2020-08-11 07:55:22

by Baoquan He

[permalink] [raw]
Subject: [PATCH] Revert "mm/vmstat.c: do not show lowmem reserve protection information of empty zone"

This reverts commit 26e7deadaae1755faf1f6d1a68988c4b8348df59.

Sonny reported that one of their tests started failing on the latest
kernel on their Chrome OS platform. The root cause is that the above
commit removed the protection line of empty zone, while the parser used
in the test relies on the protection line to mark the end of each zone.

Let's revert it to avoid breaking userspace testing or applications.

Cc: <[email protected]> # 5.8.x
Reported-by: Sonny Rao <[email protected]>
Signed-off-by: Baoquan He <[email protected]>
---
mm/vmstat.c | 12 ++++++------
1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/mm/vmstat.c b/mm/vmstat.c
index 5b35c0eb8c57..62376dc66ec0 100644
--- a/mm/vmstat.c
+++ b/mm/vmstat.c
@@ -1644,12 +1644,6 @@ static void zoneinfo_show_print(struct seq_file *m, pg_data_t *pgdat,
zone->present_pages,
zone_managed_pages(zone));

- /* If unpopulated, no other information is useful */
- if (!populated_zone(zone)) {
- seq_putc(m, '\n');
- return;
- }
-
seq_printf(m,
"\n protection: (%ld",
zone->lowmem_reserve[0]);
@@ -1657,6 +1651,12 @@ static void zoneinfo_show_print(struct seq_file *m, pg_data_t *pgdat,
seq_printf(m, ", %ld", zone->lowmem_reserve[i]);
seq_putc(m, ')');

+ /* If unpopulated, no other information is useful */
+ if (!populated_zone(zone)) {
+ seq_putc(m, '\n');
+ return;
+ }
+
for (i = 0; i < NR_VM_ZONE_STAT_ITEMS; i++)
seq_printf(m, "\n %-12s %lu", zone_stat_name(i),
zone_page_state(zone, i));
--
2.17.2


2020-08-11 08:35:25

by David Hildenbrand

[permalink] [raw]
Subject: Re: [PATCH] Revert "mm/vmstat.c: do not show lowmem reserve protection information of empty zone"

On 11.08.20 09:54, Baoquan He wrote:
> This reverts commit 26e7deadaae1755faf1f6d1a68988c4b8348df59.
>
> Sonny reported that one of their tests started failing on the latest
> kernel on their Chrome OS platform. The root cause is that the above
> commit removed the protection line of empty zone, while the parser used
> in the test relies on the protection line to mark the end of each zone.
>
> Let's revert it to avoid breaking userspace testing or applications.
>
> Cc: <[email protected]> # 5.8.x
> Reported-by: Sonny Rao <[email protected]>
> Signed-off-by: Baoquan He <[email protected]>
> ---
> mm/vmstat.c | 12 ++++++------
> 1 file changed, 6 insertions(+), 6 deletions(-)
>
> diff --git a/mm/vmstat.c b/mm/vmstat.c
> index 5b35c0eb8c57..62376dc66ec0 100644
> --- a/mm/vmstat.c
> +++ b/mm/vmstat.c
> @@ -1644,12 +1644,6 @@ static void zoneinfo_show_print(struct seq_file *m, pg_data_t *pgdat,
> zone->present_pages,
> zone_managed_pages(zone));
>
> - /* If unpopulated, no other information is useful */
> - if (!populated_zone(zone)) {
> - seq_putc(m, '\n');
> - return;
> - }
> -
> seq_printf(m,
> "\n protection: (%ld",
> zone->lowmem_reserve[0]);
> @@ -1657,6 +1651,12 @@ static void zoneinfo_show_print(struct seq_file *m, pg_data_t *pgdat,
> seq_printf(m, ", %ld", zone->lowmem_reserve[i]);
> seq_putc(m, ')');
>
> + /* If unpopulated, no other information is useful */
> + if (!populated_zone(zone)) {
> + seq_putc(m, '\n');
> + return;
> + }
> +
> for (i = 0; i < NR_VM_ZONE_STAT_ITEMS; i++)
> seq_printf(m, "\n %-12s %lu", zone_stat_name(i),
> zone_page_state(zone, i));
>

Reviewed-by: David Hildenbrand <[email protected]>

--
Thanks,

David / dhildenb

2020-08-11 21:48:46

by David Rientjes

[permalink] [raw]
Subject: Re: [PATCH] Revert "mm/vmstat.c: do not show lowmem reserve protection information of empty zone"

On Tue, 11 Aug 2020, Baoquan He wrote:

> This reverts commit 26e7deadaae1755faf1f6d1a68988c4b8348df59.
>
> Sonny reported that one of their tests started failing on the latest
> kernel on their Chrome OS platform. The root cause is that the above
> commit removed the protection line of empty zone, while the parser used
> in the test relies on the protection line to mark the end of each zone.
>
> Let's revert it to avoid breaking userspace testing or applications.
>

Acked-by: David Rientjes <[email protected]>

No objection since I noted userspace parsing as a potential risk in
https://marc.info/?l=linux-mm&m=158507790824995 but I didn't catch this
earlier patch which could have led to the same thing.