When reclaiming zones we should arbitrate between the zoned
devices to get a better throughput. So implement a simple
round-robin load balancer between the zoned devices.
Signed-off-by: Hannes Reinecke <hare@xxxxxxx>
---
drivers/md/dm-zoned-metadata.c | 8 +++++++-
1 file changed, 7 insertions(+), 1 deletion(-)
diff --git a/drivers/md/dm-zoned-metadata.c b/drivers/md/dm-zoned-metadata.c
index 87784e7785bc..25dcad2a565f 100644
--- a/drivers/md/dm-zoned-metadata.c
+++ b/drivers/md/dm-zoned-metadata.c
@@ -171,6 +171,8 @@ struct dmz_metadata {
unsigned int nr_reserved_seq;
unsigned int nr_chunks;
+ unsigned int last_alloc_idx;
+
/* Zone information array */
struct xarray zones;
@@ -2178,7 +2180,7 @@ struct dm_zone *dmz_alloc_zone(struct dmz_metadata *zmd, unsigned long flags)
{
struct list_head *list;
struct dm_zone *zone;
- unsigned int dev_idx = 0;
+ unsigned int dev_idx = zmd->last_alloc_idx;
again:
if (flags & DMZ_ALLOC_CACHE)
@@ -2214,6 +2216,9 @@ struct dm_zone *dmz_alloc_zone(struct dmz_metadata *zmd, unsigned long flags)
zone = list_first_entry(list, struct dm_zone, link);
list_del_init(&zone->link);
+ if (!(flags & DMZ_ALLOC_CACHE))
+ zmd->last_alloc_idx = (dev_idx + 1) % zmd->nr_devs;
+
if (dmz_is_cache(zone))
atomic_dec(&zmd->unmap_nr_cache);
else if (dmz_is_rnd(zone))
@@ -2839,6 +2844,7 @@ int dmz_ctr_metadata(struct dmz_dev *dev, int num_dev,
zmd->dev = dev;
zmd->nr_devs = num_dev;
zmd->mblk_rbtree = RB_ROOT;
+ zmd->last_alloc_idx = 0;
init_rwsem(&zmd->mblk_sem);
mutex_init(&zmd->mblk_flush_lock);
spin_lock_init(&zmd->mblk_lock);
--
2.16.4
--
dm-devel mailing list
dm-devel@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/dm-devel
