mm: have zonelist contains structs with both a zone pointer and zone_idx

Filtering zonelists requires very frequent use of zone_idx().  This is costly
as it involves a lookup of another structure and a substraction operation.  As
the zone_idx is often required, it should be quickly accessible.  The node idx
could also be stored here if it was found that accessing zone->node is
significant which may be the case on workloads where nodemasks are heavily
used.

This patch introduces a struct zoneref to store a zone pointer and a zone
index.  The zonelist then consists of an array of these struct zonerefs which
are looked up as necessary.  Helpers are given for accessing the zone index as
well as the node index.

[[email protected]: Suggested struct zoneref instead of embedding information in pointers]
[[email protected]: mm-have-zonelist: fix memcg ooms]
[[email protected]: just return do_try_to_free_pages]
[[email protected]: do_try_to_free_pages gfp_mask redundant]
Signed-off-by: Mel Gorman <[email protected]>
Acked-by: Christoph Lameter <[email protected]>
Acked-by: David Rientjes <[email protected]>
Signed-off-by: Lee Schermerhorn <[email protected]>
Cc: KAMEZAWA Hiroyuki <[email protected]>
Cc: Mel Gorman <[email protected]>
Cc: Christoph Lameter <[email protected]>
Cc: Nick Piggin <[email protected]>
Signed-off-by: Hugh Dickins <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>

arch/parisc/mm/init.c

@@ -608,7 +608,7 @@ void show_mem(void)
 		for (i = 0; i < npmem_ranges; i++) {
 			zl = node_zonelist(i);
 			for (j = 0; j < MAX_NR_ZONES; j++) {
-				struct zone **z;
+				struct zoneref *z;
 				struct zone *zone;
 
 				printk("Zone list for zone %d on node %d: ", j, i);

fs/buffer.c

@@ -360,16 +360,16 @@ void invalidate_bdev(struct block_device *bdev)
  */
 static void free_more_memory(void)
 {
-	struct zone **zones;
+	struct zoneref *zrefs;
 	int nid;
 
 	wakeup_pdflush(1024);
 	yield();
 
 	for_each_online_node(nid) {
-		zones = first_zones_zonelist(node_zonelist(nid, GFP_NOFS),
+		zrefs = first_zones_zonelist(node_zonelist(nid, GFP_NOFS),
 						gfp_zone(GFP_NOFS));
-		if (*zones)
+		if (zrefs->zone)
 			try_to_free_pages(node_zonelist(nid, GFP_NOFS), 0,
 						GFP_NOFS);
 	}

include/linux/mmzone.h

@@ -468,6 +468,15 @@ struct zonelist_cache {
 struct zonelist_cache;
 #endif
 
+/*
+ * This struct contains information about a zone in a zonelist. It is stored
+ * here to avoid dereferences into large structures and lookups of tables
+ */
+struct zoneref {
+	struct zone *zone;	/* Pointer to actual zone */
+	int zone_idx;		/* zone_idx(zoneref->zone) */
+};
+
 /*
  * One allocation request operates on a zonelist. A zonelist
  * is a list of zones, the first one is the 'goal' of the
@@ -476,11 +485,18 @@ struct zonelist_cache;
  *
  * If zlcache_ptr is not NULL, then it is just the address of zlcache,
  * as explained above.  If zlcache_ptr is NULL, there is no zlcache.
+ * *
+ * To speed the reading of the zonelist, the zonerefs contain the zone index
+ * of the entry being read. Helper functions to access information given
+ * a struct zoneref are
+ *
+ * zonelist_zone()	- Return the struct zone * for an entry in _zonerefs
+ * zonelist_zone_idx()	- Return the index of the zone for an entry
+ * zonelist_node_idx()	- Return the index of the node for an entry
  */
-
 struct zonelist {
 	struct zonelist_cache *zlcache_ptr;		     // NULL or &zlcache
-	struct zone *zones[MAX_ZONES_PER_ZONELIST + 1];      // NULL delimited
+	struct zoneref _zonerefs[MAX_ZONES_PER_ZONELIST + 1];
 #ifdef CONFIG_NUMA
 	struct zonelist_cache zlcache;			     // optional ...
 #endif
@@ -713,26 +729,52 @@ extern struct zone *next_zone(struct zone *zone);
 	     zone;					\
 	     zone = next_zone(zone))
 
+static inline struct zone *zonelist_zone(struct zoneref *zoneref)
+{
+	return zoneref->zone;
+}
+
+static inline int zonelist_zone_idx(struct zoneref *zoneref)
+{
+	return zoneref->zone_idx;
+}
+
+static inline int zonelist_node_idx(struct zoneref *zoneref)
+{
+#ifdef CONFIG_NUMA
+	/* zone_to_nid not available in this context */
+	return zoneref->zone->node;
+#else
+	return 0;
+#endif /* CONFIG_NUMA */
+}
+
+static inline void zoneref_set_zone(struct zone *zone, struct zoneref *zoneref)
+{
+	zoneref->zone = zone;
+	zoneref->zone_idx = zone_idx(zone);
+}
+
 /* Returns the first zone at or below highest_zoneidx in a zonelist */
-static inline struct zone **first_zones_zonelist(struct zonelist *zonelist,
+static inline struct zoneref *first_zones_zonelist(struct zonelist *zonelist,
 					enum zone_type highest_zoneidx)
 {
-	struct zone **z;
+	struct zoneref *z;
 
 	/* Find the first suitable zone to use for the allocation */
-	z = zonelist->zones;
-	while (*z && zone_idx(*z) > highest_zoneidx)
+	z = zonelist->_zonerefs;
+	while (zonelist_zone_idx(z) > highest_zoneidx)
 		z++;
 
 	return z;
 }
 
 /* Returns the next zone at or below highest_zoneidx in a zonelist */
-static inline struct zone **next_zones_zonelist(struct zone **z,
+static inline struct zoneref *next_zones_zonelist(struct zoneref *z,
 					enum zone_type highest_zoneidx)
 {
 	/* Find the next suitable zone to use for the allocation */
-	while (*z && zone_idx(*z) > highest_zoneidx)
+	while (zonelist_zone_idx(z) > highest_zoneidx)
 		z++;
 
 	return z;
@@ -748,9 +790,11 @@ static inline struct zone **next_zones_zonelist(struct zone **z,
  * This iterator iterates though all zones at or below a given zone index.
  */
 #define for_each_zone_zonelist(zone, z, zlist, highidx) \
-	for (z = first_zones_zonelist(zlist, highidx), zone = *z++;	\
+	for (z = first_zones_zonelist(zlist, highidx),			\
+					zone = zonelist_zone(z++);	\
 		zone;							\
-		z = next_zones_zonelist(z, highidx), zone = *z++)
+		z = next_zones_zonelist(z, highidx),			\
+					zone = zonelist_zone(z++))
 
 #ifdef CONFIG_SPARSEMEM
 #include <asm/sparsemem.h>

include/linux/oom.h

@@ -23,8 +23,8 @@ enum oom_constraint {
 	CONSTRAINT_MEMORY_POLICY,
 };
 
-extern int try_set_zone_oom(struct zonelist *zonelist);
-extern void clear_zonelist_oom(struct zonelist *zonelist);
+extern int try_set_zone_oom(struct zonelist *zonelist, gfp_t gfp_flags);
+extern void clear_zonelist_oom(struct zonelist *zonelist, gfp_t gfp_flags);
 
 extern void out_of_memory(struct zonelist *zonelist, gfp_t gfp_mask, int order);
 extern int register_oom_notifier(struct notifier_block *nb);

kernel/cpuset.c

@@ -1967,8 +1967,8 @@ int cpuset_zonelist_valid_mems_allowed(struct zonelist *zl)
 {
 	int i;
 
-	for (i = 0; zl->zones[i]; i++) {
-		int nid = zone_to_nid(zl->zones[i]);
+	for (i = 0; zl->_zonerefs[i].zone; i++) {
+		int nid = zonelist_node_idx(&zl->_zonerefs[i]);
 
 		if (node_isset(nid, current->mems_allowed))
 			return 1;

mm/hugetlb.c

@@ -97,7 +97,8 @@ static struct page *dequeue_huge_page_vma(struct vm_area_struct *vma,
 	struct mempolicy *mpol;
 	struct zonelist *zonelist = huge_zonelist(vma, address,
 					htlb_alloc_mask, &mpol);
-	struct zone *zone, **z;
+	struct zone *zone;
+	struct zoneref *z;
 
 	for_each_zone_zonelist(zone, z, zonelist, MAX_NR_ZONES - 1) {
 		nid = zone_to_nid(zone);

mm/mempolicy.c

@@ -186,7 +186,7 @@ static struct zonelist *bind_zonelist(nodemask_t *nodes)
 		for_each_node_mask(nd, *nodes) { 
 			struct zone *z = &NODE_DATA(nd)->node_zones[k];
 			if (z->present_pages > 0) 
-				zl->zones[num++] = z;
+				zoneref_set_zone(z, &zl->_zonerefs[num++]);
 		}
 		if (k == 0)
 			break;
@@ -196,7 +196,8 @@ static struct zonelist *bind_zonelist(nodemask_t *nodes)
 		kfree(zl);
 		return ERR_PTR(-EINVAL);
 	}
-	zl->zones[num] = NULL;
+	zl->_zonerefs[num].zone = NULL;
+	zl->_zonerefs[num].zone_idx = 0;
 	return zl;
 }
 
@@ -504,9 +505,11 @@ static void get_zonemask(struct mempolicy *p, nodemask_t *nodes)
 	nodes_clear(*nodes);
 	switch (p->policy) {
 	case MPOL_BIND:
-		for (i = 0; p->v.zonelist->zones[i]; i++)
-			node_set(zone_to_nid(p->v.zonelist->zones[i]),
-				*nodes);
+		for (i = 0; p->v.zonelist->_zonerefs[i].zone; i++) {
+			struct zoneref *zref;
+			zref = &p->v.zonelist->_zonerefs[i];
+			node_set(zonelist_node_idx(zref), *nodes);
+		}
 		break;
 	case MPOL_DEFAULT:
 		break;
@@ -1212,12 +1215,13 @@ unsigned slab_node(struct mempolicy *policy)
 	case MPOL_INTERLEAVE:
 		return interleave_nodes(policy);
 
-	case MPOL_BIND:
+	case MPOL_BIND: {
 		/*
 		 * Follow bind policy behavior and start allocation at the
 		 * first node.
 		 */
-		return zone_to_nid(policy->v.zonelist->zones[0]);
+		return zonelist_node_idx(policy->v.zonelist->_zonerefs);
+	}
 
 	case MPOL_PREFERRED:
 		if (policy->v.preferred_node >= 0)
@@ -1323,7 +1327,7 @@ static struct page *alloc_page_interleave(gfp_t gfp, unsigned order,
 
 	zl = node_zonelist(nid, gfp);
 	page = __alloc_pages(gfp, order, zl);
-	if (page && page_zone(page) == zl->zones[0])
+	if (page && page_zone(page) == zonelist_zone(&zl->_zonerefs[0]))
 		inc_zone_page_state(page, NUMA_INTERLEAVE_HIT);
 	return page;
 }
@@ -1463,10 +1467,14 @@ int __mpol_equal(struct mempolicy *a, struct mempolicy *b)
 		return a->v.preferred_node == b->v.preferred_node;
 	case MPOL_BIND: {
 		int i;
-		for (i = 0; a->v.zonelist->zones[i]; i++)
-			if (a->v.zonelist->zones[i] != b->v.zonelist->zones[i])
+		for (i = 0; a->v.zonelist->_zonerefs[i].zone; i++) {
+			struct zone *za, *zb;
+			za = zonelist_zone(&a->v.zonelist->_zonerefs[i]);
+			zb = zonelist_zone(&b->v.zonelist->_zonerefs[i]);
+			if (za != zb)
 				return 0;
-		return b->v.zonelist->zones[i] == NULL;
+		}
+		return b->v.zonelist->_zonerefs[i].zone == NULL;
 	}
 	default:
 		BUG();
@@ -1785,12 +1793,12 @@ static void mpol_rebind_policy(struct mempolicy *pol,
 		break;
 	case MPOL_BIND: {
 		nodemask_t nodes;
-		struct zone **z;
+		struct zoneref *z;
 		struct zonelist *zonelist;
 
 		nodes_clear(nodes);
-		for (z = pol->v.zonelist->zones; *z; z++)
-			node_set(zone_to_nid(*z), nodes);
+		for (z = pol->v.zonelist->_zonerefs; z->zone; z++)
+			node_set(zonelist_node_idx(z), nodes);
 		nodes_remap(tmp, nodes, *mpolmask, *newmask);
 		nodes = tmp;
 

mm/oom_kill.c

@@ -176,7 +176,7 @@ static inline enum oom_constraint constrained_alloc(struct zonelist *zonelist,
 {
 #ifdef CONFIG_NUMA
 	struct zone *zone;
-	struct zone **z;
+	struct zoneref *z;
 	enum zone_type high_zoneidx = gfp_zone(gfp_mask);
 	nodemask_t nodes = node_states[N_HIGH_MEMORY];
 
@@ -462,29 +462,29 @@ EXPORT_SYMBOL_GPL(unregister_oom_notifier);
  * if a parallel OOM killing is already taking place that includes a zone in
  * the zonelist.  Otherwise, locks all zones in the zonelist and returns 1.
  */
-int try_set_zone_oom(struct zonelist *zonelist)
+int try_set_zone_oom(struct zonelist *zonelist, gfp_t gfp_mask)
 {
-	struct zone **z;
+	struct zoneref *z;
+	struct zone *zone;
 	int ret = 1;
 
-	z = zonelist->zones;
-
 	spin_lock(&zone_scan_mutex);
-	do {
-		if (zone_is_oom_locked(*z)) {
+	for_each_zone_zonelist(zone, z, zonelist, gfp_zone(gfp_mask)) {
+		if (zone_is_oom_locked(zone)) {
 			ret = 0;
 			goto out;
 		}
-	} while (*(++z) != NULL);
+	}
+
+	for_each_zone_zonelist(zone, z, zonelist, gfp_zone(gfp_mask)) {
+		/*
+		 * Lock each zone in the zonelist under zone_scan_mutex so a
+		 * parallel invocation of try_set_zone_oom() doesn't succeed
+		 * when it shouldn't.
+		 */
+		zone_set_flag(zone, ZONE_OOM_LOCKED);
+	}
 
-	/*
-	 * Lock each zone in the zonelist under zone_scan_mutex so a parallel
-	 * invocation of try_set_zone_oom() doesn't succeed when it shouldn't.
-	 */
-	z = zonelist->zones;
-	do {
-		zone_set_flag(*z, ZONE_OOM_LOCKED);
-	} while (*(++z) != NULL);
 out:
 	spin_unlock(&zone_scan_mutex);
 	return ret;
@@ -495,16 +495,15 @@ int try_set_zone_oom(struct zonelist *zonelist)
  * allocation attempts with zonelists containing them may now recall the OOM
  * killer, if necessary.
  */
-void clear_zonelist_oom(struct zonelist *zonelist)
+void clear_zonelist_oom(struct zonelist *zonelist, gfp_t gfp_mask)
 {
-	struct zone **z;
-
-	z = zonelist->zones;
+	struct zoneref *z;
+	struct zone *zone;
 
 	spin_lock(&zone_scan_mutex);
-	do {
-		zone_clear_flag(*z, ZONE_OOM_LOCKED);
-	} while (*(++z) != NULL);
+	for_each_zone_zonelist(zone, z, zonelist, gfp_zone(gfp_mask)) {
+		zone_clear_flag(zone, ZONE_OOM_LOCKED);
+	}
 	spin_unlock(&zone_scan_mutex);
 }