[PATCH rfc 09/10] nvmet: Use non-selective polling

[Sagi Grimberg <sagi@grimberg.me>]

It doesn't really make sense to do selective polling
because we never care about specific IOs. Non selective
polling can actually help by doing some useful work
while we're submitting a command.

We ask for a batch of (magic) 4 completions which looks
like a decent network<->backend proportion, if less are
available we'll see less.

Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
---
 drivers/nvme/target/io-cmd.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/nvme/target/io-cmd.c b/drivers/nvme/target/io-cmd.c
index 4195115c7e54..8e4fd7ca4a8a 100644
--- a/drivers/nvme/target/io-cmd.c
+++ b/drivers/nvme/target/io-cmd.c
@@ -46,7 +46,6 @@ static void nvmet_execute_rw(struct nvmet_req *req)
 	struct scatterlist *sg;
 	struct bio *bio;
 	sector_t sector;
-	blk_qc_t cookie;
 	int op, op_flags = 0, i;
 
 	if (!req->sg_cnt) {
@@ -85,16 +84,17 @@ static void nvmet_execute_rw(struct nvmet_req *req)
 			bio_set_op_attrs(bio, op, op_flags);
 
 			bio_chain(bio, prev);
-			cookie = submit_bio(prev);
+			submit_bio(prev);
 		}
 
 		sector += sg->length >> 9;
 		sg_cnt--;
 	}
 
-	cookie = submit_bio(bio);
+	submit_bio(bio);
 
-	blk_mq_poll(bdev_get_queue(req->ns->bdev), cookie);
+	/* magic 4 is what we are willing to grab before we return */
+	blk_mq_poll_batch(bdev_get_queue(req->ns->bdev), 4);
 }
 
 static void nvmet_execute_flush(struct nvmet_req *req)
-- 
2.7.4


_______________________________________________
Linux-nvme mailing list
Linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme