[PATCH] mm, oom: consider present pages for the node size

[Michal Hocko]

From: Michal Hocko <mhocko@xxxxxxxx>

constrained_alloc calculates the size of the oom domain by using
node_spanned_pages which is incorrect because this is the full range of
the physical memory range that the numa node occupies rather than the
memory that backs that range which is represented by node_present_pages.

Sparsely populated nodes (e.g. after memory hot remove or simply sparse
due to memory layout) can have really a large difference between the
two. This shouldn't really cause any real user observable problems
because the oom calculates a ratio against totalpages and used memory
cannot exceed present pages but it is confusing and wrong from code
point of view.

Noticed-by: David Hildenbrand <david@xxxxxxxxxx>
Signed-off-by: Michal Hocko <mhocko@xxxxxxxx>
---
 mm/oom_kill.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/mm/oom_kill.c b/mm/oom_kill.c
index eda2e2a0bdc6..16af3da97d08 100644
--- a/mm/oom_kill.c
+++ b/mm/oom_kill.c
@@ -287,7 +287,7 @@ static enum oom_constraint constrained_alloc(struct oom_control *oc)
 	    !nodes_subset(node_states[N_MEMORY], *oc->nodemask)) {
 		oc->totalpages = total_swap_pages;
 		for_each_node_mask(nid, *oc->nodemask)
-			oc->totalpages += node_spanned_pages(nid);
+			oc->totalpages += node_present_pages(nid);
 		return CONSTRAINT_MEMORY_POLICY;
 	}
 
@@ -300,7 +300,7 @@ static enum oom_constraint constrained_alloc(struct oom_control *oc)
 	if (cpuset_limited) {
 		oc->totalpages = total_swap_pages;
 		for_each_node_mask(nid, cpuset_current_mems_allowed)
-			oc->totalpages += node_spanned_pages(nid);
+			oc->totalpages += node_present_pages(nid);
 		return CONSTRAINT_CPUSET;
 	}
 	return CONSTRAINT_NONE;
-- 
2.20.1