When zspage is migrated to the other zone, the zone page state should
be updated as well.
Signed-off-by: Chanho Min <[email protected]>
Signed-off-by: Jinsuk Choi <[email protected]>
---
mm/zsmalloc.c | 5 +++++
1 file changed, 5 insertions(+)
diff --git a/mm/zsmalloc.c b/mm/zsmalloc.c
index 2b2b9aa..22d17ec 100644
--- a/mm/zsmalloc.c
+++ b/mm/zsmalloc.c
@@ -2069,6 +2069,11 @@ static int zs_page_migrate(struct address_space *mapping, struct page *newpage,
zs_pool_dec_isolated(pool);
}
+ if (page_zone(newpage) != page_zone(page)) {
+ dec_zone_page_state(page, NR_ZSPAGES);
+ inc_zone_page_state(newpage, NR_ZSPAGES);
+ }
+
reset_page(page);
put_page(page);
page = newpage;
--
2.7.4
On 11/29/19 2:29 AM, Chanho Min wrote:
> When zspage is migrated to the other zone, the zone page state should
> be updated as well.
What are the user visible effects? I assume NR_ZSPAGES accounting can go
wrong otherwise? Has it been observed in practice?
Should we Cc stable and identify a Fixes: commit?
Thanks,
Vlastimil
> Signed-off-by: Chanho Min <[email protected]>
> Signed-off-by: Jinsuk Choi <[email protected]>
> ---
> mm/zsmalloc.c | 5 +++++
> 1 file changed, 5 insertions(+)
>
> diff --git a/mm/zsmalloc.c b/mm/zsmalloc.c
> index 2b2b9aa..22d17ec 100644
> --- a/mm/zsmalloc.c
> +++ b/mm/zsmalloc.c
> @@ -2069,6 +2069,11 @@ static int zs_page_migrate(struct address_space *mapping, struct page *newpage,
> zs_pool_dec_isolated(pool);
> }
>
> + if (page_zone(newpage) != page_zone(page)) {
> + dec_zone_page_state(page, NR_ZSPAGES);
> + inc_zone_page_state(newpage, NR_ZSPAGES);
> + }
> +
> reset_page(page);
> put_page(page);
> page = newpage;
>
On (19/11/29 10:29), Chanho Min wrote:
> When zspage is migrated to the other zone, the zone page state should
> be updated as well.
>
> Signed-off-by: Chanho Min <[email protected]>
> Signed-off-by: Jinsuk Choi <[email protected]>
Looks good to me.
Reviewed-by: Sergey Senozhatsky <[email protected]>
-ss
On Fri, Nov 29, 2019 at 10:29:27AM +0900, Chanho Min wrote:
> When zspage is migrated to the other zone, the zone page state should
> be updated as well.
>
> Signed-off-by: Chanho Min <[email protected]>
> Signed-off-by: Jinsuk Choi <[email protected]>
Hi Chanho,
It seems your previous mail included fix and stable tag was mangled by
HTML mail. Could you resend the patch it unless Andrew pick up?
Acked-by: Minchan Kim <[email protected]>
Thanks!