Linux-NVME Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Nilay Shroff <nilay@linux.ibm.com>
To: linux-nvme@lists.infradead.org
Cc: dwagner@suse.de, sagi@grimberg.me, hare@suse.de,
	kbusch@kernel.org, hch@lst.de, gjoyce@linux.ibm.com,
	chaitanyak@nvidia.com, Nilay Shroff <nilay@linux.ibm.com>
Subject: [PATCH v2 4/4] fabrics: determine --nr-io-queues when not explicitly specified
Date: Mon, 21 Sep 2026 13:25:02 +0530	[thread overview]
Message-ID: <20260921075511.2108561-5-nilay@linux.ibm.com> (raw)
In-Reply-To: <20260921075511.2108561-1-nilay@linux.ibm.com>

When performing an NVMe/TCP connect operation, if the user does not
explicitly specify --nr-io-queues, determine the default value using:

	if (nr_hw_queues > 1)
		min(nr_hw_queues, num_online_cpus)

Here, nr_hw_queues represents the number of hardware queues currently
configured on the NIC used for the NVMe/TCP connection. If nr_hw_queues
cannot be retrieved, or the NIC only supports one hardware queue, do
not calculate nr-io-queues and let the host determine the value.

Use the shr_route_get_egress_iface() and shr_netdev_get_hw_queues()
helpers to determine the egress netdev and retrieve its hardware
queue count. Use get_nprocs() to determine the number of online CPUs.

Apply the same logic when connecting to the target using a config INI
file.

Signed-off-by: Nilay Shroff <nilay@linux.ibm.com>
---
 src/fabrics.c | 79 +++++++++++++++++++++++++++++++++++++++++++++++++--
 1 file changed, 76 insertions(+), 3 deletions(-)

diff --git a/src/fabrics.c b/src/fabrics.c
index f5f1327ad..f5e71fdb8 100644
--- a/src/fabrics.c
+++ b/src/fabrics.c
@@ -19,17 +19,20 @@
  * Fabrics specification standard.
  */
 
+#include <ccan/minmax/minmax.h>
 #include <errno.h>
 #include <fcntl.h>
 #include <getopt.h>
 #include <inttypes.h>
 #include <libgen.h>
+#include <net/if.h>
 #include <stdbool.h>
 #include <stddef.h>
 #include <stdint.h>
 #include <stdio.h>
 #include <stdlib.h>
 #include <sys/stat.h>
+#include <sys/sysinfo.h>
 #include <sys/types.h>
 #include <time.h>
 #include <unistd.h>
@@ -49,6 +52,7 @@
 #include <ccan/endian/endian.h>
 #include <ccan/str/str.h>
 #include <shared/io-util.h>
+#include <shared/net-util.h>
 #include <shared/sig-util.h>
 
 #include "cleanup.h"
@@ -424,6 +428,50 @@ static int build_conn_tid(const struct libnvmf_config_conn *conn,
 			hostnqn, hostid, tid);
 }
 
+/*
+ * Return the number of I/O queues on success. Return 0 if the number
+ * cannot be determined or the NIC has only one hardware queue, in
+ * which case the host should determine the number of I/O queues.
+ */
+static int fabrics_connect_nr_io_queues(const char *transport,
+					const char *traddr,
+					const char *host_traddr,
+					const char *host_iface)
+{
+	uint32_t combined_count, tx_count, rx_count;
+	char ifname[IF_NAMESIZE] = {};
+	int nr_cpus, nr_hw_queues = 0;
+
+	if (strcmp(transport, "tcp"))
+		return 0;
+
+	nr_cpus = get_nprocs();
+	if (nr_cpus <= 0)
+		return 0;
+
+	if (host_iface && host_iface[0]) {
+		strncpy(ifname, host_iface, IF_NAMESIZE - 1);
+	} else {
+		if (shr_route_get_egress_iface(host_traddr, traddr,
+				ifname, IF_NAMESIZE))
+			return 0;
+	}
+
+	if (shr_netdev_get_hw_queues(ifname, &combined_count,
+			&tx_count, &rx_count))
+		return 0;
+
+	if (combined_count)
+		nr_hw_queues = combined_count;
+	else if (tx_count && rx_count)
+		nr_hw_queues = min(tx_count, rx_count);
+
+	if (nr_hw_queues <= 1)
+		return 0;
+
+	return min(nr_cpus, nr_hw_queues);
+}
+
 /* libnvmf_config_conn_for_each() callback: settle addressing/identity,
  * check exclusion, then discover or connect.
  */
@@ -435,6 +483,8 @@ static void consume_conn(const struct libnvmf_config_conn *conn,
 	struct hook_fabrics_data hfd = { .flags = st->flags, .raw = st->raw };
 	__cleanup_nvmf_context struct libnvmf_context *fctx = NULL;
 	__cleanup_nvmf_tid struct libnvmf_tid *tid = NULL;
+	const struct libnvmf_params *params;
+	const char *key = "nr-io-queues";
 	int err;
 
 	if (st->mode == CONSUME_ROLE_BASED && !is_dc && !st->connect)
@@ -460,9 +510,29 @@ static void consume_conn(const struct libnvmf_config_conn *conn,
 		goto record_err;
 
 	err = libnvmf_context_set_connection_from_tid(fctx, tid);
-	if (!err)
-		err = libnvmf_context_apply_params(fctx,
-				libnvmf_config_conn_get_params(conn));
+	if (err)
+		goto record_err;
+
+	params = libnvmf_config_conn_get_params(conn);
+	if (!libnvmf_params_get(params, key)) {
+		int nr_io_queues;
+
+		nr_io_queues = fabrics_connect_nr_io_queues(
+				libnvmf_config_conn_get_transport(conn),
+				libnvmf_config_conn_get_traddr(conn),
+				libnvmf_config_conn_get_host_traddr(conn),
+				libnvmf_config_conn_get_host_iface(conn));
+
+		if (nr_io_queues) {
+			char val[32];
+
+			snprintf(val, sizeof(val), "%d", nr_io_queues);
+			libnvmf_params_set((struct libnvmf_params *)params,
+					key, val);
+		}
+	}
+
+	err = libnvmf_context_apply_params(fctx, params);
 	if (err)
 		goto record_err;
 
@@ -1064,6 +1134,9 @@ int fabrics_connect(const char *desc, int argc, char **argv)
 		return ret;
 	fa.traddr = resolved_traddr;
 
+	if (!fa.nr_io_queues)
+		fa.nr_io_queues = fabrics_connect_nr_io_queues(fa.transport,
+				      fa.traddr, fa.host_traddr, fa.host_iface);
 do_connect:
 	ret = nvme_create_global_ctx_hostnqn(&ctx,
 		fa.hostnqn, fa.hostid, &hnqn, &hid);
-- 
2.53.0



      parent reply	other threads:[~2026-09-21  7:56 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-09-21  7:54 [PATCH v2 0/4] nvme-cli: NIC topology aware I/O queue scaling Nilay Shroff
2026-09-21  7:54 ` [PATCH v2 1/4] shared/net-util-linux: add support for retrieving egress netdev Nilay Shroff
2026-09-21  7:55 ` [PATCH v2 2/4] shared/net-util-linux: add support for retrieving NIC h/w queues Nilay Shroff
2026-09-21  7:55 ` [PATCH v2 3/4] shared: add more test coverage for net-util Nilay Shroff
2026-09-21  7:55 ` Nilay Shroff [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20260921075511.2108561-5-nilay@linux.ibm.com \
    --to=nilay@linux.ibm.com \
    --cc=chaitanyak@nvidia.com \
    --cc=dwagner@suse.de \
    --cc=gjoyce@linux.ibm.com \
    --cc=hare@suse.de \
    --cc=hch@lst.de \
    --cc=kbusch@kernel.org \
    --cc=linux-nvme@lists.infradead.org \
    --cc=sagi@grimberg.me \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox