On Tue, Jul 20, 2021 at 10:48:47AM +0800, Hou Pu wrote:

On Mon, 19 Jul 2021 09:44:39 -0700, Keith Busch wrote:

quoted

@@ -66,13 +66,13 @@ TRACE_EVENT(nvme_setup_cmd,
 		__entry->flags = cmd->common.flags;
 		__entry->cid = cmd->common.command_id;
 		__entry->nsid = le32_to_cpu(cmd->common.nsid);
-		__entry->metadata = le64_to_cpu(cmd->common.metadata);
+		__entry->metadata = !!blk_integrity_rq(req);
 		__entry->fctype = cmd->fabrics.fctype;
 		__assign_disk_name(__entry->disk, req->rq_disk);
 		memcpy(__entry->cdw10, &cmd->common.cdw10,
 			sizeof(__entry->cdw10));
 	    ),
-	    TP_printk("nvme%d: %sqid=%d, cmdid=%u, nsid=%u, flags=0x%x, meta=0x%llx, cmd=(%s %s)",
+	    TP_printk("nvme%d: %sqid=%d, cmdid=%u, nsid=%u, flags=0x%x, meta=0x%x, cmd=(%s %s)",

Can we also change "cmd=(%s %s)" to "cmd(%s, %s)" like the target side.

That would have to be a different patch, and you can send it if you
would like to see this change.

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