[PATCH RFC v2 3/3] nvme-rdma: use rdma_reject_msg() to log connection rejects

[Christoph Hellwig]

On Thu, Oct 20, 2016 at 03:40:29PM -0700, Steve Wise wrote:
> @@ -1237,18 +1237,22 @@ out_destroy_queue_ib:
>  static int nvme_rdma_conn_rejected(struct nvme_rdma_queue *queue,
>  		struct rdma_cm_event *ev)
>  {
> +	struct rdma_cm_id *cm_id = queue->cm_id;
> +	int rdma_status = ev->status;
> +	short nvme_status = -1;
> +
> +	if (rdma_consumer_reject(cm_id, rdma_status) &&
> +	    ev->param.conn.private_data_len) {
>  		struct nvme_rdma_cm_rej *rej =
>  			(struct nvme_rdma_cm_rej *)ev->param.conn.private_data;

Given the nasty casting issues in the current RDMA/CM API maybe we should
actually expand the scope of the rdma_consumer_reject helper to include
the above check, e.g. check that there is a private data len and then
return a pointer to the private data?

Something like

static int nvme_rdma_conn_rejected(struct nvme_rdma_queue *queue,
		struct rdma_cm_event *ev)
{
	struct rdma_cm_id *cm_id = queue->cm_id;
	struct nvme_rdma_cm_rej *rej
	short nvme_status = -1;

	rej = rdma_cm_reject_message(ev);
	if (rej)
		nvme_status = le16_to_cpu(rej->sts);

>  
> +	dev_err(queue->ctrl->ctrl.device, "Connect rejected: status %d (%s) "
> +		"nvme status %d.\n", rdma_status,
> +		rdma_reject_msg(cm_id, rdma_status), nvme_status);

And while we're pretty printing the rest it would be nice to pretty
print the NVMe status here as well.