Skip to contentSkip to conversation

NVMe cleanups for 5.8

PATCH In review Max Gurtovoy 4 participants 23 messages Jun 2, 2020 – Jun 10, 2020

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

cover NVMe cleanups for 5.8

This series introduce some cleanups and code/logic duplication for host and target side. It introduces a new header file nvme-types.h that will be used by both host and target drivers. For now it will include mainly mapping structures from types to names, but in the future it can include more common logic between host/target drivers (as done for specific TCP and RDMA transports). The series continues with adding a flags member to nvmet_fabrics_ops structure and removes the need for adding dedicated member for each feature in this struct.

Max Gurtovoy (5):
nvme: introduce nvme-types header file
nvme: use nvme_ana_type_to_name to get state string
nvme: replace transport name with trtype enum for ops
nvmet-tcp: remove has_keyed_sgls initialization
nvmet: introduce flags member in nvmet_fabrics_ops

 drivers/nvme/host/fabrics.c     |  3 +-
 drivers/nvme/host/fabrics.h     |  4 +-
 drivers/nvme/host/fc.c          |  2 +-
 drivers/nvme/host/multipath.c   | 16 ++-----
 drivers/nvme/host/nvme.h        |  1 +
 drivers/nvme/host/rdma.c        |  2 +-
 drivers/nvme/host/tcp.c         |  2 +-
 drivers/nvme/target/admin-cmd.c |  2 +-
 drivers/nvme/target/configfs.c  | 92 ++++++++-------------------------------
 drivers/nvme/target/core.c      |  2 +-
 drivers/nvme/target/discovery.c |  2 +-
 drivers/nvme/target/loop.c      |  2 +-
 drivers/nvme/target/nvmet.h     |  6 ++-
 drivers/nvme/target/rdma.c      |  3 +-
 drivers/nvme/target/tcp.c       |  1 -
 include/linux/nvme-types.h      | 95 +++++++++++++++++++++++++++++++++++++++++
 16 files changed, 132 insertions(+), 103 deletions(-)
 create mode 100644 include/linux/nvme-types.h
Signature · 7 lines
1.8.3.1


_______________________________________________
linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme
#1 lore raw

PATCH 1/5 nvme: introduce nvme-types header file

Centralize the mapping between name and type to a common header file instead of duplicating logic in both NVMe host and target drivers.

---
 drivers/nvme/target/configfs.c | 92 ++++++++--------------------------------
 drivers/nvme/target/nvmet.h    |  1 +
 include/linux/nvme-types.h     | 95 ++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 113 insertions(+), 75 deletions(-)
 create mode 100644 include/linux/nvme-types.h
drivers/nvme/target/configfs.c +17 -75
index 419e0d4..daeb17e 100644
@@ -20,27 +20,6 @@
 static LIST_HEAD(nvmet_ports_list);
 struct list_head *nvmet_ports = &nvmet_ports_list;
 
-struct nvmet_type_name_map {
-	u8		type;
-	const char	*name;
-};
-
-static struct nvmet_type_name_map nvmet_transport[] = {
-	{ NVMF_TRTYPE_RDMA,	"rdma" },
-	{ NVMF_TRTYPE_FC,	"fc" },
-	{ NVMF_TRTYPE_TCP,	"tcp" },
-	{ NVMF_TRTYPE_LOOP,	"loop" },
-};
-
-static const struct nvmet_type_name_map nvmet_addr_family[] = {
-	{ NVMF_ADDR_FAMILY_PCI,		"pcie" },
-	{ NVMF_ADDR_FAMILY_IP4,		"ipv4" },
-	{ NVMF_ADDR_FAMILY_IP6,		"ipv6" },
-	{ NVMF_ADDR_FAMILY_IB,		"ib" },
-	{ NVMF_ADDR_FAMILY_FC,		"fc" },
-	{ NVMF_ADDR_FAMILY_LOOP,	"loop" },
-};
-
 static bool nvmet_is_port_enabled(struct nvmet_port *p, const char *caller)
 {
 	if (p->enabled)
@@ -56,14 +35,8 @@ static bool nvmet_is_port_enabled(struct nvmet_port *p, const char *caller)
 static ssize_t nvmet_addr_adrfam_show(struct config_item *item, char *page)
 {
 	u8 adrfam = to_nvmet_port(item)->disc_addr.adrfam;
-	int i;
-
-	for (i = 1; i < ARRAY_SIZE(nvmet_addr_family); i++) {
-		if (nvmet_addr_family[i].type == adrfam)
-			return sprintf(page, "%s\n", nvmet_addr_family[i].name);
-	}
 
-	return sprintf(page, "\n");
+	return sprintf(page, "%s\n", nvme_adrfam_type_to_name(adrfam));
 }
 
 static ssize_t nvmet_addr_adrfam_store(struct config_item *item,
@@ -75,8 +48,8 @@ static ssize_t nvmet_addr_adrfam_store(struct config_item *item,
 	if (nvmet_is_port_enabled(port, __func__))
 		return -EACCES;
 
-	for (i = 1; i < ARRAY_SIZE(nvmet_addr_family); i++) {
-		if (sysfs_streq(page, nvmet_addr_family[i].name))
+	for (i = 1; i < ARRAY_SIZE(nvme_addr_family); i++) {
+		if (sysfs_streq(page, nvme_addr_family[i].name))
 			goto found;
 	}
 
@@ -84,7 +57,7 @@ static ssize_t nvmet_addr_adrfam_store(struct config_item *item,
 	return -EINVAL;
 
 found:
-	port->disc_addr.adrfam = nvmet_addr_family[i].type;
+	port->disc_addr.adrfam = nvme_addr_family[i].type;
 	return count;
 }
 
@@ -148,24 +121,12 @@ static ssize_t nvmet_addr_traddr_store(struct config_item *item,
 
 CONFIGFS_ATTR(nvmet_, addr_traddr);
 
-static const struct nvmet_type_name_map nvmet_addr_treq[] = {
-	{ NVMF_TREQ_NOT_SPECIFIED,	"not specified" },
-	{ NVMF_TREQ_REQUIRED,		"required" },
-	{ NVMF_TREQ_NOT_REQUIRED,	"not required" },
-};
-
 static ssize_t nvmet_addr_treq_show(struct config_item *item, char *page)
 {
 	u8 treq = to_nvmet_port(item)->disc_addr.treq &
 		NVME_TREQ_SECURE_CHANNEL_MASK;
-	int i;
-
-	for (i = 0; i < ARRAY_SIZE(nvmet_addr_treq); i++) {
-		if (treq == nvmet_addr_treq[i].type)
-			return sprintf(page, "%s\n", nvmet_addr_treq[i].name);
-	}
 
-	return sprintf(page, "\n");
+	return sprintf(page, "%s\n", nvme_treq_type_to_name(treq));
 }
 
 static ssize_t nvmet_addr_treq_store(struct config_item *item,
@@ -178,8 +139,8 @@ static ssize_t nvmet_addr_treq_store(struct config_item *item,
 	if (nvmet_is_port_enabled(port, __func__))
 		return -EACCES;
 
-	for (i = 0; i < ARRAY_SIZE(nvmet_addr_treq); i++) {
-		if (sysfs_streq(page, nvmet_addr_treq[i].name))
+	for (i = 0; i < ARRAY_SIZE(nvme_addr_treq); i++) {
+		if (sysfs_streq(page, nvme_addr_treq[i].name))
 			goto found;
 	}
 
@@ -187,7 +148,7 @@ static ssize_t nvmet_addr_treq_store(struct config_item *item,
 	return -EINVAL;
 
 found:
-	treq |= nvmet_addr_treq[i].type;
+	treq |= nvme_addr_treq[i].type;
 	port->disc_addr.treq = treq;
 	return count;
 }
@@ -282,14 +243,9 @@ static ssize_t nvmet_addr_trtype_show(struct config_item *item,
 		char *page)
 {
 	struct nvmet_port *port = to_nvmet_port(item);
-	int i;
-
-	for (i = 0; i < ARRAY_SIZE(nvmet_transport); i++) {
-		if (port->disc_addr.trtype == nvmet_transport[i].type)
-			return sprintf(page, "%s\n", nvmet_transport[i].name);
-	}
 
-	return sprintf(page, "\n");
+	return sprintf(page, "%s\n",
+		       nvme_trtype_to_name(port->disc_addr.trtype));
 }
 
 static void nvmet_port_init_tsas_rdma(struct nvmet_port *port)
@@ -308,8 +264,8 @@ static ssize_t nvmet_addr_trtype_store(struct config_item *item,
 	if (nvmet_is_port_enabled(port, __func__))
 		return -EACCES;
 
-	for (i = 0; i < ARRAY_SIZE(nvmet_transport); i++) {
-		if (sysfs_streq(page, nvmet_transport[i].name))
+	for (i = 0; i < ARRAY_SIZE(nvme_transport); i++) {
+		if (sysfs_streq(page, nvme_transport[i].name))
 			goto found;
 	}
 
@@ -318,7 +274,7 @@ static ssize_t nvmet_addr_trtype_store(struct config_item *item,
 
 found:
 	memset(&port->disc_addr.tsas, 0, NVMF_TSAS_SIZE);
-	port->disc_addr.trtype = nvmet_transport[i].type;
+	port->disc_addr.trtype = nvme_transport[i].type;
 	if (port->disc_addr.trtype == NVMF_TRTYPE_RDMA)
 		nvmet_port_init_tsas_rdma(port);
 	return count;
@@ -1227,27 +1183,13 @@ static struct config_group *nvmet_referral_make(
 	.ct_group_ops	= &nvmet_referral_group_ops,
 };
 
-static struct nvmet_type_name_map nvmet_ana_state[] = {
-	{ NVME_ANA_OPTIMIZED,		"optimized" },
-	{ NVME_ANA_NONOPTIMIZED,	"non-optimized" },
-	{ NVME_ANA_INACCESSIBLE,	"inaccessible" },
-	{ NVME_ANA_PERSISTENT_LOSS,	"persistent-loss" },
-	{ NVME_ANA_CHANGE,		"change" },
-};
-
 static ssize_t nvmet_ana_group_ana_state_show(struct config_item *item,
 		char *page)
 {
 	struct nvmet_ana_group *grp = to_ana_group(item);
 	enum nvme_ana_state state = grp->port->ana_state[grp->grpid];
-	int i;
-
-	for (i = 0; i < ARRAY_SIZE(nvmet_ana_state); i++) {
-		if (state == nvmet_ana_state[i].type)
-			return sprintf(page, "%s\n", nvmet_ana_state[i].name);
-	}
 
-	return sprintf(page, "\n");
+	return sprintf(page, "%s\n", nvme_ana_type_to_name(state));
 }
 
 static ssize_t nvmet_ana_group_ana_state_store(struct config_item *item,
@@ -1257,8 +1199,8 @@ static ssize_t nvmet_ana_group_ana_state_store(struct config_item *item,
 	enum nvme_ana_state *ana_state = grp->port->ana_state;
 	int i;
 
-	for (i = 0; i < ARRAY_SIZE(nvmet_ana_state); i++) {
-		if (sysfs_streq(page, nvmet_ana_state[i].name))
+	for (i = 0; i < ARRAY_SIZE(nvme_ana_states); i++) {
+		if (sysfs_streq(page, nvme_ana_states[i].name))
 			goto found;
 	}
 
@@ -1267,7 +1209,7 @@ static ssize_t nvmet_ana_group_ana_state_store(struct config_item *item,
 
 found:
 	down_write(&nvmet_ana_sem);
-	ana_state[grp->grpid] = (enum nvme_ana_state) nvmet_ana_state[i].type;
+	ana_state[grp->grpid] = (enum nvme_ana_state) nvme_ana_states[i].type;
 	nvmet_ana_chgcnt++;
 	up_write(&nvmet_ana_sem);
 	nvmet_port_send_ana_event(grp->port);
drivers/nvme/target/nvmet.h +1 -0
index 8096912..ce57d5f 100644
@@ -15,6 +15,7 @@
 #include <linux/mutex.h>
 #include <linux/uuid.h>
 #include <linux/nvme.h>
+#include <linux/nvme-types.h>
 #include <linux/configfs.h>
 #include <linux/rcupdate.h>
 #include <linux/blkdev.h>
include/linux/nvme-types.h new file +95 -0

95 lines not on this page · raw message

Signature · 7 lines
1.8.3.1


_______________________________________________
linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme
Reviewed-by on v1 1/5
#2 lore raw

Re: [PATCH 1/5] nvme: introduce nvme-types header file

Quoted 316 lines · Max Gurtovoy

Centralize the mapping between name and type to a common header file instead of duplicating logic in both NVMe host and target drivers.

---
  drivers/nvme/target/configfs.c | 92 ++++++++--------------------------------
  drivers/nvme/target/nvmet.h    |  1 +
  include/linux/nvme-types.h     | 95 ++++++++++++++++++++++++++++++++++++++++++
  3 files changed, 113 insertions(+), 75 deletions(-)

create mode 100644 include/linux/nvme-types.h

drivers/nvme/target/configfs.c +17 -75
index 419e0d4..daeb17e 100644
@@ -20,27 +20,6 @@
  static LIST_HEAD(nvmet_ports_list);
  struct list_head *nvmet_ports = &nvmet_ports_list;
 
-struct nvmet_type_name_map {
-	u8		type;
-	const char	*name;
-};
-
-static struct nvmet_type_name_map nvmet_transport[] = {
-	{ NVMF_TRTYPE_RDMA,	"rdma" },
-	{ NVMF_TRTYPE_FC,	"fc" },
-	{ NVMF_TRTYPE_TCP,	"tcp" },
-	{ NVMF_TRTYPE_LOOP,	"loop" },
-};
-
-static const struct nvmet_type_name_map nvmet_addr_family[] = {
-	{ NVMF_ADDR_FAMILY_PCI,		"pcie" },
-	{ NVMF_ADDR_FAMILY_IP4,		"ipv4" },
-	{ NVMF_ADDR_FAMILY_IP6,		"ipv6" },
-	{ NVMF_ADDR_FAMILY_IB,		"ib" },
-	{ NVMF_ADDR_FAMILY_FC,		"fc" },
-	{ NVMF_ADDR_FAMILY_LOOP,	"loop" },
-};
-
  static bool nvmet_is_port_enabled(struct nvmet_port *p, const char *caller)
  {
  	if (p->enabled)
@@ -56,14 +35,8 @@ static bool nvmet_is_port_enabled(struct nvmet_port *p, const char *caller)
  static ssize_t nvmet_addr_adrfam_show(struct config_item *item, char *page)
  {
  	u8 adrfam = to_nvmet_port(item)->disc_addr.adrfam;
-	int i;
-
-	for (i = 1; i < ARRAY_SIZE(nvmet_addr_family); i++) {
-		if (nvmet_addr_family[i].type == adrfam)
-			return sprintf(page, "%s\n", nvmet_addr_family[i].name);
-	}
 
-	return sprintf(page, "\n");
+	return sprintf(page, "%s\n", nvme_adrfam_type_to_name(adrfam));
  }
 
  static ssize_t nvmet_addr_adrfam_store(struct config_item *item,
@@ -75,8 +48,8 @@ static ssize_t nvmet_addr_adrfam_store(struct config_item *item,
  	if (nvmet_is_port_enabled(port, __func__))
  		return -EACCES;
 
-	for (i = 1; i < ARRAY_SIZE(nvmet_addr_family); i++) {
-		if (sysfs_streq(page, nvmet_addr_family[i].name))
+	for (i = 1; i < ARRAY_SIZE(nvme_addr_family); i++) {
+		if (sysfs_streq(page, nvme_addr_family[i].name))
  			goto found;
  	}
 
@@ -84,7 +57,7 @@ static ssize_t nvmet_addr_adrfam_store(struct config_item *item,
  	return -EINVAL;
 
  found:
-	port->disc_addr.adrfam = nvmet_addr_family[i].type;
+	port->disc_addr.adrfam = nvme_addr_family[i].type;
  	return count;
  }
 
@@ -148,24 +121,12 @@ static ssize_t nvmet_addr_traddr_store(struct config_item *item,
 
  CONFIGFS_ATTR(nvmet_, addr_traddr);
 
-static const struct nvmet_type_name_map nvmet_addr_treq[] = {
-	{ NVMF_TREQ_NOT_SPECIFIED,	"not specified" },
-	{ NVMF_TREQ_REQUIRED,		"required" },
-	{ NVMF_TREQ_NOT_REQUIRED,	"not required" },
-};
-
  static ssize_t nvmet_addr_treq_show(struct config_item *item, char *page)
  {
  	u8 treq = to_nvmet_port(item)->disc_addr.treq &
  		NVME_TREQ_SECURE_CHANNEL_MASK;
-	int i;
-
-	for (i = 0; i < ARRAY_SIZE(nvmet_addr_treq); i++) {
-		if (treq == nvmet_addr_treq[i].type)
-			return sprintf(page, "%s\n", nvmet_addr_treq[i].name);
-	}
 
-	return sprintf(page, "\n");
+	return sprintf(page, "%s\n", nvme_treq_type_to_name(treq));
  }
 
  static ssize_t nvmet_addr_treq_store(struct config_item *item,
@@ -178,8 +139,8 @@ static ssize_t nvmet_addr_treq_store(struct config_item *item,
  	if (nvmet_is_port_enabled(port, __func__))
  		return -EACCES;
 
-	for (i = 0; i < ARRAY_SIZE(nvmet_addr_treq); i++) {
-		if (sysfs_streq(page, nvmet_addr_treq[i].name))
+	for (i = 0; i < ARRAY_SIZE(nvme_addr_treq); i++) {
+		if (sysfs_streq(page, nvme_addr_treq[i].name))
  			goto found;
  	}
 
@@ -187,7 +148,7 @@ static ssize_t nvmet_addr_treq_store(struct config_item *item,
  	return -EINVAL;
 
  found:
-	treq |= nvmet_addr_treq[i].type;
+	treq |= nvme_addr_treq[i].type;
  	port->disc_addr.treq = treq;
  	return count;
  }
@@ -282,14 +243,9 @@ static ssize_t nvmet_addr_trtype_show(struct config_item *item,
  		char *page)
  {
  	struct nvmet_port *port = to_nvmet_port(item);
-	int i;
-
-	for (i = 0; i < ARRAY_SIZE(nvmet_transport); i++) {
-		if (port->disc_addr.trtype == nvmet_transport[i].type)
-			return sprintf(page, "%s\n", nvmet_transport[i].name);
-	}
 
-	return sprintf(page, "\n");
+	return sprintf(page, "%s\n",
+		       nvme_trtype_to_name(port->disc_addr.trtype));
  }
 
  static void nvmet_port_init_tsas_rdma(struct nvmet_port *port)
@@ -308,8 +264,8 @@ static ssize_t nvmet_addr_trtype_store(struct config_item *item,
  	if (nvmet_is_port_enabled(port, __func__))
  		return -EACCES;
 
-	for (i = 0; i < ARRAY_SIZE(nvmet_transport); i++) {
-		if (sysfs_streq(page, nvmet_transport[i].name))
+	for (i = 0; i < ARRAY_SIZE(nvme_transport); i++) {
+		if (sysfs_streq(page, nvme_transport[i].name))
  			goto found;
  	}
 
@@ -318,7 +274,7 @@ static ssize_t nvmet_addr_trtype_store(struct config_item *item,
 
  found:
  	memset(&port->disc_addr.tsas, 0, NVMF_TSAS_SIZE);
-	port->disc_addr.trtype = nvmet_transport[i].type;
+	port->disc_addr.trtype = nvme_transport[i].type;
  	if (port->disc_addr.trtype == NVMF_TRTYPE_RDMA)
  		nvmet_port_init_tsas_rdma(port);
  	return count;
@@ -1227,27 +1183,13 @@ static struct config_group *nvmet_referral_make(
  	.ct_group_ops	= &nvmet_referral_group_ops,
  };
 
-static struct nvmet_type_name_map nvmet_ana_state[] = {
-	{ NVME_ANA_OPTIMIZED,		"optimized" },
-	{ NVME_ANA_NONOPTIMIZED,	"non-optimized" },
-	{ NVME_ANA_INACCESSIBLE,	"inaccessible" },
-	{ NVME_ANA_PERSISTENT_LOSS,	"persistent-loss" },
-	{ NVME_ANA_CHANGE,		"change" },
-};
-
  static ssize_t nvmet_ana_group_ana_state_show(struct config_item *item,
  		char *page)
  {
  	struct nvmet_ana_group *grp = to_ana_group(item);
  	enum nvme_ana_state state = grp->port->ana_state[grp->grpid];
-	int i;
-
-	for (i = 0; i < ARRAY_SIZE(nvmet_ana_state); i++) {
-		if (state == nvmet_ana_state[i].type)
-			return sprintf(page, "%s\n", nvmet_ana_state[i].name);
-	}
 
-	return sprintf(page, "\n");
+	return sprintf(page, "%s\n", nvme_ana_type_to_name(state));
  }
 
  static ssize_t nvmet_ana_group_ana_state_store(struct config_item *item,
@@ -1257,8 +1199,8 @@ static ssize_t nvmet_ana_group_ana_state_store(struct config_item *item,
  	enum nvme_ana_state *ana_state = grp->port->ana_state;
  	int i;
 
-	for (i = 0; i < ARRAY_SIZE(nvmet_ana_state); i++) {
-		if (sysfs_streq(page, nvmet_ana_state[i].name))
+	for (i = 0; i < ARRAY_SIZE(nvme_ana_states); i++) {
+		if (sysfs_streq(page, nvme_ana_states[i].name))
  			goto found;
  	}
 
@@ -1267,7 +1209,7 @@ static ssize_t nvmet_ana_group_ana_state_store(struct config_item *item,
 
  found:
  	down_write(&nvmet_ana_sem);
-	ana_state[grp->grpid] = (enum nvme_ana_state) nvmet_ana_state[i].type;
+	ana_state[grp->grpid] = (enum nvme_ana_state) nvme_ana_states[i].type;
  	nvmet_ana_chgcnt++;
  	up_write(&nvmet_ana_sem);
  	nvmet_port_send_ana_event(grp->port);
drivers/nvme/target/nvmet.h +1 -0
index 8096912..ce57d5f 100644
@@ -15,6 +15,7 @@
  #include <linux/mutex.h>
  #include <linux/uuid.h>
  #include <linux/nvme.h>
+#include <linux/nvme-types.h>
  #include <linux/configfs.h>
  #include <linux/rcupdate.h>
  #include <linux/blkdev.h>
include/linux/nvme-types.h new file +85 -0
new file mode 100644
index 0000000..38ed980
@@ -0,0 +1,95 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Copyright (c) 2020 Mellanox Technologies. All rights reserved.
+ */
+
+#ifndef _LINUX_NVME_TYPES_H
+#define _LINUX_NVME_TYPES_H
+
+#include <linux/kernel.h>
+#include <linux/nvme.h>
+
+struct nvme_type_name_map {
+	u8		type;
+	const char	*name;
+};
+
+static const struct nvme_type_name_map nvme_transport[] = {
+	{ NVMF_TRTYPE_RDMA,	"rdma" },
+	{ NVMF_TRTYPE_FC,	"fc" },
+	{ NVMF_TRTYPE_TCP,	"tcp" },
+	{ NVMF_TRTYPE_LOOP,	"loop" },
+};
+
+static inline const char *nvme_trtype_to_name(u8 type)
+{
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(nvme_transport); i++) {
+		if (type == nvme_transport[i].type)
+			return nvme_transport[i].name;
+	}
+
+	return "invalid";
+}
+
+static const struct nvme_type_name_map nvme_addr_family[] = {
+	{ NVMF_ADDR_FAMILY_PCI,		"pcie" },
+	{ NVMF_ADDR_FAMILY_IP4,		"ipv4" },
+	{ NVMF_ADDR_FAMILY_IP6,		"ipv6" },
+	{ NVMF_ADDR_FAMILY_IB,		"ib" },
+	{ NVMF_ADDR_FAMILY_FC,		"fc" },
+	{ NVMF_ADDR_FAMILY_LOOP,	"loop" },
+};
+
+static inline const char *nvme_adrfam_type_to_name(u8 type)
+{
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(nvme_addr_family); i++) {
+		if (type == nvme_addr_family[i].type)
+			return nvme_addr_family[i].name;
+	}
+
+	return "invalid";
+}
+
+static const struct nvme_type_name_map nvme_addr_treq[] = {
+	{ NVMF_TREQ_NOT_SPECIFIED,	"not specified" },
+	{ NVMF_TREQ_REQUIRED,		"required" },
+	{ NVMF_TREQ_NOT_REQUIRED,	"not required" },
+};
+
+static inline const char *nvme_treq_type_to_name(u8 type)
+{
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(nvme_addr_treq); i++) {
+		if (type == nvme_addr_treq[i].type)
+			return nvme_addr_treq[i].name;
+	}
+
+	return "invalid";
+}
+
+static const struct nvme_type_name_map nvme_ana_states[] = {
+	{ NVME_ANA_OPTIMIZED,		"optimized" },
+	{ NVME_ANA_NONOPTIMIZED,	"non-optimized" },
+	{ NVME_ANA_INACCESSIBLE,	"inaccessible" },
+	{ NVME_ANA_PERSISTENT_LOSS,	"persistent-loss" },
+	{ NVME_ANA_CHANGE,		"change" },
+};
+
+static inline const char *nvme_ana_type_to_name(u8 type)
+{
+	int i;

include/linux/nvme-types.h line 95

Quoted 10 lines
include/linux/nvme-types.h new file +10 -0
+
+	for (i = 0; i < ARRAY_SIZE(nvme_ana_states); i++) {
+		if (type == nvme_ana_states[i].type)
+			return nvme_ana_states[i].name;
+	}
+
+	return "invalid";
+}
+
+#endif /* _LINUX_NVME_TYPES_H */

Looks Good.

Signature · 6 lines
Himanshu Madhani                     Oracle Linux Engineering

_______________________________________________
linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme
maintainer
#3 lore raw

Re: [PATCH 1/5] nvme: introduce nvme-types header file

In reply to Max Gurtovoy
Quoted 253 lines · Max Gurtovoy

Centralize the mapping between name and type to a common header file instead of duplicating logic in both NVMe host and target drivers.

---
  drivers/nvme/target/configfs.c | 92 ++++++++--------------------------------
  drivers/nvme/target/nvmet.h    |  1 +
  include/linux/nvme-types.h     | 95 ++++++++++++++++++++++++++++++++++++++++++
  3 files changed, 113 insertions(+), 75 deletions(-)

create mode 100644 include/linux/nvme-types.h

drivers/nvme/target/configfs.c +17 -75
index 419e0d4..daeb17e 100644
@@ -20,27 +20,6 @@
  static LIST_HEAD(nvmet_ports_list);
  struct list_head *nvmet_ports = &nvmet_ports_list;
 
-struct nvmet_type_name_map {
-	u8		type;
-	const char	*name;
-};
-
-static struct nvmet_type_name_map nvmet_transport[] = {
-	{ NVMF_TRTYPE_RDMA,	"rdma" },
-	{ NVMF_TRTYPE_FC,	"fc" },
-	{ NVMF_TRTYPE_TCP,	"tcp" },
-	{ NVMF_TRTYPE_LOOP,	"loop" },
-};
-
-static const struct nvmet_type_name_map nvmet_addr_family[] = {
-	{ NVMF_ADDR_FAMILY_PCI,		"pcie" },
-	{ NVMF_ADDR_FAMILY_IP4,		"ipv4" },
-	{ NVMF_ADDR_FAMILY_IP6,		"ipv6" },
-	{ NVMF_ADDR_FAMILY_IB,		"ib" },
-	{ NVMF_ADDR_FAMILY_FC,		"fc" },
-	{ NVMF_ADDR_FAMILY_LOOP,	"loop" },
-};
-
  static bool nvmet_is_port_enabled(struct nvmet_port *p, const char *caller)
  {
  	if (p->enabled)
@@ -56,14 +35,8 @@ static bool nvmet_is_port_enabled(struct nvmet_port *p, const char *caller)
  static ssize_t nvmet_addr_adrfam_show(struct config_item *item, char *page)
  {
  	u8 adrfam = to_nvmet_port(item)->disc_addr.adrfam;
-	int i;
-
-	for (i = 1; i < ARRAY_SIZE(nvmet_addr_family); i++) {
-		if (nvmet_addr_family[i].type == adrfam)
-			return sprintf(page, "%s\n", nvmet_addr_family[i].name);
-	}
 
-	return sprintf(page, "\n");
+	return sprintf(page, "%s\n", nvme_adrfam_type_to_name(adrfam));
  }
 
  static ssize_t nvmet_addr_adrfam_store(struct config_item *item,
@@ -75,8 +48,8 @@ static ssize_t nvmet_addr_adrfam_store(struct config_item *item,
  	if (nvmet_is_port_enabled(port, __func__))
  		return -EACCES;
 
-	for (i = 1; i < ARRAY_SIZE(nvmet_addr_family); i++) {
-		if (sysfs_streq(page, nvmet_addr_family[i].name))
+	for (i = 1; i < ARRAY_SIZE(nvme_addr_family); i++) {
+		if (sysfs_streq(page, nvme_addr_family[i].name))
  			goto found;
  	}
 
@@ -84,7 +57,7 @@ static ssize_t nvmet_addr_adrfam_store(struct config_item *item,
  	return -EINVAL;
 
  found:
-	port->disc_addr.adrfam = nvmet_addr_family[i].type;
+	port->disc_addr.adrfam = nvme_addr_family[i].type;
  	return count;
  }
 
@@ -148,24 +121,12 @@ static ssize_t nvmet_addr_traddr_store(struct config_item *item,
 
  CONFIGFS_ATTR(nvmet_, addr_traddr);
 
-static const struct nvmet_type_name_map nvmet_addr_treq[] = {
-	{ NVMF_TREQ_NOT_SPECIFIED,	"not specified" },
-	{ NVMF_TREQ_REQUIRED,		"required" },
-	{ NVMF_TREQ_NOT_REQUIRED,	"not required" },
-};
-
  static ssize_t nvmet_addr_treq_show(struct config_item *item, char *page)
  {
  	u8 treq = to_nvmet_port(item)->disc_addr.treq &
  		NVME_TREQ_SECURE_CHANNEL_MASK;
-	int i;
-
-	for (i = 0; i < ARRAY_SIZE(nvmet_addr_treq); i++) {
-		if (treq == nvmet_addr_treq[i].type)
-			return sprintf(page, "%s\n", nvmet_addr_treq[i].name);
-	}
 
-	return sprintf(page, "\n");
+	return sprintf(page, "%s\n", nvme_treq_type_to_name(treq));
  }
 
  static ssize_t nvmet_addr_treq_store(struct config_item *item,
@@ -178,8 +139,8 @@ static ssize_t nvmet_addr_treq_store(struct config_item *item,
  	if (nvmet_is_port_enabled(port, __func__))
  		return -EACCES;
 
-	for (i = 0; i < ARRAY_SIZE(nvmet_addr_treq); i++) {
-		if (sysfs_streq(page, nvmet_addr_treq[i].name))
+	for (i = 0; i < ARRAY_SIZE(nvme_addr_treq); i++) {
+		if (sysfs_streq(page, nvme_addr_treq[i].name))
  			goto found;
  	}
 
@@ -187,7 +148,7 @@ static ssize_t nvmet_addr_treq_store(struct config_item *item,
  	return -EINVAL;
 
  found:
-	treq |= nvmet_addr_treq[i].type;
+	treq |= nvme_addr_treq[i].type;
  	port->disc_addr.treq = treq;
  	return count;
  }
@@ -282,14 +243,9 @@ static ssize_t nvmet_addr_trtype_show(struct config_item *item,
  		char *page)
  {
  	struct nvmet_port *port = to_nvmet_port(item);
-	int i;
-
-	for (i = 0; i < ARRAY_SIZE(nvmet_transport); i++) {
-		if (port->disc_addr.trtype == nvmet_transport[i].type)
-			return sprintf(page, "%s\n", nvmet_transport[i].name);
-	}
 
-	return sprintf(page, "\n");
+	return sprintf(page, "%s\n",
+		       nvme_trtype_to_name(port->disc_addr.trtype));
  }
 
  static void nvmet_port_init_tsas_rdma(struct nvmet_port *port)
@@ -308,8 +264,8 @@ static ssize_t nvmet_addr_trtype_store(struct config_item *item,
  	if (nvmet_is_port_enabled(port, __func__))
  		return -EACCES;
 
-	for (i = 0; i < ARRAY_SIZE(nvmet_transport); i++) {
-		if (sysfs_streq(page, nvmet_transport[i].name))
+	for (i = 0; i < ARRAY_SIZE(nvme_transport); i++) {
+		if (sysfs_streq(page, nvme_transport[i].name))
  			goto found;
  	}
 
@@ -318,7 +274,7 @@ static ssize_t nvmet_addr_trtype_store(struct config_item *item,
 
  found:
  	memset(&port->disc_addr.tsas, 0, NVMF_TSAS_SIZE);
-	port->disc_addr.trtype = nvmet_transport[i].type;
+	port->disc_addr.trtype = nvme_transport[i].type;
  	if (port->disc_addr.trtype == NVMF_TRTYPE_RDMA)
  		nvmet_port_init_tsas_rdma(port);
  	return count;
@@ -1227,27 +1183,13 @@ static struct config_group *nvmet_referral_make(
  	.ct_group_ops	= &nvmet_referral_group_ops,
  };
 
-static struct nvmet_type_name_map nvmet_ana_state[] = {
-	{ NVME_ANA_OPTIMIZED,		"optimized" },
-	{ NVME_ANA_NONOPTIMIZED,	"non-optimized" },
-	{ NVME_ANA_INACCESSIBLE,	"inaccessible" },
-	{ NVME_ANA_PERSISTENT_LOSS,	"persistent-loss" },
-	{ NVME_ANA_CHANGE,		"change" },
-};
-
  static ssize_t nvmet_ana_group_ana_state_show(struct config_item *item,
  		char *page)
  {
  	struct nvmet_ana_group *grp = to_ana_group(item);
  	enum nvme_ana_state state = grp->port->ana_state[grp->grpid];
-	int i;
-
-	for (i = 0; i < ARRAY_SIZE(nvmet_ana_state); i++) {
-		if (state == nvmet_ana_state[i].type)
-			return sprintf(page, "%s\n", nvmet_ana_state[i].name);
-	}
 
-	return sprintf(page, "\n");
+	return sprintf(page, "%s\n", nvme_ana_type_to_name(state));
  }
 
  static ssize_t nvmet_ana_group_ana_state_store(struct config_item *item,
@@ -1257,8 +1199,8 @@ static ssize_t nvmet_ana_group_ana_state_store(struct config_item *item,
  	enum nvme_ana_state *ana_state = grp->port->ana_state;
  	int i;
 
-	for (i = 0; i < ARRAY_SIZE(nvmet_ana_state); i++) {
-		if (sysfs_streq(page, nvmet_ana_state[i].name))
+	for (i = 0; i < ARRAY_SIZE(nvme_ana_states); i++) {
+		if (sysfs_streq(page, nvme_ana_states[i].name))
  			goto found;
  	}
 
@@ -1267,7 +1209,7 @@ static ssize_t nvmet_ana_group_ana_state_store(struct config_item *item,
 
  found:
  	down_write(&nvmet_ana_sem);
-	ana_state[grp->grpid] = (enum nvme_ana_state) nvmet_ana_state[i].type;
+	ana_state[grp->grpid] = (enum nvme_ana_state) nvme_ana_states[i].type;
  	nvmet_ana_chgcnt++;
  	up_write(&nvmet_ana_sem);
  	nvmet_port_send_ana_event(grp->port);
drivers/nvme/target/nvmet.h +1 -0
index 8096912..ce57d5f 100644
@@ -15,6 +15,7 @@
  #include <linux/mutex.h>
  #include <linux/uuid.h>
  #include <linux/nvme.h>
+#include <linux/nvme-types.h>
  #include <linux/configfs.h>
  #include <linux/rcupdate.h>
  #include <linux/blkdev.h>
include/linux/nvme-types.h new file +23 -0
new file mode 100644
index 0000000..38ed980
@@ -0,0 +1,95 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Copyright (c) 2020 Mellanox Technologies. All rights reserved.
+ */
+
+#ifndef _LINUX_NVME_TYPES_H
+#define _LINUX_NVME_TYPES_H
+
+#include <linux/kernel.h>
+#include <linux/nvme.h>
+
+struct nvme_type_name_map {
+	u8		type;
+	const char	*name;
+};
+
+static const struct nvme_type_name_map nvme_transport[] = {
+	{ NVMF_TRTYPE_RDMA,	"rdma" },
+	{ NVMF_TRTYPE_FC,	"fc" },
+	{ NVMF_TRTYPE_TCP,	"tcp" },
+	{ NVMF_TRTYPE_LOOP,	"loop" },
+};
+

include/linux/nvme-types.h line 33

Quoted 10 lines
include/linux/nvme-types.h new file +10 -0
+static inline const char *nvme_trtype_to_name(u8 type)
+{
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(nvme_transport); i++) {
+		if (type == nvme_transport[i].type)
+			return nvme_transport[i].name;
+	}
+
+	return "invalid";

Why does this need a loop?

maybe just:

	return (type < ARRAY_SIZE(nvme_transport) && nvme_transport[type]) ? 

nvme_transport[i].name : "invalid";

Same for the rest...

_______________________________________________
linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme

#4 lore raw

Re: [PATCH 1/5] nvme: introduce nvme-types header file

Quoted 25 lines · Sagi Grimberg
Quoted 12 lines
include/linux/nvme-types.h new file +12 -0
+struct nvme_type_name_map {
+    u8        type;
+    const char    *name;
+};
+
+static const struct nvme_type_name_map nvme_transport[] = {
+    { NVMF_TRTYPE_RDMA,    "rdma" },
+    { NVMF_TRTYPE_FC,    "fc" },
+    { NVMF_TRTYPE_TCP,    "tcp" },
+    { NVMF_TRTYPE_LOOP,    "loop" },
+};
+

include/linux/nvme-types.h line 33

Quoted 10 lines
include/linux/nvme-types.h new file +10 -0
+static inline const char *nvme_trtype_to_name(u8 type)
+{
+    int i;
+
+    for (i = 0; i < ARRAY_SIZE(nvme_transport); i++) {
+        if (type == nvme_transport[i].type)
+            return nvme_transport[i].name;
+    }
+
+    return "invalid";

Why does this need a loop?

Quoted 4 lines

maybe just:
    return (type < ARRAY_SIZE(nvme_transport) && nvme_transport[type]) ? nvme_transport[i].name : "invalid";

Same for the rest...

It won't work since i != type.

I didn't change logic here, and mainly moved the target logic for type_name mappings to common code.

_______________________________________________
linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme

maintainer
#5 lore raw

Re: [PATCH 1/5] nvme: introduce nvme-types header file

Quoted 9 lines
Quoted 7 lines

Why does this need a loop?

maybe just:
    return (type < ARRAY_SIZE(nvme_transport) && nvme_transport[type]) ? nvme_transport[i].name : "invalid";

Same for the rest...

It won't work since i != type.

Yea, would be nice to fix it so we don't need the loop.

_______________________________________________
linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme

#6 lore raw

Re: [PATCH 1/5] nvme: introduce nvme-types header file

Quoted 11 lines · Sagi Grimberg
Quoted 9 lines
Quoted 7 lines

Why does this need a loop?

maybe just:
    return (type < ARRAY_SIZE(nvme_transport) && nvme_transport[type]) ? nvme_transport[i].name : "invalid";

Same for the rest...

It won't work since i != type.

Yea, would be nice to fix it so we don't need the loop.

I'm not sure that wasting memory on arrays that are bigger than needed is better.

NVMF_TRTYPE_LOOP        = 254

NVMF_ADDR_FAMILY_LOOP   = 254

I can create inline functions that have switch/case that will return type_to_name/name_to_type mappings instead of arrays (like we do in nvme-rdma.h).

is this better ?

Quoted 6 lines

_______________________________________________
linux-nvme mailing list
[email protected]
https://eur03.safelinks.protection.outlook.com/?url=http%3A%2F%2Flists.infradead.org%2Fmailman%2Flistinfo%2Flinux-nvme&amp;data=02%7C01%7Cmaxg%40mellanox.com%7C849cad0b92424000a08808d8092692ad%7Ca652971c7d2e4d9ba6a4d149256f461b%7C0%7C0%7C637269408560493242&amp;sdata=y4PYjHBCWEzm6QzVIP%2FyhatTDRAVYvxMfztosIN8pzw%3D&amp;reserved=0

_______________________________________________
linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme

#7 lore raw

Re: [PATCH 1/5] nvme: introduce nvme-types header file

In reply to Max Gurtovoy
Quoted 2 lines · Max Gurtovoy

Centralize the mapping between name and type to a common header file instead of duplicating logic in both NVMe host and target drivers.

The idea looks ok to me, but nvme_types suggested a header ala <sys/types.h> or <linux/types.h> to me. Not sure what a good name would be, though. I also suspect that keeping it under drivers/nvme/ in some fork might make more sense.

_______________________________________________
linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme

#8 lore raw

Re: [PATCH 1/5] nvme: introduce nvme-types header file

Quoted 7 lines · Christoph Hellwig
Quoted 2 lines · Max Gurtovoy

Centralize the mapping between name and type to a common header file instead of duplicating logic in both NVMe host and target drivers.

The idea looks ok to me, but nvme_types suggested a header ala <sys/types.h> or <linux/types.h> to me. Not sure what a good name would be, though. I also suspect that keeping it under drivers/nvme/ in some fork might make more sense.

how about <linux/nvme-common.h>  that will be used for common code between host/target subsystems ?

_______________________________________________
linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme

#9 lore raw

PATCH 2/5 nvme: use nvme_ana_type_to_name to get state string

In reply to Max Gurtovoy

Remove code duplication from nvme-core.ko and use common function to translate enum to ANA state string.

---
 drivers/nvme/host/multipath.c | 16 +++-------------
 drivers/nvme/host/nvme.h      |  1 +
 2 files changed, 4 insertions(+), 13 deletions(-)
drivers/nvme/host/multipath.c +3 -13
index da78e49..37cb9df 100644
@@ -124,15 +124,6 @@ void nvme_kick_requeue_lists(struct nvme_ctrl *ctrl)
 	up_read(&ctrl->namespaces_rwsem);
 }
 
-static const char *nvme_ana_state_names[] = {
-	[0]				= "invalid state",
-	[NVME_ANA_OPTIMIZED]		= "optimized",
-	[NVME_ANA_NONOPTIMIZED]		= "non-optimized",
-	[NVME_ANA_INACCESSIBLE]		= "inaccessible",
-	[NVME_ANA_PERSISTENT_LOSS]	= "persistent-loss",
-	[NVME_ANA_CHANGE]		= "change",
-};
-
 bool nvme_mpath_clear_current_path(struct nvme_ns *ns)
 {
 	struct nvme_ns_head *head = ns->head;
@@ -500,9 +491,8 @@ static int nvme_update_ana_state(struct nvme_ctrl *ctrl,
 	unsigned *nr_change_groups = data;
 	struct nvme_ns *ns;
 
-	dev_dbg(ctrl->device, "ANA group %d: %s.\n",
-			le32_to_cpu(desc->grpid),
-			nvme_ana_state_names[desc->state]);
+	dev_dbg(ctrl->device, "ANA group %d: %s.\n", le32_to_cpu(desc->grpid),
+		nvme_ana_type_to_name(desc->state));
 
 	if (desc->state == NVME_ANA_CHANGE)
 		(*nr_change_groups)++;
@@ -636,7 +626,7 @@ static ssize_t ana_state_show(struct device *dev, struct device_attribute *attr,
 {
 	struct nvme_ns *ns = nvme_get_ns_from_dev(dev);
 
-	return sprintf(buf, "%s\n", nvme_ana_state_names[ns->ana_state]);
+	return sprintf(buf, "%s\n", nvme_ana_type_to_name(ns->ana_state));
 }
 DEVICE_ATTR_RO(ana_state);
 
drivers/nvme/host/nvme.h +1 -0
index fa5c755..eef0e88 100644
@@ -7,6 +7,7 @@
 #define _NVME_H
 
 #include <linux/nvme.h>
+#include <linux/nvme-types.h>
 #include <linux/cdev.h>
 #include <linux/pci.h>
 #include <linux/kref.h>
Signature · 7 lines
1.8.3.1


_______________________________________________
linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme
Reviewed-by on v1 2/5
#10 lore raw

Re: [PATCH 2/5] nvme: use nvme_ana_type_to_name to get state string

Quoted 57 lines · Max Gurtovoy

Remove code duplication from nvme-core.ko and use common function to translate enum to ANA state string.

---
  drivers/nvme/host/multipath.c | 16 +++-------------
  drivers/nvme/host/nvme.h      |  1 +
  2 files changed, 4 insertions(+), 13 deletions(-)
drivers/nvme/host/multipath.c +3 -13
index da78e49..37cb9df 100644
@@ -124,15 +124,6 @@ void nvme_kick_requeue_lists(struct nvme_ctrl *ctrl)
  	up_read(&ctrl->namespaces_rwsem);
  }
 
-static const char *nvme_ana_state_names[] = {
-	[0]				= "invalid state",
-	[NVME_ANA_OPTIMIZED]		= "optimized",
-	[NVME_ANA_NONOPTIMIZED]		= "non-optimized",
-	[NVME_ANA_INACCESSIBLE]		= "inaccessible",
-	[NVME_ANA_PERSISTENT_LOSS]	= "persistent-loss",
-	[NVME_ANA_CHANGE]		= "change",
-};
-
  bool nvme_mpath_clear_current_path(struct nvme_ns *ns)
  {
  	struct nvme_ns_head *head = ns->head;
@@ -500,9 +491,8 @@ static int nvme_update_ana_state(struct nvme_ctrl *ctrl,
  	unsigned *nr_change_groups = data;
  	struct nvme_ns *ns;
 
-	dev_dbg(ctrl->device, "ANA group %d: %s.\n",
-			le32_to_cpu(desc->grpid),
-			nvme_ana_state_names[desc->state]);
+	dev_dbg(ctrl->device, "ANA group %d: %s.\n", le32_to_cpu(desc->grpid),
+		nvme_ana_type_to_name(desc->state));
 
  	if (desc->state == NVME_ANA_CHANGE)
  		(*nr_change_groups)++;
@@ -636,7 +626,7 @@ static ssize_t ana_state_show(struct device *dev, struct device_attribute *attr,
  {
  	struct nvme_ns *ns = nvme_get_ns_from_dev(dev);
 
-	return sprintf(buf, "%s\n", nvme_ana_state_names[ns->ana_state]);
+	return sprintf(buf, "%s\n", nvme_ana_type_to_name(ns->ana_state));
  }
  DEVICE_ATTR_RO(ana_state);
 
drivers/nvme/host/nvme.h +0 -0
index fa5c755..eef0e88 100644

drivers/nvme/host/nvme.h line 13

Quoted 7 lines
drivers/nvme/host/nvme.h +1 -0
@@ -7,6 +7,7 @@
  #define _NVME_H
 
  #include <linux/nvme.h>
+#include <linux/nvme-types.h>
  #include <linux/cdev.h>
  #include <linux/pci.h>
  #include <linux/kref.h>

Looks Good.

Signature · 6 lines
Himanshu Madhani                     Oracle Linux Engineering

_______________________________________________
linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme
#11 lore raw

PATCH 3/5 nvme: replace transport name with trtype enum for ops

In reply to Max Gurtovoy

Use common code to map transport type to transport name instead of duplicating logic.

---
 drivers/nvme/host/fabrics.c | 3 ++-
 drivers/nvme/host/fabrics.h | 4 ++--
 drivers/nvme/host/fc.c      | 2 +-
 drivers/nvme/host/rdma.c    | 2 +-
 drivers/nvme/host/tcp.c     | 2 +-
 drivers/nvme/target/loop.c  | 2 +-
 6 files changed, 8 insertions(+), 7 deletions(-)
drivers/nvme/host/fabrics.c +2 -1
index 2a6c819..bd5e44e 100644
@@ -528,7 +528,8 @@ static struct nvmf_transport_ops *nvmf_lookup_transport(
 	lockdep_assert_held(&nvmf_transports_rwsem);
 
 	list_for_each_entry(ops, &nvmf_transports, entry) {
-		if (strcmp(ops->name, opts->transport) == 0)
+		if (strcmp(nvme_trtype_to_name(ops->trtype),
+			   opts->transport) == 0)
 			return ops;
 	}
 
drivers/nvme/host/fabrics.h +2 -2
index a0ec40a..1fd52f0 100644
@@ -119,7 +119,7 @@ struct nvmf_ctrl_options {
  * @entry:		Used by the fabrics library to add the new
  *			registration entry to its linked-list internal tree.
  * @module:             Transport module reference
- * @name:		Name of the NVMe fabric driver implementation.
+ * @trtype:		NVMe fabric transport type
  * @required_opts:	sysfs command-line options that must be specified
  *			when adding a new NVMe controller.
  * @allowed_opts:	sysfs command-line options that can be specified
@@ -141,7 +141,7 @@ struct nvmf_ctrl_options {
 struct nvmf_transport_ops {
 	struct list_head	entry;
 	struct module		*module;
-	const char		*name;
+	u8			trtype;
 	int			required_opts;
 	int			allowed_opts;
 	struct nvme_ctrl	*(*create_ctrl)(struct device *dev,
drivers/nvme/host/fc.c +1 -1
index cb00075..67db0ae 100644
@@ -3694,7 +3694,7 @@ struct nvmet_fc_traddr {
 
 
 static struct nvmf_transport_ops nvme_fc_transport = {
-	.name		= "fc",
+	.trtype		= NVMF_TRTYPE_FC,
 	.module		= THIS_MODULE,
 	.required_opts	= NVMF_OPT_TRADDR | NVMF_OPT_HOST_TRADDR,
 	.allowed_opts	= NVMF_OPT_RECONNECT_DELAY | NVMF_OPT_CTRL_LOSS_TMO,
drivers/nvme/host/rdma.c +1 -1
index f8f856d..5e7dda3 100644
@@ -2329,7 +2329,7 @@ static struct nvme_ctrl *nvme_rdma_create_ctrl(struct device *dev,
 }
 
 static struct nvmf_transport_ops nvme_rdma_transport = {
-	.name		= "rdma",
+	.trtype		= NVMF_TRTYPE_RDMA,
 	.module		= THIS_MODULE,
 	.required_opts	= NVMF_OPT_TRADDR,
 	.allowed_opts	= NVMF_OPT_TRSVCID | NVMF_OPT_RECONNECT_DELAY |
drivers/nvme/host/tcp.c +1 -1
index 7c7c188..c678f97 100644
@@ -2487,7 +2487,7 @@ static struct nvme_ctrl *nvme_tcp_create_ctrl(struct device *dev,
 }
 
 static struct nvmf_transport_ops nvme_tcp_transport = {
-	.name		= "tcp",
+	.trtype		= NVMF_TRTYPE_TCP,
 	.module		= THIS_MODULE,
 	.required_opts	= NVMF_OPT_TRADDR,
 	.allowed_opts	= NVMF_OPT_TRSVCID | NVMF_OPT_RECONNECT_DELAY |
drivers/nvme/target/loop.c +1 -1
index 0d54e73..257ee60 100644
@@ -674,7 +674,7 @@ static void nvme_loop_remove_port(struct nvmet_port *port)
 };
 
 static struct nvmf_transport_ops nvme_loop_transport = {
-	.name		= "loop",
+	.trtype		= NVMF_TRTYPE_LOOP,
 	.module		= THIS_MODULE,
 	.create_ctrl	= nvme_loop_create_ctrl,
 	.allowed_opts	= NVMF_OPT_TRADDR,
Signature · 7 lines
1.8.3.1


_______________________________________________
linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme
Reviewed-by on v1 3/5
#12 lore raw

Re: [PATCH 3/5] nvme: replace transport name with trtype enum for ops

Quoted 95 lines · Max Gurtovoy

Use common code to map transport type to transport name instead of duplicating logic.

---
  drivers/nvme/host/fabrics.c | 3 ++-
  drivers/nvme/host/fabrics.h | 4 ++--
  drivers/nvme/host/fc.c      | 2 +-
  drivers/nvme/host/rdma.c    | 2 +-
  drivers/nvme/host/tcp.c     | 2 +-
  drivers/nvme/target/loop.c  | 2 +-
  6 files changed, 8 insertions(+), 7 deletions(-)
drivers/nvme/host/fabrics.c +2 -1
index 2a6c819..bd5e44e 100644
@@ -528,7 +528,8 @@ static struct nvmf_transport_ops *nvmf_lookup_transport(
  	lockdep_assert_held(&nvmf_transports_rwsem);
 
  	list_for_each_entry(ops, &nvmf_transports, entry) {
-		if (strcmp(ops->name, opts->transport) == 0)
+		if (strcmp(nvme_trtype_to_name(ops->trtype),
+			   opts->transport) == 0)
  			return ops;
  	}
 
drivers/nvme/host/fabrics.h +2 -2
index a0ec40a..1fd52f0 100644
@@ -119,7 +119,7 @@ struct nvmf_ctrl_options {
   * @entry:		Used by the fabrics library to add the new
   *			registration entry to its linked-list internal tree.
   * @module:             Transport module reference
- * @name:		Name of the NVMe fabric driver implementation.
+ * @trtype:		NVMe fabric transport type
   * @required_opts:	sysfs command-line options that must be specified
   *			when adding a new NVMe controller.
   * @allowed_opts:	sysfs command-line options that can be specified
@@ -141,7 +141,7 @@ struct nvmf_ctrl_options {
  struct nvmf_transport_ops {
  	struct list_head	entry;
  	struct module		*module;
-	const char		*name;
+	u8			trtype;
  	int			required_opts;
  	int			allowed_opts;
  	struct nvme_ctrl	*(*create_ctrl)(struct device *dev,
drivers/nvme/host/fc.c +1 -1
index cb00075..67db0ae 100644
@@ -3694,7 +3694,7 @@ struct nvmet_fc_traddr {
 
 
  static struct nvmf_transport_ops nvme_fc_transport = {
-	.name		= "fc",
+	.trtype		= NVMF_TRTYPE_FC,
  	.module		= THIS_MODULE,
  	.required_opts	= NVMF_OPT_TRADDR | NVMF_OPT_HOST_TRADDR,
  	.allowed_opts	= NVMF_OPT_RECONNECT_DELAY | NVMF_OPT_CTRL_LOSS_TMO,
drivers/nvme/host/rdma.c +1 -1
index f8f856d..5e7dda3 100644
@@ -2329,7 +2329,7 @@ static struct nvme_ctrl *nvme_rdma_create_ctrl(struct device *dev,
  }
 
  static struct nvmf_transport_ops nvme_rdma_transport = {
-	.name		= "rdma",
+	.trtype		= NVMF_TRTYPE_RDMA,
  	.module		= THIS_MODULE,
  	.required_opts	= NVMF_OPT_TRADDR,
  	.allowed_opts	= NVMF_OPT_TRSVCID | NVMF_OPT_RECONNECT_DELAY |
drivers/nvme/host/tcp.c +1 -1
index 7c7c188..c678f97 100644
@@ -2487,7 +2487,7 @@ static struct nvme_ctrl *nvme_tcp_create_ctrl(struct device *dev,
  }
 
  static struct nvmf_transport_ops nvme_tcp_transport = {
-	.name		= "tcp",
+	.trtype		= NVMF_TRTYPE_TCP,
  	.module		= THIS_MODULE,
  	.required_opts	= NVMF_OPT_TRADDR,
  	.allowed_opts	= NVMF_OPT_TRSVCID | NVMF_OPT_RECONNECT_DELAY |
drivers/nvme/target/loop.c +0 -0
index 0d54e73..257ee60 100644

drivers/nvme/target/loop.c static void nvme_loop_remove_port(struct nvmet_port *port) · line 680

Quoted 8 lines
drivers/nvme/target/loop.c +1 -1
@@ -674,7 +674,7 @@ static void nvme_loop_remove_port(struct nvmet_port *port)
  };
 
  static struct nvmf_transport_ops nvme_loop_transport = {
-	.name		= "loop",
+	.trtype		= NVMF_TRTYPE_LOOP,
  	.module		= THIS_MODULE,
  	.create_ctrl	= nvme_loop_create_ctrl,
  	.allowed_opts	= NVMF_OPT_TRADDR,
Signature · 6 lines
Himanshu Madhani                     Oracle Linux Engineering

_______________________________________________
linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme
maintainer
#13 lore raw

Re: [PATCH 3/5] nvme: replace transport name with trtype enum for ops

In reply to Max Gurtovoy
Quoted 3 lines

Use common code to map transport type to transport name instead of duplicating logic.

I'd prefer to keep the driver ops name and not rely on the trtype, userspace passes a name so I'd rather keep the transport registered with a name.

_______________________________________________
linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme

#14 lore raw

Re: [PATCH 3/5] nvme: replace transport name with trtype enum for ops

Quoted 7 lines · Sagi Grimberg
Quoted 3 lines

Use common code to map transport type to transport name instead of duplicating logic.

I'd prefer to keep the driver ops name and not rely on the trtype, userspace passes a name so I'd rather keep the transport registered with a name.

Userspace can continue passing names, but why not using common code ? as we do in the target side..

_______________________________________________
linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme

maintainer
#15 lore raw

Re: [PATCH 3/5] nvme: replace transport name with trtype enum for ops

Quoted 9 lines
Quoted 7 lines
Quoted 3 lines

Use common code to map transport type to transport name instead of duplicating logic.

I'd prefer to keep the driver ops name and not rely on the trtype, userspace passes a name so I'd rather keep the transport registered with a name.

Userspace can continue passing names, but why not using common code ? as we do in the target side..

We can still use common code, but I'd like to keep the driver name in the ops instead of trtype enumeration.

_______________________________________________
linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme

#16 lore raw

PATCH 4/5 nvmet-tcp: remove has_keyed_sgls initialization

In reply to Max Gurtovoy

Since the nvmet_tcp_ops is static, there is no need to initialize values to zero.

---
 drivers/nvme/target/tcp.c | 1 -
 1 file changed, 1 deletion(-)
drivers/nvme/target/tcp.c +0 -1
index 9e4cb90..3099237 100644
@@ -1751,7 +1751,6 @@ static void nvmet_tcp_disc_port_addr(struct nvmet_req *req,
 	.owner			= THIS_MODULE,
 	.type			= NVMF_TRTYPE_TCP,
 	.msdbd			= 1,
-	.has_keyed_sgls		= 0,
 	.add_port		= nvmet_tcp_add_port,
 	.remove_port		= nvmet_tcp_remove_port,
 	.queue_response		= nvmet_tcp_queue_response,
Signature · 7 lines
1.8.3.1


_______________________________________________
linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme
Reviewed-by on v1 4/5
#17 lore raw

Re: [PATCH 4/5] nvmet-tcp: remove has_keyed_sgls initialization

Quoted 15 lines · Max Gurtovoy

Since the nvmet_tcp_ops is static, there is no need to initialize values to zero.

---
  drivers/nvme/target/tcp.c | 1 -
  1 file changed, 1 deletion(-)
drivers/nvme/target/tcp.c +0 -0
index 9e4cb90..3099237 100644

drivers/nvme/target/tcp.c static void nvmet_tcp_disc_port_addr(struct nvmet_req *req, · line 1756

Quoted 7 lines
drivers/nvme/target/tcp.c +0 -1
@@ -1751,7 +1751,6 @@ static void nvmet_tcp_disc_port_addr(struct nvmet_req *req,
  	.owner			= THIS_MODULE,
  	.type			= NVMF_TRTYPE_TCP,
  	.msdbd			= 1,
-	.has_keyed_sgls		= 0,
  	.add_port		= nvmet_tcp_add_port,
  	.remove_port		= nvmet_tcp_remove_port,
  	.queue_response		= nvmet_tcp_queue_response,
Signature · 6 lines
Himanshu Madhani                     Oracle Linux Engineering

_______________________________________________
linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme
#18 lore raw

PATCH 5/5 nvmet: introduce flags member in nvmet_fabrics_ops

In reply to Max Gurtovoy

Replace has_keyed_sgls and metadata_support booleans with a flags member that will be used for adding more features in the future.

---
 drivers/nvme/target/admin-cmd.c | 2 +-
 drivers/nvme/target/core.c      | 2 +-
 drivers/nvme/target/discovery.c | 2 +-
 drivers/nvme/target/nvmet.h     | 5 +++--
 drivers/nvme/target/rdma.c      | 3 +--
 5 files changed, 7 insertions(+), 7 deletions(-)
drivers/nvme/target/admin-cmd.c +1 -1
index 1db8c04..95bb3bc 100644
@@ -427,7 +427,7 @@ static void nvmet_execute_identify_ctrl(struct nvmet_req *req)
 	id->awupf = 0;
 
 	id->sgls = cpu_to_le32(1 << 0);	/* we always support SGLs */
-	if (ctrl->ops->has_keyed_sgls)
+	if (ctrl->ops->flags & NVMF_KEYED_SGLS)
 		id->sgls |= cpu_to_le32(1 << 2);
 	if (req->port->inline_data_size)
 		id->sgls |= cpu_to_le32(1 << 20);
drivers/nvme/target/core.c +1 -1
index 6392bcd..86882d4 100644
@@ -322,7 +322,7 @@ int nvmet_enable_port(struct nvmet_port *port)
 	 * If the user requested PI support and the transport isn't pi capable,
 	 * don't enable the port.
 	 */
-	if (port->pi_enable && !ops->metadata_support) {
+	if (port->pi_enable && !(ops->flags & NVMF_METADATA_SUPPORTED)) {
 		pr_err("T10-PI is not supported by transport type %d\n",
 		       port->disc_addr.trtype);
 		ret = -EINVAL;
drivers/nvme/target/discovery.c +1 -1
index 40cf0b6..f40c05c 100644
@@ -277,7 +277,7 @@ static void nvmet_execute_disc_identify(struct nvmet_req *req)
 	id->maxcmd = cpu_to_le16(NVMET_MAX_CMD);
 
 	id->sgls = cpu_to_le32(1 << 0);	/* we always support SGLs */
-	if (ctrl->ops->has_keyed_sgls)
+	if (ctrl->ops->flags & NVMF_KEYED_SGLS)
 		id->sgls |= cpu_to_le32(1 << 2);
 	if (req->port->inline_data_size)
 		id->sgls |= cpu_to_le32(1 << 20);
drivers/nvme/target/nvmet.h +3 -2
index ce57d5f..daf3838 100644
@@ -287,8 +287,9 @@ struct nvmet_fabrics_ops {
 	struct module *owner;
 	unsigned int type;
 	unsigned int msdbd;
-	bool has_keyed_sgls : 1;
-	bool metadata_support : 1;
+	unsigned int flags;
+#define NVMF_KEYED_SGLS			(1 << 0)
+#define NVMF_METADATA_SUPPORTED		(1 << 1)
 	void (*queue_response)(struct nvmet_req *req);
 	int (*add_port)(struct nvmet_port *port);
 	void (*remove_port)(struct nvmet_port *port);
drivers/nvme/target/rdma.c +1 -2
index d514178..4c89f58 100644
@@ -1968,8 +1968,7 @@ static u8 nvmet_rdma_get_mdts(const struct nvmet_ctrl *ctrl)
 	.owner			= THIS_MODULE,
 	.type			= NVMF_TRTYPE_RDMA,
 	.msdbd			= 1,
-	.has_keyed_sgls		= 1,
-	.metadata_support	= 1,
+	.flags			= NVMF_KEYED_SGLS | NVMF_METADATA_SUPPORTED,
 	.add_port		= nvmet_rdma_add_port,
 	.remove_port		= nvmet_rdma_remove_port,
 	.queue_response		= nvmet_rdma_queue_response,
Signature · 7 lines
1.8.3.1


_______________________________________________
linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme
Reviewed-by on v1 5/5
#19 lore raw

Re: [PATCH 5/5] nvmet: introduce flags member in nvmet_fabrics_ops

Quoted 74 lines · Max Gurtovoy

Replace has_keyed_sgls and metadata_support booleans with a flags member that will be used for adding more features in the future.

---
  drivers/nvme/target/admin-cmd.c | 2 +-
  drivers/nvme/target/core.c      | 2 +-
  drivers/nvme/target/discovery.c | 2 +-
  drivers/nvme/target/nvmet.h     | 5 +++--
  drivers/nvme/target/rdma.c      | 3 +--
  5 files changed, 7 insertions(+), 7 deletions(-)
drivers/nvme/target/admin-cmd.c +1 -1
index 1db8c04..95bb3bc 100644
@@ -427,7 +427,7 @@ static void nvmet_execute_identify_ctrl(struct nvmet_req *req)
  	id->awupf = 0;
 
  	id->sgls = cpu_to_le32(1 << 0);	/* we always support SGLs */
-	if (ctrl->ops->has_keyed_sgls)
+	if (ctrl->ops->flags & NVMF_KEYED_SGLS)
  		id->sgls |= cpu_to_le32(1 << 2);
  	if (req->port->inline_data_size)
  		id->sgls |= cpu_to_le32(1 << 20);
drivers/nvme/target/core.c +1 -1
index 6392bcd..86882d4 100644
@@ -322,7 +322,7 @@ int nvmet_enable_port(struct nvmet_port *port)
  	 * If the user requested PI support and the transport isn't pi capable,
  	 * don't enable the port.
  	 */
-	if (port->pi_enable && !ops->metadata_support) {
+	if (port->pi_enable && !(ops->flags & NVMF_METADATA_SUPPORTED)) {
  		pr_err("T10-PI is not supported by transport type %d\n",
  		       port->disc_addr.trtype);
  		ret = -EINVAL;
drivers/nvme/target/discovery.c +1 -1
index 40cf0b6..f40c05c 100644
@@ -277,7 +277,7 @@ static void nvmet_execute_disc_identify(struct nvmet_req *req)
  	id->maxcmd = cpu_to_le16(NVMET_MAX_CMD);
 
  	id->sgls = cpu_to_le32(1 << 0);	/* we always support SGLs */
-	if (ctrl->ops->has_keyed_sgls)
+	if (ctrl->ops->flags & NVMF_KEYED_SGLS)
  		id->sgls |= cpu_to_le32(1 << 2);
  	if (req->port->inline_data_size)
  		id->sgls |= cpu_to_le32(1 << 20);
drivers/nvme/target/nvmet.h +3 -2
index ce57d5f..daf3838 100644
@@ -287,8 +287,9 @@ struct nvmet_fabrics_ops {
  	struct module *owner;
  	unsigned int type;
  	unsigned int msdbd;
-	bool has_keyed_sgls : 1;
-	bool metadata_support : 1;
+	unsigned int flags;
+#define NVMF_KEYED_SGLS			(1 << 0)
+#define NVMF_METADATA_SUPPORTED		(1 << 1)
  	void (*queue_response)(struct nvmet_req *req);
  	int (*add_port)(struct nvmet_port *port);
  	void (*remove_port)(struct nvmet_port *port);
drivers/nvme/target/rdma.c +0 -0
index d514178..4c89f58 100644

drivers/nvme/target/rdma.c static u8 nvmet_rdma_get_mdts(const struct nvmet_ctrl *ctrl) · line 1974

Quoted 9 lines
drivers/nvme/target/rdma.c +1 -2
@@ -1968,8 +1968,7 @@ static u8 nvmet_rdma_get_mdts(const struct nvmet_ctrl *ctrl)
  	.owner			= THIS_MODULE,
  	.type			= NVMF_TRTYPE_RDMA,
  	.msdbd			= 1,
-	.has_keyed_sgls		= 1,
-	.metadata_support	= 1,
+	.flags			= NVMF_KEYED_SGLS | NVMF_METADATA_SUPPORTED,
  	.add_port		= nvmet_rdma_add_port,
  	.remove_port		= nvmet_rdma_remove_port,
  	.queue_response		= nvmet_rdma_queue_response,

Looks Good.

Signature · 6 lines
Himanshu Madhani                     Oracle Linux Engineering

_______________________________________________
linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme
#20 lore raw
In reply to Max Gurtovoy
Quoted 8 lines · Max Gurtovoy

This series introduce some cleanups and code/logic duplication for host and target side. It introduces a new header file nvme-types.h that will be used by both host and target drivers. For now it will include mainly mapping structures from types to names, but in the future it can include more common logic between host/target drivers (as done for specific TCP and RDMA transports). The series continues with adding a flags member to nvmet_fabrics_ops structure and removes the need for adding dedicated member for each feature in this struct.

From a quick look this looks ok, but w're done with the 5.8 merge window except for bug fixes.

_______________________________________________
linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme

#21 lore raw
Quoted 11 lines · Christoph Hellwig
Quoted 8 lines · Max Gurtovoy

This series introduce some cleanups and code/logic duplication for host and target side. It introduces a new header file nvme-types.h that will be used by both host and target drivers. For now it will include mainly mapping structures from types to names, but in the future it can include more common logic between host/target drivers (as done for specific TCP and RDMA transports). The series continues with adding a flags member to nvmet_fabrics_ops structure and removes the need for adding dedicated member for each feature in this struct.

From a quick look this looks ok, but w're done with the 5.8 merge
window except for bug fixes.

Ok so it can go to nvme-5.9..

Quoted 4 lines

_______________________________________________
linux-nvme mailing list
[email protected]
https://eur03.safelinks.protection.outlook.com/?url=http%3A%2F%2Flists.infradead.org%2Fmailman%2Flistinfo%2Flinux-nvme&amp;data=02%7C01%7Cmaxg%40mellanox.com%7Ca5d736a6fdf14b4e6e8108d807bf3ab4%7Ca652971c7d2e4d9ba6a4d149256f461b%7C0%7C0%7C637267865201210781&amp;sdata=BHo4gR2xiT6WIgiE8BEReIiwVhfUfKcJl7zw651W0l4%3D&amp;reserved=0

_______________________________________________
linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme