2017-04-20 10:24:34

by Rabin Vincent

Subject: [PATCH] mm: prevent NR_ISOLATE_* stats from going negative

From: Rabin Vincent <[email protected]>

Commit 6afcf8ef0ca0 ("mm, compaction: fix NR_ISOLATED_* stats for pfn
based migration") moved the dec_node_page_state() call (along with the
page_is_file_cache() call) to after putback_lru_page(). But
page_is_file_cache() can change after putback_lru_page() is called, so
it should be called before putback_lru_page(), as it was before that
patch, to prevent NR_ISOLATE_* stats from going negative.

Without this fix, non-CONFIG_SMP kernels end up hanging in the
while(too_many_isolated()) { congestion_wait() } loop in
shrink_active_list() due to the negative stats.

Mem-Info:
active_anon:32567 inactive_anon:121 isolated_anon:1
active_file:6066 inactive_file:6639 isolated_file:4294967295
^^^^^^^^^^
unevictable:0 dirty:115 writeback:0 unstable:0
slab_reclaimable:2086 slab_unreclaimable:3167
mapped:3398 shmem:18366 pagetables:1145 bounce:0
free:1798 free_pcp:13 free_cma:0

Fixes: 6afcf8ef0ca0 ("mm, compaction: fix NR_ISOLATED_* stats for pfn based migration")
Cc: Ming Ling <[email protected]>
Cc: Michal Hocko <[email protected]>
Cc: Minchan Kim <[email protected]>
Cc: Vlastimil Babka <[email protected]>
Cc: <[email protected]>
Signed-off-by: Rabin Vincent <[email protected]>
---
mm/migrate.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/mm/migrate.c b/mm/migrate.c
index ed97c2c..738f1d5 100644
--- a/mm/migrate.c
+++ b/mm/migrate.c
@@ -184,9 +184,9 @@ void putback_movable_pages(struct list_head *l)
unlock_page(page);
put_page(page);
} else {
- putback_lru_page(page);
dec_node_page_state(page, NR_ISOLATED_ANON +
page_is_file_cache(page));
+ putback_lru_page(page);
}
}
}
--
2.7.0

2017-04-20 12:32:42

by Michal Hocko

Subject: Re: [PATCH] mm: prevent NR_ISOLATE_* stats from going negative

On Thu 20-04-17 12:24:25, Rabin Vincent wrote:
> From: Rabin Vincent <[email protected]>
>
> Commit 6afcf8ef0ca0 ("mm, compaction: fix NR_ISOLATED_* stats for pfn
> based migration") moved the dec_node_page_state() call (along with the
> page_is_file_cache() call) to after putback_lru_page(). But
> page_is_file_cache() can change after putback_lru_page() is called, so
> it should be called before putback_lru_page(), as it was before that
> patch, to prevent NR_ISOLATE_* stats from going negative.
>
> Without this fix, non-CONFIG_SMP kernels end up hanging in the
> while(too_many_isolated()) { congestion_wait() } loop in
> shrink_active_list() due to the negative stats.
>
> Mem-Info:
> active_anon:32567 inactive_anon:121 isolated_anon:1
> active_file:6066 inactive_file:6639 isolated_file:4294967295
> ^^^^^^^^^^
> unevictable:0 dirty:115 writeback:0 unstable:0
> slab_reclaimable:2086 slab_unreclaimable:3167
> mapped:3398 shmem:18366 pagetables:1145 bounce:0
> free:1798 free_pcp:13 free_cma:0
>
> Fixes: 6afcf8ef0ca0 ("mm, compaction: fix NR_ISOLATED_* stats for pfn based migration")
> Cc: Ming Ling <[email protected]>
> Cc: Michal Hocko <[email protected]>
> Cc: Minchan Kim <[email protected]>
> Cc: Vlastimil Babka <[email protected]>
> Cc: <[email protected]>
> Signed-off-by: Rabin Vincent <[email protected]>

Thanks for catching and fixing this. This is definitely my fault whe
reworking Ming's original patch

Acked-by: Michal Hocko <[email protected]>

> ---
> mm/migrate.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/mm/migrate.c b/mm/migrate.c
> index ed97c2c..738f1d5 100644
> --- a/mm/migrate.c
> +++ b/mm/migrate.c
> @@ -184,9 +184,9 @@ void putback_movable_pages(struct list_head *l)
> unlock_page(page);
> put_page(page);
> } else {
> - putback_lru_page(page);
> dec_node_page_state(page, NR_ISOLATED_ANON +
> page_is_file_cache(page));
> + putback_lru_page(page);
> }
> }
> }
> --
> 2.7.0
>

--
Michal Hocko
SUSE Labs