nvme: avoid duplicate SET_NUM_QUEUES commands

nvme(4) issues a SET_NUM_QUEUES command during device
initialization to ensure enough I/O queues exists for each
of the MSI-X vectors we have allocated.  The SET_NUM_QUEUES
command is then issued again during nvme_ctrlr_start(), to
ensure that is properly set after any controller reset.

At least one NVMe drive exists which fails this second
SET_NUM_QUEUES command during device initialization.  So
change nvme_ctrlr_start() to only issue its SET_NUM_QUEUES
command when it is coming out of a reset - avoiding the
duplicate SET_NUM_QUEUES during device initialization.

Reported by:	gallatin
MFC after:	3 days
Sponsored by:	Intel

1 files changed, 10 insertions, 8 deletions

diff --git a/sys/dev/nvme/nvme_ctrlr.c b/sys/dev/nvme/nvme_ctrlr.c
index e400f52..4c8adc0 100644
--- a/sys/dev/nvme/nvme_ctrlr.c
+++ b/sys/dev/nvme/nvme_ctrlr.c
@@ -725,15 +725,17 @@ nvme_ctrlr_start(void *ctrlr_arg)
 	 *  explicit specify how many queues it will use.  This value should
 	 *  never change between resets, so panic if somehow that does happen.
 	 */
-	old_num_io_queues = ctrlr->num_io_queues;
-	if (nvme_ctrlr_set_num_qpairs(ctrlr) != 0) {
-		nvme_ctrlr_fail(ctrlr);
-		return;
-	}
+	if (ctrlr->is_resetting) {
+		old_num_io_queues = ctrlr->num_io_queues;
+		if (nvme_ctrlr_set_num_qpairs(ctrlr) != 0) {
+			nvme_ctrlr_fail(ctrlr);
+			return;
+		}
 
-	if (old_num_io_queues != ctrlr->num_io_queues) {
-		panic("num_io_queues changed from %u to %u", old_num_io_queues,
-		    ctrlr->num_io_queues);
+		if (old_num_io_queues != ctrlr->num_io_queues) {
+			panic("num_io_queues changed from %u to %u",
+			      old_num_io_queues, ctrlr->num_io_queues);
+		}
 	}
 
 	if (nvme_ctrlr_create_qpairs(ctrlr) != 0) {