[BUG ?] oom with empty nodes.

[KAMEZAWA Hiroyuki]

On some system, there are memory-less-nodes. (IOW, cpu-only-node)
Then,there are online nodes which has no memory.

Now, below code is used to detect the context where oom happens.

===
static inline int constrained_alloc(struct zonelist *zonelist, gfp_t gfp_mask)
{
#ifdef CONFIG_NUMA
        struct zone **z;
        nodemask_t nodes = node_online_map;

        for (z = zonelist->zones; *z; z++)
                if (cpuset_zone_allowed_softwall(*z, gfp_mask))
                        node_clear(zone_to_nid(*z), nodes);
                else
                        return CONSTRAINT_CPUSET;

        if (!nodes_empty(nodes))
                return CONSTRAINT_MEMORY_POLICY;
#endif

        return CONSTRAINT_NONE;
}
==
Because the zonelist never incldues memory-less-zone, above nodemask cannot be
empty if there is a memoly-less-node, never returns CONSTRAINT_NONE.

It looks select_bad_process() and panic_on_oom is checked only when
above function returns CONSTRAINT_NONE. This means current dies every time.

Does anyone have good idea for fix this ?

-Kame


 

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/