dm cache policy smq: don't do any writebacks unless IDLE

If there are no clean blocks to be demoted the writeback will be
triggered at that point.  Preemptively writing back can hurt high IO
load scenarios.

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

diff --git a/drivers/md/dm-cache-policy-smq.c b/drivers/md/dm-cache-policy-smq.c
index 758480a1893dc46eb0b084f909a7087abe833d26..e5eb9c9b4bc8e8265c90debdc3d86eea98f08d96 100644 (file)
--- a/drivers/md/dm-cache-policy-smq.c
+++ b/drivers/md/dm-cache-policy-smq.c
@@ -1120,8 +1120,6 @@ static bool clean_target_met(struct smq_policy *mq, bool idle)
         * Cache entries may not be populated.  So we cannot rely on the
         * size of the clean queue.
         */
-       unsigned nr_clean;
-
        if (idle) {
                /*
                 * We'd like to clean everything.
@@ -1129,9 +1127,10 @@ static bool clean_target_met(struct smq_policy *mq, bool idle)
                return q_size(&mq->dirty) == 0u;
        }
 
-       nr_clean = from_cblock(mq->cache_size) - q_size(&mq->dirty);
-       return (nr_clean + btracker_nr_writebacks_queued(mq->bg_work)) >=
-               percent_to_target(mq, CLEAN_TARGET);
+       /*
+        * If we're busy we don't worry about cleaning at all.
+        */
+       return true;
 }
 
 static bool free_target_met(struct smq_policy *mq)