1 files changed, 191 insertions, 39 deletions

diff --git a/lib/rbtree_test.c b/lib/rbtree_test.c
index 8b3c9dc88262..191a238e5a9d 100644
--- a/lib/rbtree_test.c
+++ b/lib/rbtree_test.c
@@ -1,11 +1,18 @@
 #include <linux/module.h>
+#include <linux/moduleparam.h>
 #include <linux/rbtree_augmented.h>
 #include <linux/random.h>
+#include <linux/slab.h>
 #include <asm/timex.h>
 
-#define NODES       100
-#define PERF_LOOPS  100000
-#define CHECK_LOOPS 100
+#define __param(type, name, init, msg)		\
+	static type name = init;		\
+	module_param(name, type, 0444);		\
+	MODULE_PARM_DESC(name, msg);
+
+__param(int, nnodes, 100, "Number of nodes in the rb-tree");
+__param(int, perf_loops, 100000, "Number of iterations modifying the rb-tree");
+__param(int, check_loops, 100, "Number of iterations modifying and verifying the rb-tree");
 
 struct test_node {
 	u32 key;
@@ -16,14 +23,14 @@ struct test_node {
 	u32 augmented;
 };
 
-static struct rb_root root = RB_ROOT;
-static struct test_node nodes[NODES];
+static struct rb_root_cached root = RB_ROOT_CACHED;
+static struct test_node *nodes = NULL;
 
 static struct rnd_state rnd;
 
-static void insert(struct test_node *node, struct rb_root *root)
+static void insert(struct test_node *node, struct rb_root_cached *root)
 {
-	struct rb_node **new = &root->rb_node, *parent = NULL;
+	struct rb_node **new = &root->rb_root.rb_node, *parent = NULL;
 	u32 key = node->key;
 
 	while (*new) {
@@ -35,14 +42,40 @@ static void insert(struct test_node *node, struct rb_root *root)
 	}
 
 	rb_link_node(&node->rb, parent, new);
-	rb_insert_color(&node->rb, root);
+	rb_insert_color(&node->rb, &root->rb_root);
+}
+
+static void insert_cached(struct test_node *node, struct rb_root_cached *root)
+{
+	struct rb_node **new = &root->rb_root.rb_node, *parent = NULL;
+	u32 key = node->key;
+	bool leftmost = true;
+
+	while (*new) {
+		parent = *new;
+		if (key < rb_entry(parent, struct test_node, rb)->key)
+			new = &parent->rb_left;
+		else {
+			new = &parent->rb_right;
+			leftmost = false;
+		}
+	}
+
+	rb_link_node(&node->rb, parent, new);
+	rb_insert_color_cached(&node->rb, root, leftmost);
 }
 
-static inline void erase(struct test_node *node, struct rb_root *root)
+static inline void erase(struct test_node *node, struct rb_root_cached *root)
 {
-	rb_erase(&node->rb, root);
+	rb_erase(&node->rb, &root->rb_root);
 }
 
+static inline void erase_cached(struct test_node *node, struct rb_root_cached *root)
+{
+	rb_erase_cached(&node->rb, root);
+}
+
+
 static inline u32 augment_recompute(struct test_node *node)
 {
 	u32 max = node->val, child_augmented;
@@ -64,9 +97,10 @@ static inline u32 augment_recompute(struct test_node *node)
 RB_DECLARE_CALLBACKS(static, augment_callbacks, struct test_node, rb,
 		     u32, augmented, augment_recompute)
 
-static void insert_augmented(struct test_node *node, struct rb_root *root)
+static void insert_augmented(struct test_node *node,
+			     struct rb_root_cached *root)
 {
-	struct rb_node **new = &root->rb_node, *rb_parent = NULL;
+	struct rb_node **new = &root->rb_root.rb_node, *rb_parent = NULL;
 	u32 key = node->key;
 	u32 val = node->val;
 	struct test_node *parent;
@@ -84,18 +118,53 @@ static void insert_augmented(struct test_node *node, struct rb_root *root)
 
 	node->augmented = val;
 	rb_link_node(&node->rb, rb_parent, new);
-	rb_insert_augmented(&node->rb, root, &augment_callbacks);
+	rb_insert_augmented(&node->rb, &root->rb_root, &augment_callbacks);
+}
+
+static void insert_augmented_cached(struct test_node *node,
+				    struct rb_root_cached *root)
+{
+	struct rb_node **new = &root->rb_root.rb_node, *rb_parent = NULL;
+	u32 key = node->key;
+	u32 val = node->val;
+	struct test_node *parent;
+	bool leftmost = true;
+
+	while (*new) {
+		rb_parent = *new;
+		parent = rb_entry(rb_parent, struct test_node, rb);
+		if (parent->augmented < val)
+			parent->augmented = val;
+		if (key < parent->key)
+			new = &parent->rb.rb_left;
+		else {
+			new = &parent->rb.rb_right;
+			leftmost = false;
+		}
+	}
+
+	node->augmented = val;
+	rb_link_node(&node->rb, rb_parent, new);
+	rb_insert_augmented_cached(&node->rb, root,
+				   leftmost, &augment_callbacks);
+}
+
+
+static void erase_augmented(struct test_node *node, struct rb_root_cached *root)
+{
+	rb_erase_augmented(&node->rb, &root->rb_root, &augment_callbacks);
 }
 
-static void erase_augmented(struct test_node *node, struct rb_root *root)
+static void erase_augmented_cached(struct test_node *node,
+				   struct rb_root_cached *root)
 {
-	rb_erase_augmented(&node->rb, root, &augment_callbacks);
+	rb_erase_augmented_cached(&node->rb, root, &augment_callbacks);
 }
 
 static void init(void)
 {
 	int i;
-	for (i = 0; i < NODES; i++) {
+	for (i = 0; i < nnodes; i++) {
 		nodes[i].key = prandom_u32_state(&rnd);
 		nodes[i].val = prandom_u32_state(&rnd);
 	}
@@ -118,7 +187,7 @@ static void check_postorder_foreach(int nr_nodes)
 {
 	struct test_node *cur, *n;
 	int count = 0;
-	rbtree_postorder_for_each_entry_safe(cur, n, &root, rb)
+	rbtree_postorder_for_each_entry_safe(cur, n, &root.rb_root, rb)
 		count++;
 
 	WARN_ON_ONCE(count != nr_nodes);
@@ -128,7 +197,7 @@ static void check_postorder(int nr_nodes)
 {
 	struct rb_node *rb;
 	int count = 0;
-	for (rb = rb_first_postorder(&root); rb; rb = rb_next_postorder(rb))
+	for (rb = rb_first_postorder(&root.rb_root); rb; rb = rb_next_postorder(rb))
 		count++;
 
 	WARN_ON_ONCE(count != nr_nodes);
@@ -140,7 +209,7 @@ static void check(int nr_nodes)
 	int count = 0, blacks = 0;
 	u32 prev_key = 0;
 
-	for (rb = rb_first(&root); rb; rb = rb_next(rb)) {
+	for (rb = rb_first(&root.rb_root); rb; rb = rb_next(rb)) {
 		struct test_node *node = rb_entry(rb, struct test_node, rb);
 		WARN_ON_ONCE(node->key < prev_key);
 		WARN_ON_ONCE(is_red(rb) &&
@@ -155,7 +224,7 @@ static void check(int nr_nodes)
 	}
 
 	WARN_ON_ONCE(count != nr_nodes);
-	WARN_ON_ONCE(count < (1 << black_path_count(rb_last(&root))) - 1);
+	WARN_ON_ONCE(count < (1 << black_path_count(rb_last(&root.rb_root))) - 1);
 
 	check_postorder(nr_nodes);
 	check_postorder_foreach(nr_nodes);
@@ -166,7 +235,7 @@ static void check_augmented(int nr_nodes)
 	struct rb_node *rb;
 
 	check(nr_nodes);
-	for (rb = rb_first(&root); rb; rb = rb_next(rb)) {
+	for (rb = rb_first(&root.rb_root); rb; rb = rb_next(rb)) {
 		struct test_node *node = rb_entry(rb, struct test_node, rb);
 		WARN_ON_ONCE(node->augmented != augment_recompute(node));
 	}
@@ -176,6 +245,11 @@ static int __init rbtree_test_init(void)
 {
 	int i, j;
 	cycles_t time1, time2, time;
+	struct rb_node *node;
+
+	nodes = kmalloc(nnodes * sizeof(*nodes), GFP_KERNEL);
+	if (!nodes)
+		return -ENOMEM;
 
 	printk(KERN_ALERT "rbtree testing");
 
@@ -184,27 +258,88 @@ static int __init rbtree_test_init(void)
 
 	time1 = get_cycles();
 
-	for (i = 0; i < PERF_LOOPS; i++) {
-		for (j = 0; j < NODES; j++)
+	for (i = 0; i < perf_loops; i++) {
+		for (j = 0; j < nnodes; j++)
 			insert(nodes + j, &root);
-		for (j = 0; j < NODES; j++)
+		for (j = 0; j < nnodes; j++)
 			erase(nodes + j, &root);
 	}
 
 	time2 = get_cycles();
 	time = time2 - time1;
 
-	time = div_u64(time, PERF_LOOPS);
-	printk(" -> %llu cycles\n", (unsigned long long)time);
+	time = div_u64(time, perf_loops);
+	printk(" -> test 1 (latency of nnodes insert+delete): %llu cycles\n",
+	       (unsigned long long)time);
+
+	time1 = get_cycles();
+
+	for (i = 0; i < perf_loops; i++) {
+		for (j = 0; j < nnodes; j++)
+			insert_cached(nodes + j, &root);
+		for (j = 0; j < nnodes; j++)
+			erase_cached(nodes + j, &root);
+	}
+
+	time2 = get_cycles();
+	time = time2 - time1;
+
+	time = div_u64(time, perf_loops);
+	printk(" -> test 2 (latency of nnodes cached insert+delete): %llu cycles\n",
+	       (unsigned long long)time);
+
+	for (i = 0; i < nnodes; i++)
+		insert(nodes + i, &root);
+
+	time1 = get_cycles();
+
+	for (i = 0; i < perf_loops; i++) {
+		for (node = rb_first(&root.rb_root); node; node = rb_next(node))
+			;
+	}
+
+	time2 = get_cycles();
+	time = time2 - time1;
+
+	time = div_u64(time, perf_loops);
+	printk(" -> test 3 (latency of inorder traversal): %llu cycles\n",
+	       (unsigned long long)time);
+
+	time1 = get_cycles();
+
+	for (i = 0; i < perf_loops; i++)
+		node = rb_first(&root.rb_root);
+
+	time2 = get_cycles();
+	time = time2 - time1;
+
+	time = div_u64(time, perf_loops);
+	printk(" -> test 4 (latency to fetch first node)\n");
+	printk("        non-cached: %llu cycles\n", (unsigned long long)time);
+
+	time1 = get_cycles();
+
+	for (i = 0; i < perf_loops; i++)
+		node = rb_first_cached(&root);
+
+	time2 = get_cycles();
+	time = time2 - time1;
+
+	time = div_u64(time, perf_loops);
+	printk("        cached: %llu cycles\n", (unsigned long long)time);
 
-	for (i = 0; i < CHECK_LOOPS; i++) {
+	for (i = 0; i < nnodes; i++)
+		erase(nodes + i, &root);
+
+	/* run checks */
+	for (i = 0; i < check_loops; i++) {
 		init();
-		for (j = 0; j < NODES; j++) {
+		for (j = 0; j < nnodes; j++) {
 			check(j);
 			insert(nodes + j, &root);
 		}
-		for (j = 0; j < NODES; j++) {
-			check(NODES - j);
+		for (j = 0; j < nnodes; j++) {
+			check(nnodes - j);
 			erase(nodes + j, &root);
 		}
 		check(0);
@@ -216,32 +351,49 @@ static int __init rbtree_test_init(void)
 
 	time1 = get_cycles();
 
-	for (i = 0; i < PERF_LOOPS; i++) {
-		for (j = 0; j < NODES; j++)
+	for (i = 0; i < perf_loops; i++) {
+		for (j = 0; j < nnodes; j++)
 			insert_augmented(nodes + j, &root);
-		for (j = 0; j < NODES; j++)
+		for (j = 0; j < nnodes; j++)
 			erase_augmented(nodes + j, &root);
 	}
 
 	time2 = get_cycles();
 	time = time2 - time1;
 
-	time = div_u64(time, PERF_LOOPS);
-	printk(" -> %llu cycles\n", (unsigned long long)time);
+	time = div_u64(time, perf_loops);
+	printk(" -> test 1 (latency of nnodes insert+delete): %llu cycles\n", (unsigned long long)time);
+
+	time1 = get_cycles();
+
+	for (i = 0; i < perf_loops; i++) {
+		for (j = 0; j < nnodes; j++)
+			insert_augmented_cached(nodes + j, &root);
+		for (j = 0; j < nnodes; j++)
+			erase_augmented_cached(nodes + j, &root);
+	}
+
+	time2 = get_cycles();
+	time = time2 - time1;
+
+	time = div_u64(time, perf_loops);
+	printk(" -> test 2 (latency of nnodes cached insert+delete): %llu cycles\n", (unsigned long long)time);
 
-	for (i = 0; i < CHECK_LOOPS; i++) {
+	for (i = 0; i < check_loops; i++) {
 		init();
-		for (j = 0; j < NODES; j++) {
+		for (j = 0; j < nnodes; j++) {
 			check_augmented(j);
 			insert_augmented(nodes + j, &root);
 		}
-		for (j = 0; j < NODES; j++) {
-			check_augmented(NODES - j);
+		for (j = 0; j < nnodes; j++) {
+			check_augmented(nnodes - j);
 			erase_augmented(nodes + j, &root);
 		}
 		check_augmented(0);
 	}
 
+	kfree(nodes);
+
 	return -EAGAIN; /* Fail will directly unload the module */
 }