fix OOM killing processes wrongly thought MPOL_BIND
authorHugh Dickins <[email protected]>
Mon, 23 Apr 2007 21:41:02 +0000 (14:41 -0700)
committerLinus Torvalds <[email protected]>
Tue, 24 Apr 2007 15:23:07 +0000 (08:23 -0700)
I only have CONFIG_NUMA=y for build testing: surprised when trying a memhog
to see lots of other processes killed with "No available memory
(MPOL_BIND)".  memhog is killed correctly once we initialize nodemask in
constrained_alloc().

Signed-off-by: Hugh Dickins <[email protected]>
Acked-by: Christoph Lameter <[email protected]>
Acked-by: William Irwin <[email protected]>
Acked-by: KAMEZAWA Hiroyuki <[email protected]>
Cc: <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
mm/oom_kill.c

index af981b645a6904affee5320ef77869e10b8db278..3791edfffeebe9595f54ce7d04fe424832aa6af5 100644 (file)
@@ -176,6 +176,8 @@ static inline int constrained_alloc(struct zonelist *zonelist, gfp_t gfp_mask)
        struct zone **z;
        nodemask_t nodes;
        int node;
+
+       nodes_clear(nodes);
        /* node has memory ? */
        for_each_online_node(node)
                if (NODE_DATA(node)->node_present_pages)