nvme-pci: reverse request order in nvme_queue_rqs

blk_mq_flush_plug_list submits requests in the reverse order that they
were submitted, which leads to a rather suboptimal I/O pattern especially
in rotational devices.  Fix this by rewriting nvme_queue_rqs so that it
always pops the requests from the passed in request list, and then adds
them to the head of a local submit list.  This actually simplifies the
code a bit as it removes the complicated list splicing, at the cost of
extra updates of the rq_next pointer.  As that should be cache hot
anyway it should be an easy price to pay.

Fixes: d62cbcf62f2f ("nvme: add support for mq_ops->queue_rqs()")
Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20241113152050.157179-2-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
index 0aa26a33f23153942bc4cf409a771d7fb1c6ab80..ec1c44c75d9241d39f4e1db82e53ee7ff1091db2 100644 (file)
--- a/drivers/nvme/host/pci.c
+++ b/drivers/nvme/host/pci.c
@@ -906,9 +906,10 @@ static blk_status_t nvme_queue_rq(struct blk_mq_hw_ctx *hctx,
 
 static void nvme_submit_cmds(struct nvme_queue *nvmeq, struct request **rqlist)
 {
+       struct request *req;
+
        spin_lock(&nvmeq->sq_lock);
-       while (!rq_list_empty(*rqlist)) {
-               struct request *req = rq_list_pop(rqlist);
+       while ((req = rq_list_pop(rqlist))) {
                struct nvme_iod *iod = blk_mq_rq_to_pdu(req);
 
                nvme_sq_copy_cmd(nvmeq, &iod->cmd);
@@ -933,31 +934,25 @@ static bool nvme_prep_rq_batch(struct nvme_queue *nvmeq, struct request *req)
 
 static void nvme_queue_rqs(struct request **rqlist)
 {
-       struct request *req, *next, *prev = NULL;
+       struct request *submit_list = NULL;
        struct request *requeue_list = NULL;
+       struct request **requeue_lastp = &requeue_list;
+       struct nvme_queue *nvmeq = NULL;
+       struct request *req;
 
-       rq_list_for_each_safe(rqlist, req, next) {
-               struct nvme_queue *nvmeq = req->mq_hctx->driver_data;
-
-               if (!nvme_prep_rq_batch(nvmeq, req)) {
-                       /* detach 'req' and add to remainder list */
-                       rq_list_move(rqlist, &requeue_list, req, prev);
-
-                       req = prev;
-                       if (!req)
-                               continue;
-               }
+       while ((req = rq_list_pop(rqlist))) {
+               if (nvmeq && nvmeq != req->mq_hctx->driver_data)
+                       nvme_submit_cmds(nvmeq, &submit_list);
+               nvmeq = req->mq_hctx->driver_data;
 
-               if (!next || req->mq_hctx != next->mq_hctx) {
-                       /* detach rest of list, and submit */
-                       req->rq_next = NULL;
-                       nvme_submit_cmds(nvmeq, rqlist);
-                       *rqlist = next;
-                       prev = NULL;
-               } else
-                       prev = req;
+               if (nvme_prep_rq_batch(nvmeq, req))
+                       rq_list_add(&submit_list, req); /* reverse order */
+               else
+                       rq_list_add_tail(&requeue_lastp, req);
        }
 
+       if (nvmeq)
+               nvme_submit_cmds(nvmeq, &submit_list);
        *rqlist = requeue_list;
 }