Re: [PATCH 13/13] mm: memcontrol: hook up vmpressure to socket pressure
From: Vladimir Davydov <hidden>
Date: 2015-11-30 11:36:48
Also in:
cgroups, linux-mm, lkml
On Tue, Nov 24, 2015 at 04:59:40PM -0500, Johannes Weiner wrote: ...
quoted hunk ↗ jump to hunk
@@ -2396,6 +2396,7 @@ static bool shrink_zone(struct zone *zone, struct scan_control *sc, memcg = mem_cgroup_iter(root, NULL, &reclaim); do { unsigned long lru_pages; + unsigned long reclaimed; unsigned long scanned; struct lruvec *lruvec; int swappiness;@@ -2408,6 +2409,7 @@ static bool shrink_zone(struct zone *zone, struct scan_control *sc, lruvec = mem_cgroup_zone_lruvec(zone, memcg); swappiness = mem_cgroup_swappiness(memcg); + reclaimed = sc->nr_reclaimed; scanned = sc->nr_scanned; shrink_lruvec(lruvec, swappiness, sc, &lru_pages);@@ -2418,6 +2420,11 @@ static bool shrink_zone(struct zone *zone, struct scan_control *sc, memcg, sc->nr_scanned - scanned, lru_pages); + /* Record the group's reclaim efficiency */ + vmpressure(sc->gfp_mask, memcg, false, + sc->nr_scanned - scanned, + sc->nr_reclaimed - reclaimed); +
Suppose we have the following cgroup configuration. A __ B \_ C A is empty (which is natural for the unified hierarchy AFAIU). B has some workload running in it, and C generates socket pressure. Due to the socket pressure coming from C we start reclaim in A, which results in thrashing of B, but we might not put sockets under pressure in A or C, because vmpressure does not account pages scanned/reclaimed in B when generating a vmpressure event for A or C. This might result in aggressive reclaim and thrashing in B w/o generating a signal for C to stop growing socket buffers. Do you think such a situation is possible? If so, would it make sense to switch to post-order walk in shrink_zone and pass sub-tree scanned/reclaimed stats to vmpressure for each scanned memcg? Thanks, Vladimir
quoted hunk ↗ jump to hunk
/* * Direct reclaim and kswapd have to scan all memory * cgroups to fulfill the overall scan target for the@@ -2449,7 +2456,8 @@ static bool shrink_zone(struct zone *zone, struct scan_control *sc, reclaim_state->reclaimed_slab = 0; } - vmpressure(sc->gfp_mask, sc->target_mem_cgroup, + /* Record the subtree's reclaim efficiency */ + vmpressure(sc->gfp_mask, sc->target_mem_cgroup, true, sc->nr_scanned - nr_scanned, sc->nr_reclaimed - nr_reclaimed);