Re: [PATCH] mm-vmscan-dont-mess-with-pgdat-flags-in-memcg-reclaim-v2-fix

[Shakeel Butt]

On Fri, Apr 6, 2018 at 8:09 AM, Andrey Ryabinin <aryabinin@xxxxxxxxxxxxx> wrote:
>
>
> On 04/06/2018 05:37 PM, Shakeel Butt wrote:
>
>>>
>>> @@ -2482,7 +2494,7 @@ static inline bool should_continue_reclaim(struct pglist_data *pgdat,
>>>  static bool pgdat_memcg_congested(pg_data_t *pgdat, struct mem_cgroup *memcg)
>>>  {
>>>         return test_bit(PGDAT_CONGESTED, &pgdat->flags) ||
>>> -               (memcg && test_memcg_bit(PGDAT_CONGESTED, memcg));
>>> +               (memcg && memcg_congested(pgdat, memcg));
>>
>> I am wondering if we should check all ancestors for congestion as
>> well. Maybe a parallel memcg reclaimer might have set some ancestor of
>> this memcg to congested.
>>
>
> Why? If ancestor is congested but its child (the one we currently reclaim) is not,
> it could mean only 2 things:
>  - Either child use mostly anon and inactive file lru is small (file_lru >> priority == 0)
>    so it's not congested.
>  - Or the child was congested recently (at the time when ancestor scanned this group),
>    but not anymore. So the information from ancestor is simply outdated.
>

Oh yeah, you explained in the other email as well. Thanks.

I think Andrew will squash this patch with the previous one. Andrew,
please add following in the squashed patch.

Reviewed-by: Shakeel Butt <shakeelb@xxxxxxxxxx>