimed_slab. It doesn't get renamed from
"reclaimed_slab" to "reclaim" until the next patch. When I moved
flush_reclaim_state() from patch 2 to patch 1 I forgot to augment it.
My bad.

The break is fixed by the very next patch, and the patches have
already landed in Linus's tree, so there isn't much that can be done
at this point. Sorry about that. Just wondering, why wouldn't this
breakage be caught by any of the build bots?

> +       }
> +}
> +
>  static long xchg_nr_deferred(struct shrinker *shrinker,
>                              struct shrink_control *sc)
>  {
> @@ -5346,8 +5386,7 @@ static int shrink_one(struct lruvec *lruvec, struct scan_control *sc)
>                 vmpressure(sc->gfp_mask, memcg, false, sc->nr_scanned - scanned,
>                            sc->nr_reclaimed - reclaimed);
>
> -       sc->nr_reclaimed += current->reclaim_state->reclaimed_slab;
> -       current->reclaim_state->reclaimed_slab = 0;
> +       flush_reclaim_state(sc);
>
>         return success ? MEMCG_LRU_YOUNG : 0;
>  }
> @@ -6450,7 +6489,6 @@ static void shrink_node_memcgs(pg_data_t *pgdat, struct scan_control *sc)
>
>  static void shrink_node(pg_data_t *pgdat, struct scan_control *sc)
>  {
> -       struct reclaim_state *reclaim_state = current->reclaim_state;
>         unsigned long nr_reclaimed, nr_scanned;
>         struct lruvec *target_lruvec;
>         bool reclaimable = false;
> @@ -6472,10 +6510,7 @@ static void shrink_node(pg_data_t *pgdat, struct scan_control *sc)
>
>         shrink_node_memcgs(pgdat, sc);
>
> -       if (reclaim_state) {
> -               sc->nr_reclaimed += reclaim_state->reclaimed_slab;
> -               reclaim_state->reclaimed_slab = 0;
> -       }
> +       flush_reclaim_state(sc);
>
>         /* Record the subtree's reclaim efficiency */
>         if (!sc->proactive)
> --
> 2.40.0.577.gac1e443424-goog
>