[PATCHv3 2/4] nvme: extend show-topology command to add support for multipath

[Nilay Shroff <nilay@linux.ibm.com>]

This commit enhances the show-topology command by adding support for
NVMe multipath. With this change, users can now list all paths to a
namespace from its corresponding head node device. Each NVMe path
entry then also includes additional details such as ANA state, NUMA
node, and queue depth, improving visibility into multipath configs.
This information can be particularly helpful for debugging and
analyzing NVMe multipath setups.

To support this functionality, the "--ranking" option of the nvme
show-topology command has been extended with a new sub-option:
"multipath".

Since this enhancement is specific to NVMe multipath, the iopolicy
configured under each subsystem is now always displayed. Previously,
iopolicy was shown only with nvme show-topology verbose output, but
it is now included by default to improve usability and provide better
context when reviewing multipath configurations via show-topology.

With this update, users can view the multipath topology of a multi
controller/port NVMe disk.

Examples:

$ nvme show-topology -r multipath

nvme-subsys2 - NQN=nvmet_subsystem
               hostnqn=nqn.2014-08.org.nvmexpress:uuid:12b49f6e-0276-4746-b10c-56815b7e6dc2
               iopolicy=numa

          _ _ _<head-node>
         /              _ _ _ <ana-state>
        /              /      _ _ _ <numa-node-list>
       /              /      /
      |              /      /
 +- nvme2n1 (ns 1)  /      /
 \                 |      |
  +- nvme2c2n1 optimized 1,2 nvme2 tcp traddr=127.0.0.2,trsvcid=4460,src_addr=127.0.0.1 live
  +- nvme2c3n1 optimized 3,4 nvme3 tcp traddr=127.0.0.3,trsvcid=4460,src_addr=127.0.0.1 live

For iopolicy=numa, only NUMA node list is shown (queue depth is hidden).

$ nvme show-topology -r multipath

nvme-subsys2 - NQN=nvmet_subsystem
               hostnqn=nqn.2014-08.org.nvmexpress:uuid:12b49f6e-0276-4746-b10c-56815b7e6dc2
               iopolicy=queue-depth

          _ _ _<head-node>
         /              _ _ _ <ana-state>
        /              /
       /              /     _ _ _<queue-depth>
      |              /     /
 +- nvme2n1 (ns 1)  /     /
 \                 |     |
  +- nvme2c2n1 optimized 0 nvme2 tcp traddr=127.0.0.2,trsvcid=4460,src_addr=127.0.0.1 live
  +- nvme2c3n1 optimized 0 nvme3 tcp traddr=127.0.0.3,trsvcid=4460,src_addr=127.0.0.1 live

For iopolicy=queue-depth, queue depth is shown (NUMA node list is hidden).

$ nvme show-topology -r multipath

nvme-subsys2 - NQN=nvmet_subsystem
               hostnqn=nqn.2014-08.org.nvmexpress:uuid:12b49f6e-0276-4746-b10c-56815b7e6dc2
               iopolicy=round-robin

          _ _ _<head-node>
         /              _ _ _ <ana-state>
        /              /
       /              /
      |              /
 +- nvme2n1 (ns 1)  /
 \                 |
  +- nvme2c2n1 optimized nvme2 tcp traddr=127.0.0.2,trsvcid=4460,src_addr=127.0.0.1 live
  +- nvme2c3n1 optimized nvme3 tcp traddr=127.0.0.3,trsvcid=4460,src_addr=127.0.0.1 live

For iopolicy=round-robin, both NUMA node list and queue depth are hidden.

Note:
The annotations above (e.g., <numa-node-list>, <ana-state>, <head-node>,
<queue-depth>) are for illustration only and are not part of the actual
command output. A more human-friendly tabular format will be introduced
in a follow-up patches.

Signed-off-by: Nilay Shroff <nilay@linux.ibm.com>
---
 nvme-print-binary.c |  1 +
 nvme-print-json.c   | 36 +++++++++++++++----
 nvme-print-stdout.c | 85 ++++++++++++++++++++++++++++++++++++++++++---
 nvme-print.c        |  4 ++-
 nvme-print.h        |  1 +
 nvme.c              |  4 ++-
 nvme.h              |  1 +
 7 files changed, 120 insertions(+), 12 deletions(-)

diff --git a/nvme-print-binary.c b/nvme-print-binary.c
index 351513b9..17def142 100644
--- a/nvme-print-binary.c
+++ b/nvme-print-binary.c
@@ -428,6 +428,7 @@ static struct print_ops binary_print_ops = {
 	.print_nvme_subsystem_list	= NULL,
 	.topology_ctrl			= NULL,
 	.topology_namespace		= NULL,
+	.topology_multipath		= NULL,
 
 	/* status and error messages */
 	.connect_msg			= NULL,
diff --git a/nvme-print-json.c b/nvme-print-json.c
index 829ba718..cac9e25a 100644
--- a/nvme-print-json.c
+++ b/nvme-print-json.c
@@ -4743,6 +4743,7 @@ static unsigned int json_subsystem_topology_multipath(nvme_subsystem_t s,
 	nvme_ns_t n;
 	nvme_path_t p;
 	unsigned int i = 0;
+	const char *iopolicy = nvme_subsystem_get_iopolicy(s);
 
 	nvme_subsystem_for_each_ns(s, n) {
 		struct json_object *ns_attrs;
@@ -4750,19 +4751,40 @@ static unsigned int json_subsystem_topology_multipath(nvme_subsystem_t s,
 
 		ns_attrs = json_create_object();
 		obj_add_int(ns_attrs, "NSID", nvme_ns_get_nsid(n));
+		obj_add_str(ns_attrs, "Name", nvme_ns_get_name(n));
 
 		paths = json_create_array();
 		nvme_namespace_for_each_path(n, p) {
 			struct json_object *path_attrs;
-
-			nvme_ctrl_t c = nvme_path_get_ctrl(p);
+			struct json_object *ctrls, *ctrl_attrs;
+			nvme_ctrl_t c;
 
 			path_attrs = json_create_object();
-			obj_add_str(path_attrs, "Name", nvme_ctrl_get_name(c));
-			obj_add_str(path_attrs, "Transport", nvme_ctrl_get_transport(c));
-			obj_add_str(path_attrs, "Address", nvme_ctrl_get_address(c));
-			obj_add_str(path_attrs, "State", nvme_ctrl_get_state(c));
+			obj_add_str(path_attrs, "Path", nvme_path_get_name(p));
 			obj_add_str(path_attrs, "ANAState", nvme_path_get_ana_state(p));
+
+			/*
+			 * For iopolicy numa exclude "Qdepth", for iopolicy
+			 * queue-depth exclude "NUMANodes" and for iopolicy
+			 * round-robin exclude both "Qdepth" and "NUMANodes".
+			 */
+			if (!strcmp(iopolicy, "numa"))
+				obj_add_str(path_attrs, "NUMANodes",
+						nvme_path_get_numa_nodes(p));
+			else if (!strcmp(iopolicy, "queue-depth"))
+				obj_add_int(path_attrs, "Qdepth",
+						nvme_path_get_queue_depth(p));
+
+			c = nvme_path_get_ctrl(p);
+			ctrls = json_create_array();
+			ctrl_attrs = json_create_object();
+			obj_add_str(ctrl_attrs, "Name", nvme_ctrl_get_name(c));
+			obj_add_str(ctrl_attrs, "Transport", nvme_ctrl_get_transport(c));
+			obj_add_str(ctrl_attrs, "Address", nvme_ctrl_get_address(c));
+			obj_add_str(ctrl_attrs, "State", nvme_ctrl_get_state(c));
+			array_add_obj(ctrls, ctrl_attrs);
+			obj_add_array(path_attrs, "Controller", ctrls);
+
 			array_add_obj(paths, path_attrs);
 		}
 		obj_add_array(ns_attrs, "Paths", paths);
@@ -4787,6 +4809,7 @@ static void json_print_nvme_subsystem_topology(nvme_subsystem_t s,
 
 			ns_attrs = json_create_object();
 			obj_add_int(ns_attrs, "NSID", nvme_ns_get_nsid(n));
+			obj_add_str(ns_attrs, "Name", nvme_ns_get_name(n));
 
 			ctrl = json_create_array();
 			ctrl_attrs = json_create_object();
@@ -5546,6 +5569,7 @@ static struct print_ops json_print_ops = {
 	.print_nvme_subsystem_list	= json_print_nvme_subsystem_list,
 	.topology_ctrl			= json_simple_topology,
 	.topology_namespace		= json_simple_topology,
+	.topology_multipath		= json_simple_topology,
 
 	/* status and error messages */
 	.connect_msg			= json_connect_msg,
diff --git a/nvme-print-stdout.c b/nvme-print-stdout.c
index eb56349a..204c8d9d 100644
--- a/nvme-print-stdout.c
+++ b/nvme-print-stdout.c
@@ -1126,6 +1126,8 @@ static void stdout_subsys_config(nvme_subsystem_t s)
 	       nvme_subsystem_get_nqn(s));
 	printf("%*s   hostnqn=%s\n", len, " ",
 	       nvme_host_get_hostnqn(nvme_subsystem_get_host(s)));
+	printf("%*s   iopolicy=%s\n", len, " ",
+		nvme_subsystem_get_iopolicy(s));
 
 	if (stdout_print_ops.flags & VERBOSE) {
 		printf("%*s   model=%s\n", len, " ",
@@ -1134,8 +1136,6 @@ static void stdout_subsys_config(nvme_subsystem_t s)
 			nvme_subsystem_get_serial(s));
 		printf("%*s   firmware=%s\n", len, " ",
 			nvme_subsystem_get_fw_rev(s));
-		printf("%*s   iopolicy=%s\n", len, " ",
-			nvme_subsystem_get_iopolicy(s));
 		printf("%*s   type=%s\n", len, " ",
 			nvme_subsystem_get_type(s));
 	}
@@ -5595,6 +5595,7 @@ static void stdout_subsystem_topology_multipath(nvme_subsystem_t s,
 	nvme_ns_t n;
 	nvme_path_t p;
 	nvme_ctrl_t c;
+	const char *iopolicy = nvme_subsystem_get_iopolicy(s);
 
 	if (ranking == NVME_CLI_TOPO_NAMESPACE) {
 		nvme_subsystem_for_each_ns(s, n) {
@@ -5615,7 +5616,7 @@ static void stdout_subsystem_topology_multipath(nvme_subsystem_t s,
 				       nvme_path_get_ana_state(p));
 			}
 		}
-	} else {
+	} else if (ranking == NVME_CLI_TOPO_CTRL) {
 		/* NVME_CLI_TOPO_CTRL */
 		nvme_subsystem_for_each_ctrl(s, c) {
 			printf(" +- %s %s %s\n",
@@ -5636,6 +5637,59 @@ static void stdout_subsystem_topology_multipath(nvme_subsystem_t s,
 				}
 			}
 		}
+	} else {
+		/* NVME_CLI_TOPO_MULTIPATH */
+		nvme_subsystem_for_each_ns(s, n) {
+			printf(" +- %s (ns %d)\n",
+					nvme_ns_get_name(n),
+					nvme_ns_get_nsid(n));
+			printf(" \\\n");
+			nvme_namespace_for_each_path(n, p) {
+				c = nvme_path_get_ctrl(p);
+
+				if (!strcmp(iopolicy, "numa")) {
+					/*
+					 * For iopolicy numa, exclude printing
+					 * qdepth.
+					 */
+					printf("  +- %s %s %s %s %s %s %s\n",
+						nvme_path_get_name(p),
+						nvme_path_get_ana_state(p),
+						nvme_path_get_numa_nodes(p),
+						nvme_ctrl_get_name(c),
+						nvme_ctrl_get_transport(c),
+						nvme_ctrl_get_address(c),
+						nvme_ctrl_get_state(c));
+
+				} else if (!strcmp(iopolicy, "queue-depth")) {
+					/*
+					 * For iopolicy queue-depth, exclude
+					 * printing numa nodes.
+					 */
+					printf("  +- %s %s %d %s %s %s %s\n",
+						nvme_path_get_name(p),
+						nvme_path_get_ana_state(p),
+						nvme_path_get_queue_depth(p),
+						nvme_ctrl_get_name(c),
+						nvme_ctrl_get_transport(c),
+						nvme_ctrl_get_address(c),
+						nvme_ctrl_get_state(c));
+
+				} else { /* round-robin */
+					/*
+					 * For iopolicy round-robin, exclude
+					 * printing numa nodes and qdepth.
+					 */
+					printf("  +- %s %s %s %s %s %s\n",
+						nvme_path_get_name(p),
+						nvme_path_get_ana_state(p),
+						nvme_ctrl_get_name(c),
+						nvme_ctrl_get_transport(c),
+						nvme_ctrl_get_address(c),
+						nvme_ctrl_get_state(c));
+				}
+			}
+		}
 	}
 }
 
@@ -5657,7 +5711,7 @@ static void stdout_subsystem_topology(nvme_subsystem_t s,
 				       nvme_ctrl_get_state(c));
 			}
 		}
-	} else {
+	} else if (ranking == NVME_CLI_TOPO_CTRL) {
 		/* NVME_CLI_TOPO_CTRL */
 		nvme_subsystem_for_each_ctrl(s, c) {
 			printf(" +- %s %s %s\n",
@@ -5671,6 +5725,23 @@ static void stdout_subsystem_topology(nvme_subsystem_t s,
 				       nvme_ctrl_get_state(c));
 			}
 		}
+	} else {
+		/* NVME_CLI_TOPO_MULTIPATH */
+		nvme_subsystem_for_each_ctrl(s, c) {
+			nvme_ctrl_for_each_ns(c, n) {
+				c = nvme_ns_get_ctrl(n);
+
+				printf(" +- %s (ns %d)\n",
+						nvme_ns_get_name(n),
+						nvme_ns_get_nsid(n));
+				printf(" \\\n");
+				printf("  +- %s %s %s %s\n",
+						nvme_ctrl_get_name(c),
+						nvme_ctrl_get_transport(c),
+						nvme_ctrl_get_address(c),
+						nvme_ctrl_get_state(c));
+			}
+		}
 	}
 }
 
@@ -5717,6 +5788,11 @@ static void stdout_topology_ctrl(nvme_root_t r)
 	stdout_simple_topology(r, NVME_CLI_TOPO_CTRL);
 }
 
+static void stdout_topology_multipath(nvme_root_t r)
+{
+	stdout_simple_topology(r, NVME_CLI_TOPO_MULTIPATH);
+}
+
 static void stdout_message(bool error, const char *msg, va_list ap)
 {
 	vfprintf(error ? stderr : stdout, msg, ap);
@@ -6286,6 +6362,7 @@ static struct print_ops stdout_print_ops = {
 	.print_nvme_subsystem_list	= stdout_subsystem_list,
 	.topology_ctrl			= stdout_topology_ctrl,
 	.topology_namespace		= stdout_topology_namespace,
+	.topology_multipath		= stdout_topology_multipath,
 
 	/* status and error messages */
 	.connect_msg			= stdout_connect_msg,
diff --git a/nvme-print.c b/nvme-print.c
index 3a71dffc..473a6814 100644
--- a/nvme-print.c
+++ b/nvme-print.c
@@ -1551,8 +1551,10 @@ void nvme_show_topology(nvme_root_t r,
 {
 	if (ranking == NVME_CLI_TOPO_NAMESPACE)
 		nvme_print(topology_namespace, flags, r);
-	else
+	else if (ranking == NVME_CLI_TOPO_CTRL)
 		nvme_print(topology_ctrl, flags, r);
+	else
+		nvme_print(topology_multipath, flags, r);
 }
 
 void nvme_show_message(bool error, const char *msg, ...)
diff --git a/nvme-print.h b/nvme-print.h
index 4ccf5e63..0f23b711 100644
--- a/nvme-print.h
+++ b/nvme-print.h
@@ -106,6 +106,7 @@ struct print_ops {
 	void (*print_nvme_subsystem_list)(nvme_root_t r, bool show_ana);
 	void (*topology_ctrl)(nvme_root_t r);
 	void (*topology_namespace)(nvme_root_t r);
+	void (*topology_multipath)(nvme_root_t r);
 
 	/* status and error messages */
 	void (*connect_msg)(nvme_ctrl_t c);
diff --git a/nvme.c b/nvme.c
index f17a8091..41d98a78 100644
--- a/nvme.c
+++ b/nvme.c
@@ -10179,7 +10179,7 @@ static int tls_key(int argc, char **argv, struct command *command, struct plugin
 static int show_topology_cmd(int argc, char **argv, struct command *command, struct plugin *plugin)
 {
 	const char *desc = "Show the topology\n";
-	const char *ranking = "Ranking order: namespace|ctrl";
+	const char *ranking = "Ranking order: namespace|ctrl|multipath";
 	nvme_print_flags_t flags;
 	_cleanup_nvme_root_ nvme_root_t r = NULL;
 	char *devname = NULL;
@@ -10215,6 +10215,8 @@ static int show_topology_cmd(int argc, char **argv, struct command *command, str
 		rank = NVME_CLI_TOPO_NAMESPACE;
 	} else if (!strcmp(cfg.ranking, "ctrl")) {
 		rank = NVME_CLI_TOPO_CTRL;
+	} else if (!strcmp(cfg.ranking, "multipath")) {
+		rank = NVME_CLI_TOPO_MULTIPATH;
 	} else {
 		nvme_show_error("Invalid ranking argument: %s", cfg.ranking);
 		return -EINVAL;
diff --git a/nvme.h b/nvme.h
index 248c5c7c..f02f39ca 100644
--- a/nvme.h
+++ b/nvme.h
@@ -46,6 +46,7 @@ typedef uint32_t nvme_print_flags_t;
 enum nvme_cli_topo_ranking {
 	NVME_CLI_TOPO_NAMESPACE,
 	NVME_CLI_TOPO_CTRL,
+	NVME_CLI_TOPO_MULTIPATH,
 };
 
 #define SYS_NVME "/sys/class/nvme"
-- 
2.51.0