dm space maps: improve performance with inc/dec on ranges of blocks

When we break sharing on btree nodes we typically need to increment
the reference counts to every value held in the node.  This can
cause a lot of repeated calls to the space maps.  Fix this by changing
the interface to the space map inc/dec methods to take ranges of
adjacent blocks to be operated on.

For installations that are using a lot of snapshots this will reduce
cpu overhead of fundamental operations such as provisioning a new block,
or deleting a snapshot, by as much as 10 times.

Signed-off-by: Joe Thornber <ejt@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>

1 files changed, 28 insertions, 24 deletions

diff --git a/drivers/md/persistent-data/dm-array.c b/drivers/md/persistent-data/dm-array.c
index 185dc60360b5..3a963d783a86 100644
--- a/drivers/md/persistent-data/dm-array.c
+++ b/drivers/md/persistent-data/dm-array.c
@@ -108,12 +108,10 @@ static void *element_at(struct dm_array_info *info, struct array_block *ab,
  * in an array block.
  */
 static void on_entries(struct dm_array_info *info, struct array_block *ab,
-		       void (*fn)(void *, const void *))
+		       void (*fn)(void *, const void *, unsigned))
 {
-	unsigned i, nr_entries = le32_to_cpu(ab->nr_entries);
-
-	for (i = 0; i < nr_entries; i++)
-		fn(info->value_type.context, element_at(info, ab, i));
+	unsigned nr_entries = le32_to_cpu(ab->nr_entries);
+	fn(info->value_type.context, element_at(info, ab, 0), nr_entries);
 }
 
 /*
@@ -175,19 +173,18 @@ static int alloc_ablock(struct dm_array_info *info, size_t size_of_block,
 static void fill_ablock(struct dm_array_info *info, struct array_block *ab,
 			const void *value, unsigned new_nr)
 {
-	unsigned i;
-	uint32_t nr_entries;
+	uint32_t nr_entries, delta, i;
 	struct dm_btree_value_type *vt = &info->value_type;
 
 	BUG_ON(new_nr > le32_to_cpu(ab->max_entries));
 	BUG_ON(new_nr < le32_to_cpu(ab->nr_entries));
 
 	nr_entries = le32_to_cpu(ab->nr_entries);
-	for (i = nr_entries; i < new_nr; i++) {
-		if (vt->inc)
-			vt->inc(vt->context, value);
+	delta = new_nr - nr_entries;
+	if (vt->inc)
+		vt->inc(vt->context, value, delta);
+	for (i = nr_entries; i < new_nr; i++)
 		memcpy(element_at(info, ab, i), value, vt->size);
-	}
 	ab->nr_entries = cpu_to_le32(new_nr);
 }
 
@@ -199,17 +196,16 @@ static void fill_ablock(struct dm_array_info *info, struct array_block *ab,
 static void trim_ablock(struct dm_array_info *info, struct array_block *ab,
 			unsigned new_nr)
 {
-	unsigned i;
-	uint32_t nr_entries;
+	uint32_t nr_entries, delta;
 	struct dm_btree_value_type *vt = &info->value_type;
 
 	BUG_ON(new_nr > le32_to_cpu(ab->max_entries));
 	BUG_ON(new_nr > le32_to_cpu(ab->nr_entries));
 
 	nr_entries = le32_to_cpu(ab->nr_entries);
-	for (i = nr_entries; i > new_nr; i--)
-		if (vt->dec)
-			vt->dec(vt->context, element_at(info, ab, i - 1));
+	delta = nr_entries - new_nr;
+	if (vt->dec)
+		vt->dec(vt->context, element_at(info, ab, new_nr - 1), delta);
 	ab->nr_entries = cpu_to_le32(new_nr);
 }
 
@@ -573,16 +569,17 @@ static int grow(struct resize *resize)
  * These are the value_type functions for the btree elements, which point
  * to array blocks.
  */
-static void block_inc(void *context, const void *value)
+static void block_inc(void *context, const void *value, unsigned count)
 {
-	__le64 block_le;
+	const __le64 *block_le = value;
 	struct dm_array_info *info = context;
+	unsigned i;
 
-	memcpy(&block_le, value, sizeof(block_le));
-	dm_tm_inc(info->btree_info.tm, le64_to_cpu(block_le));
+	for (i = 0; i < count; i++, block_le++)
+		dm_tm_inc(info->btree_info.tm, le64_to_cpu(*block_le));
 }
 
-static void block_dec(void *context, const void *value)
+static void __block_dec(void *context, const void *value)
 {
 	int r;
 	uint64_t b;
@@ -621,6 +618,13 @@ static void block_dec(void *context, const void *value)
 	dm_tm_dec(info->btree_info.tm, b);
 }
 
+static void block_dec(void *context, const void *value, unsigned count)
+{
+	unsigned i;
+	for (i = 0; i < count; i++, value += sizeof(__le64))
+		__block_dec(context, value);
+}
+
 static int block_equal(void *context, const void *value1, const void *value2)
 {
 	return !memcmp(value1, value2, sizeof(__le64));
@@ -711,7 +715,7 @@ static int populate_ablock_with_values(struct dm_array_info *info, struct array_
 			return r;
 
 		if (vt->inc)
-			vt->inc(vt->context, element_at(info, ab, i));
+			vt->inc(vt->context, element_at(info, ab, i), 1);
 	}
 
 	ab->nr_entries = cpu_to_le32(new_nr);
@@ -822,9 +826,9 @@ static int array_set_value(struct dm_array_info *info, dm_block_t root,
 	old_value = element_at(info, ab, entry);
 	if (vt->dec &&
 	    (!vt->equal || !vt->equal(vt->context, old_value, value))) {
-		vt->dec(vt->context, old_value);
+		vt->dec(vt->context, old_value, 1);
 		if (vt->inc)
-			vt->inc(vt->context, value);
+			vt->inc(vt->context, value, 1);
 	}
 
 	memcpy(old_value, value, info->value_type.size);