On Tue 06-08-19 10:19:49, Konstantin Khlebnikov wrote:

On 8/6/19 10:07 AM, Michal Hocko wrote:

quoted

On Fri 02-08-19 13:44:38, Michal Hocko wrote:
[...]

quoted

quoted

quoted

diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index ba9138a4a1de..53a35c526e43 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c

@@ -2429,8 +2429,12 @@ static int try_charge(struct mem_cgroup *memcg, gfp_t gfp_mask,
   				schedule_work(&memcg->high_work);
   				break;
   			}
-			current->memcg_nr_pages_over_high += batch;
-			set_notify_resume(current);
+			if (gfpflags_allow_blocking(gfp_mask)) {
+				reclaim_high(memcg, nr_pages, GFP_KERNEL);

ups, this should be s@GFP_KERNEL@gfp_mask@

quoted

quoted

+			} else {
+				current->memcg_nr_pages_over_high += batch;
+				set_notify_resume(current);
+			}
   			break;
   		}
   	} while ((memcg = parent_mem_cgroup(memcg)));

Should I send an official patch for this?

I prefer to keep it as is while we have no better solution.

Fine with me.

-- 
Michal Hocko
SUSE Labs