[net-next-2.6 V6 PATCH 0/2] Add virtual port netlink support

[net-next-2.6 V6 PATCH 0/2] Add virtual port netlink support

[Scott Feldman]

The following series adds virtual port netlink support and adds an
implementation to Cisco's enic netdev driver:

	1/2: Adds virtual netlink RTM_SETLINK/RTM_GETLINK support, and
	     adds matching netdev ops net_{set|get}_vf_port.

	2/2: Adds enic support for net_{set|get}_vf_port for enic
	     dynamic devices.

Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Scott Feldman <scofeldm@cisco.com>
Signed-off-by: Roopa Prabhu<roprabhu@cisco.com>

[net-next-2.6 V6 PATCH 1/2] Add netlink support for virtual port management (was iovnl)

[Scott Feldman]

From: Scott Feldman <scofeldm@cisco.com>

Add new netdev ops ndo_{set|get}_vf_port to allow setting of
port-profile on a netdev interface.  Extends netlink socket RTM_SETLINK/
RTM_GETLINK with new sub cmd called IFLA_VF_PORT (added to end of
IFLA_cmd list).

A port-profile is used to configure/enable the external switch virtual port
backing the netdev interface, not to configure the host-facing side of the
netdev.  A port-profile is an identifier known to the switch.  How port-
profiles are installed on the switch or how available port-profiles are
made know to the host is outside the scope of this patch.

There are two types of port-profiles specs in the netlink msg.  The first spec
is for 802.1Qbg (pre-)standard, VDP protocol.  The second spec is for devices
that run a similar protocol as VDP but in firmware, thus hiding the protocol
details.  In either case, the specs have much in common and makes sense to
define the netlink msg as the union of the two specs.  For example, both specs
have a notition of associating/deassociating a port-profile.  And both specs
require some information from the hypervisor manager, such as client port
instance ID.

The general flow is the port-profile is applied to a host netdev interface
using RTM_SETLINK, the receiver of the RTM_SETLINK msg communicates with the
switch, and the switch virtual port backing the host netdev interface is
configured/enabled based on the settings defined by the port-profile.  What
those settings comprise, and how those settings are managed is again
outside the scope of this patch, since this patch only deals with the
first step in the flow.

There is a RTM_GETLINK cmd to to return port-profile setting of an
interface and to also return the status of the last port-profile
association.

IFLA_VF_PORT is modeled after the existing IFLA_VF_* cmd where a
VF number is passed in to identify the virtual function (VF) of an SR-IOV-
capable device.  In this case, the target of IFLA_VF_PORT msg is the
netdev physical function (PF) device.  The PF will apply the port-profile
to the VF.  IFLA_VF_PORT can also be used for devices that don't
adhere to SR-IOV and can apply the port-profile directly to the netdev
target.  In this case, the VF number is ignored.

Passing in a NULL port-profile is used to delete the port-profile association.

Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Scott Feldman <scofeldm@cisco.com>
Signed-off-by: Roopa Prabhu<roprabhu@cisco.com>
---
 include/linux/if_link.h   |   52 +++++++++++++++++++
 include/linux/netdevice.h |   10 ++++
 net/core/rtnetlink.c      |  122 +++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 183 insertions(+), 1 deletions(-)

diff --git a/include/linux/if_link.h b/include/linux/if_link.h
index cfd420b..d93a4a5 100644
--- a/include/linux/if_link.h
+++ b/include/linux/if_link.h
@@ -116,6 +116,7 @@ enum {
 	IFLA_VF_TX_RATE,	/* TX Bandwidth Allocation */
 	IFLA_VFINFO,
 	IFLA_STATS64,
+	IFLA_VF_PORT,
 	__IFLA_MAX
 };
 
@@ -259,4 +260,55 @@ struct ifla_vf_info {
 	__u32 qos;
 	__u32 tx_rate;
 };
+
+/* VF Port management section */
+
+enum {
+	IFLA_VF_PORT_UNSPEC,
+	IFLA_VF_PORT_VF,		/* __u32 */
+	IFLA_VF_PORT_PROFILE,		/* string */
+	IFLA_VF_PORT_VSI_TYPE,		/* 802.1Qbg (pre-)standard VDP */
+	IFLA_VF_PORT_INSTANCE_UUID,	/* binary UUID */
+	IFLA_VF_PORT_HOST_UUID,		/* binary UUID */
+	IFLA_VF_PORT_REQUEST,		/* __u8 */
+	IFLA_VF_PORT_RESPONSE,		/* __u16, output only */
+	__IFLA_VF_PORT_MAX,
+};
+
+#define IFLA_VF_PORT_MAX (__IFLA_VF_PORT_MAX - 1)
+
+#define VF_PORT_PROFILE_MAX	40
+#define VF_PORT_UUID_MAX	16
+
+enum {
+	VF_PORT_REQUEST_PREASSOCIATE = 0,
+	VF_PORT_REQUEST_PREASSOCIATE_RR,
+	VF_PORT_REQUEST_ASSOCIATE,
+	VF_PORT_REQUEST_DISASSOCIATE,
+};
+
+enum {
+	VF_PORT_VDP_RESPONSE_SUCCESS = 0,
+	VF_PORT_VDP_RESPONSE_INVALID_FORMAT,
+	VF_PORT_VDP_RESPONSE_INSUFFICIENT_RESOURCES,
+	VF_PORT_VDP_RESPONSE_UNUSED_VTID,
+	VF_PORT_VDP_RESPONSE_VTID_VIOLATION,
+	VF_PORT_VDP_RESPONSE_VTID_VERSION_VIOALTION,
+	VF_PORT_VDP_RESPONSE_OUT_OF_SYNC,
+	/* 0x08-0xFF reserved for future VDP use */
+	VF_PORT_PROFILE_RESPONSE_SUCCESS = 0x100,
+	VF_PORT_PROFILE_RESPONSE_INPROGRESS,
+	VF_PORT_PROFILE_RESPONSE_INVALID,
+	VF_PORT_PROFILE_RESPONSE_BADSTATE,
+	VF_PORT_PROFILE_RESPONSE_INSUFFICIENT_RESOURCES,
+	VF_PORT_PROFILE_RESPONSE_ERROR,
+};
+
+struct ifla_vf_port_vsi {
+	__u8 vsi_mgr_id;
+	__u8 vsi_type_id[3];
+	__u8 vsi_type_version;
+	__u8 pad[3];
+};
+
 #endif /* _LINUX_IF_LINK_H */
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 69022d4..c2ba8d4 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -686,6 +686,10 @@ struct netdev_rx_queue {
  * int (*ndo_set_vf_tx_rate)(struct net_device *dev, int vf, int rate);
  * int (*ndo_get_vf_config)(struct net_device *dev,
  *			    int vf, struct ifla_vf_info *ivf);
+ * int (*ndo_set_vf_port)(struct net_device *dev, int vf,
+ *			  struct nlattr *vf_port[]);
+ * int (*ndo_get_vf_port)(struct net_device *dev, int vf,
+ *			  struct sk_buff *skb);
  */
 #define HAVE_NET_DEVICE_OPS
 struct net_device_ops {
@@ -735,6 +739,12 @@ struct net_device_ops {
 	int			(*ndo_get_vf_config)(struct net_device *dev,
 						     int vf,
 						     struct ifla_vf_info *ivf);
+	int			(*ndo_set_vf_port)(struct net_device *dev,
+						   int vf,
+						   struct nlattr *vf_port[]);
+	int			(*ndo_get_vf_port)(struct net_device *dev,
+						   int vf,
+						   struct sk_buff *skb);
 #if defined(CONFIG_FCOE) || defined(CONFIG_FCOE_MODULE)
 	int			(*ndo_fcoe_enable)(struct net_device *dev);
 	int			(*ndo_fcoe_disable)(struct net_device *dev);
diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
index 23a71cb..de14d36 100644
--- a/net/core/rtnetlink.c
+++ b/net/core/rtnetlink.c
@@ -653,6 +653,26 @@ static inline int rtnl_vfinfo_size(const struct net_device *dev)
 		return 0;
 }
 
+static size_t rtnl_vf_port_size(const struct net_device *dev)
+{
+	size_t vf_port_size = nla_total_size(sizeof(struct nlattr))
+						     /* VF_PORT_VF */
+		+ nla_total_size(VF_PORT_PROFILE_MAX)/* VF_PORT_PROFILE */
+		+ nla_total_size(sizeof(struct ifla_vf_port_vsi))
+						     /* VF_PORT_VSI_TYPE */
+		+ nla_total_size(VF_PORT_UUID_MAX)   /* VF_PORT_VSI_INSTANCE */
+		+ nla_total_size(VF_PORT_UUID_MAX)   /* VF_PORT_HOST_UUID */
+		+ nla_total_size(1)		     /* VF_PROT_VDP_REQUEST */
+		+ nla_total_size(2);		     /* VF_PORT_VDP_RESPONSE */
+
+	if (!dev->netdev_ops->ndo_get_vf_port || !dev->dev.parent)
+		return 0;
+	if (dev_num_vf(dev->dev.parent))
+		return vf_port_size * dev_num_vf(dev->dev.parent);
+	else
+		return vf_port_size;
+}
+
 static inline size_t if_nlmsg_size(const struct net_device *dev)
 {
 	return NLMSG_ALIGN(sizeof(struct ifinfomsg))
@@ -673,9 +693,67 @@ static inline size_t if_nlmsg_size(const struct net_device *dev)
 	       + nla_total_size(1) /* IFLA_LINKMODE */
 	       + nla_total_size(4) /* IFLA_NUM_VF */
 	       + nla_total_size(rtnl_vfinfo_size(dev)) /* IFLA_VFINFO */
+	       + rtnl_vf_port_size(dev) /* IFLA_VF_PORT */
 	       + rtnl_link_get_size(dev); /* IFLA_LINKINFO */
 }
 
+static int rtnl_vf_port_fill_nest(struct sk_buff *skb, struct net_device *dev,
+	int vf)
+{
+	struct nlattr *data;
+	int err;
+
+	data = nla_nest_start(skb, IFLA_VF_PORT);
+	if (!data)
+		return -EMSGSIZE;
+
+	if (vf >= 0)
+		nla_put_u32(skb, IFLA_VF_PORT_VF, vf);
+
+	err = dev->netdev_ops->ndo_get_vf_port(dev, vf, skb);
+	if (err == -EMSGSIZE) {
+		nla_nest_cancel(skb, data);
+		return -EMSGSIZE;
+	} else if (err) {
+		nla_nest_cancel(skb, data);
+		return 0;
+	}
+
+	nla_nest_end(skb, data);
+
+	return 0;
+}
+
+static int rtnl_vf_port_fill(struct sk_buff *skb, struct net_device *dev)
+{
+	int num_vf;
+	int err;
+
+	if (!dev->netdev_ops->ndo_get_vf_port || !dev->dev.parent)
+		return 0;
+
+	num_vf = dev_num_vf(dev->dev.parent);
+
+	if (num_vf) {
+		int i;
+
+		for (i = 0; i < num_vf; i++) {
+			err = rtnl_vf_port_fill_nest(skb, dev, i);
+			if (err)
+				goto nla_put_failure;
+		}
+	} else  {
+		err = rtnl_vf_port_fill_nest(skb, dev, -1);
+		if (err)
+			goto nla_put_failure;
+	}
+
+	return 0;
+
+nla_put_failure:
+	return -EMSGSIZE;
+}
+
 static int rtnl_fill_ifinfo(struct sk_buff *skb, struct net_device *dev,
 			    int type, u32 pid, u32 seq, u32 change,
 			    unsigned int flags)
@@ -747,17 +825,23 @@ static int rtnl_fill_ifinfo(struct sk_buff *skb, struct net_device *dev,
 		goto nla_put_failure;
 	copy_rtnl_link_stats64(nla_data(attr), stats);
 
+	if (dev->dev.parent)
+		NLA_PUT_U32(skb, IFLA_NUM_VF, dev_num_vf(dev->dev.parent));
+
 	if (dev->netdev_ops->ndo_get_vf_config && dev->dev.parent) {
 		int i;
 		struct ifla_vf_info ivi;
 
-		NLA_PUT_U32(skb, IFLA_NUM_VF, dev_num_vf(dev->dev.parent));
 		for (i = 0; i < dev_num_vf(dev->dev.parent); i++) {
 			if (dev->netdev_ops->ndo_get_vf_config(dev, i, &ivi))
 				break;
 			NLA_PUT(skb, IFLA_VFINFO, sizeof(ivi), &ivi);
 		}
 	}
+
+	if (rtnl_vf_port_fill(skb, dev))
+		goto nla_put_failure;
+
 	if (dev->rtnl_link_ops) {
 		if (rtnl_link_fill(skb, dev) < 0)
 			goto nla_put_failure;
@@ -824,6 +908,7 @@ const struct nla_policy ifla_policy[IFLA_MAX+1] = {
 				    .len = sizeof(struct ifla_vf_vlan) },
 	[IFLA_VF_TX_RATE]	= { .type = NLA_BINARY,
 				    .len = sizeof(struct ifla_vf_tx_rate) },
+	[IFLA_VF_PORT]		= { .type = NLA_NESTED },
 };
 EXPORT_SYMBOL(ifla_policy);
 
@@ -832,6 +917,20 @@ static const struct nla_policy ifla_info_policy[IFLA_INFO_MAX+1] = {
 	[IFLA_INFO_DATA]	= { .type = NLA_NESTED },
 };
 
+static const struct nla_policy ifla_vf_port_policy[IFLA_VF_PORT_MAX+1] = {
+	[IFLA_VF_PORT_VF]		= { .type = NLA_U32 },
+	[IFLA_VF_PORT_PROFILE]		= { .type = NLA_STRING,
+				.len = VF_PORT_PROFILE_MAX },
+	[IFLA_VF_PORT_VSI_TYPE]		= { .type = NLA_BINARY,
+				.len = sizeof(struct ifla_vf_port_vsi)},
+	[IFLA_VF_PORT_INSTANCE_UUID]	= { .type = NLA_BINARY,
+				.len = VF_PORT_UUID_MAX },
+	[IFLA_VF_PORT_HOST_UUID]	= { .type = NLA_STRING,
+				.len = VF_PORT_UUID_MAX },
+	[IFLA_VF_PORT_REQUEST]		= { .type = NLA_U8, },
+	[IFLA_VF_PORT_RESPONSE]		= { .type = NLA_U16, },
+};
+
 struct net *rtnl_link_get_net(struct net *src_net, struct nlattr *tb[])
 {
 	struct net *net;
@@ -1028,6 +1127,27 @@ static int do_setlink(struct net_device *dev, struct ifinfomsg *ifm,
 	}
 	err = 0;
 
+	if (tb[IFLA_VF_PORT]) {
+		struct nlattr *vf_port[IFLA_VF_PORT_MAX+1];
+		int vf = -1;
+
+		err = nla_parse_nested(vf_port, IFLA_VF_PORT_MAX,
+			tb[IFLA_VF_PORT], ifla_vf_port_policy);
+		if (err < 0)
+			goto errout;
+
+		if (vf_port[IFLA_VF_PORT_VF])
+			vf = nla_get_u32(vf_port[IFLA_VF_PORT_VF]);
+
+		err = -EOPNOTSUPP;
+		if (ops->ndo_set_vf_port)
+			err = ops->ndo_set_vf_port(dev, vf, vf_port);
+		if (err < 0)
+			goto errout;
+		modified = 1;
+	}
+	err = 0;
+
 errout:
 	if (err < 0 && modified && net_ratelimit())
 		printk(KERN_WARNING "A link change request failed with "

[net-next-2.6 V6 PATCH 2/2] Add ndo_{set|get}_vf_port support for enic dynamic vnics

[Scott Feldman]

From: Scott Feldman <scofeldm@cisco.com>

Add enic ndo_{set|get}_vf_port ops to support setting/getting
port-profile for enic dynamic devices.  Enic dynamic devices are just like
normal enic eth devices except dynamic enics require an extra configuration
step to assign a port-profile identifier to the interface before the
interface is useable.  Once a port-profile is assigned, link comes up on the
interface and is ready for I/O.  The port-profile is used to configure the
network port assigned to the interface.  The network port configuration
includes VLAN membership, QoS policies, and port security settings typical
of a data center network.

A dynamic enic initially has a zero-mac address.  Before a port-profile is
assigned, a valid non-zero unicast mac address should be assign to the
dynamic enic interface.

Signed-off-by: Scott Feldman <scofeldm@cisco.com>
Signed-off-by: Roopa Prabhu<roprabhu@cisco.com>
---
 drivers/net/enic/Makefile    |    2 
 drivers/net/enic/enic.h      |    9 +
 drivers/net/enic/enic_main.c |  321 +++++++++++++++++++++++++++++++++++++++---
 drivers/net/enic/enic_res.c  |    5 -
 drivers/net/enic/enic_res.h  |    1 
 drivers/net/enic/vnic_dev.c  |   58 +++++++-
 drivers/net/enic/vnic_dev.h  |    7 +
 drivers/net/enic/vnic_vic.c  |   73 ++++++++++
 drivers/net/enic/vnic_vic.h  |   59 ++++++++
 9 files changed, 500 insertions(+), 35 deletions(-)

diff --git a/drivers/net/enic/Makefile b/drivers/net/enic/Makefile
index 391c3bc..e7b6c31 100644
--- a/drivers/net/enic/Makefile
+++ b/drivers/net/enic/Makefile
@@ -1,5 +1,5 @@
 obj-$(CONFIG_ENIC) := enic.o
 
 enic-y := enic_main.o vnic_cq.o vnic_intr.o vnic_wq.o \
-	enic_res.o vnic_dev.o vnic_rq.o
+	enic_res.o vnic_dev.o vnic_rq.o vnic_vic.o
 
diff --git a/drivers/net/enic/enic.h b/drivers/net/enic/enic.h
index 5fa56f1..91088be 100644
--- a/drivers/net/enic/enic.h
+++ b/drivers/net/enic/enic.h
@@ -34,7 +34,7 @@
 
 #define DRV_NAME		"enic"
 #define DRV_DESCRIPTION		"Cisco VIC Ethernet NIC Driver"
-#define DRV_VERSION		"1.3.1.1"
+#define DRV_VERSION		"1.3.1.1-pp"
 #define DRV_COPYRIGHT		"Copyright 2008-2009 Cisco Systems, Inc"
 #define PFX			DRV_NAME ": "
 
@@ -74,6 +74,12 @@ struct enic_msix_entry {
 	void *devid;
 };
 
+struct enic_port_profile {
+	char name[VF_PORT_PROFILE_MAX];
+	u8 instance_uuid[VF_PORT_UUID_MAX];
+	u8 host_uuid[VF_PORT_UUID_MAX];
+};
+
 /* Per-instance private data structure */
 struct enic {
 	struct net_device *netdev;
@@ -95,6 +101,7 @@ struct enic {
 	u32 port_mtu;
 	u32 rx_coalesce_usecs;
 	u32 tx_coalesce_usecs;
+	struct enic_port_profile pp;
 
 	/* work queue cache line section */
 	____cacheline_aligned struct vnic_wq wq[ENIC_WQ_MAX];
diff --git a/drivers/net/enic/enic_main.c b/drivers/net/enic/enic_main.c
index 1232887..fb8bf4a 100644
--- a/drivers/net/enic/enic_main.c
+++ b/drivers/net/enic/enic_main.c
@@ -29,6 +29,7 @@
 #include <linux/etherdevice.h>
 #include <linux/if_ether.h>
 #include <linux/if_vlan.h>
+#include <linux/if_link.h>
 #include <linux/ethtool.h>
 #include <linux/in.h>
 #include <linux/ip.h>
@@ -40,6 +41,7 @@
 #include "vnic_dev.h"
 #include "vnic_intr.h"
 #include "vnic_stats.h"
+#include "vnic_vic.h"
 #include "enic_res.h"
 #include "enic.h"
 
@@ -49,10 +51,12 @@
 #define ENIC_DESC_MAX_SPLITS		(MAX_TSO / WQ_ENET_MAX_DESC_LEN + 1)
 
 #define PCI_DEVICE_ID_CISCO_VIC_ENET         0x0043  /* ethernet vnic */
+#define PCI_DEVICE_ID_CISCO_VIC_ENET_DYN     0x0044  /* enet dynamic vnic */
 
 /* Supported devices */
 static DEFINE_PCI_DEVICE_TABLE(enic_id_table) = {
 	{ PCI_VDEVICE(CISCO, PCI_DEVICE_ID_CISCO_VIC_ENET) },
+	{ PCI_VDEVICE(CISCO, PCI_DEVICE_ID_CISCO_VIC_ENET_DYN) },
 	{ 0, }	/* end of table */
 };
 
@@ -113,6 +117,11 @@ static const struct enic_stat enic_rx_stats[] = {
 static const unsigned int enic_n_tx_stats = ARRAY_SIZE(enic_tx_stats);
 static const unsigned int enic_n_rx_stats = ARRAY_SIZE(enic_rx_stats);
 
+static int enic_is_dynamic(struct enic *enic)
+{
+	return enic->pdev->device == PCI_DEVICE_ID_CISCO_VIC_ENET_DYN;
+}
+
 static int enic_get_settings(struct net_device *netdev,
 	struct ethtool_cmd *ecmd)
 {
@@ -810,14 +819,73 @@ static void enic_reset_mcaddrs(struct enic *enic)
 
 static int enic_set_mac_addr(struct net_device *netdev, char *addr)
 {
-	if (!is_valid_ether_addr(addr))
-		return -EADDRNOTAVAIL;
+	struct enic *enic = netdev_priv(netdev);
+
+	if (enic_is_dynamic(enic)) {
+		if (!is_valid_ether_addr(addr) && !is_zero_ether_addr(addr))
+			return -EADDRNOTAVAIL;
+	} else {
+		if (!is_valid_ether_addr(addr))
+			return -EADDRNOTAVAIL;
+	}
 
 	memcpy(netdev->dev_addr, addr, netdev->addr_len);
 
 	return 0;
 }
 
+static int enic_dev_add_station_addr(struct enic *enic)
+{
+	int err = 0;
+
+	if (is_valid_ether_addr(enic->netdev->dev_addr)) {
+		spin_lock(&enic->devcmd_lock);
+		err = vnic_dev_add_addr(enic->vdev, enic->netdev->dev_addr);
+		spin_unlock(&enic->devcmd_lock);
+	}
+
+	return err;
+}
+
+static int enic_dev_del_station_addr(struct enic *enic)
+{
+	int err = 0;
+
+	if (is_valid_ether_addr(enic->netdev->dev_addr)) {
+		spin_lock(&enic->devcmd_lock);
+		err = vnic_dev_del_addr(enic->vdev, enic->netdev->dev_addr);
+		spin_unlock(&enic->devcmd_lock);
+	}
+
+	return err;
+}
+
+static int enic_set_mac_address(struct net_device *netdev, void *p)
+{
+	struct enic *enic = netdev_priv(netdev);
+	struct sockaddr *saddr = p;
+	char *addr = saddr->sa_data;
+	int err = -EOPNOTSUPP;
+
+	if (enic_is_dynamic(enic)) {
+		if (netif_running(enic->netdev)) {
+			err = enic_dev_del_station_addr(enic);
+			if (err)
+				return err;
+		}
+		err = enic_set_mac_addr(netdev, addr);
+		if (err)
+			return err;
+		if (netif_running(enic->netdev)) {
+			err = enic_dev_add_station_addr(enic);
+			if (err)
+				return err;
+		}
+	}
+
+	return err;
+}
+
 /* netif_tx_lock held, BHs disabled */
 static void enic_set_multicast_list(struct net_device *netdev)
 {
@@ -922,6 +990,209 @@ static void enic_tx_timeout(struct net_device *netdev)
 	schedule_work(&enic->reset);
 }
 
+static int enic_vnic_dev_deinit(struct enic *enic)
+{
+	int err;
+
+	spin_lock(&enic->devcmd_lock);
+	err = vnic_dev_deinit(enic->vdev);
+	spin_unlock(&enic->devcmd_lock);
+
+	return err;
+}
+
+static int enic_dev_init_prov(struct enic *enic, struct vic_provinfo *vp)
+{
+	int err;
+
+	spin_lock(&enic->devcmd_lock);
+	err = vnic_dev_init_prov(enic->vdev,
+		(u8 *)vp, vic_provinfo_size(vp));
+	spin_unlock(&enic->devcmd_lock);
+
+	return err;
+}
+
+static int enic_dev_init_done(struct enic *enic, int *done, int *error)
+{
+	int err;
+
+	spin_lock(&enic->devcmd_lock);
+	err = vnic_dev_init_done(enic->vdev, done, error);
+	spin_unlock(&enic->devcmd_lock);
+
+	return err;
+}
+
+static int enic_set_port_profile(struct enic *enic, int vf, u8 *mac,
+	char *name, u8 *instance_uuid, u8 *host_uuid)
+{
+	struct vic_provinfo *vp;
+	u8 oui[3] = VIC_PROVINFO_CISCO_OUI;
+	unsigned short *uuid;
+	char uuid_str[38];
+	static char *uuid_fmt = "%04X%04X-%04X-%04X-%04X-%04X%04X%04X";
+	int err;
+
+	if (!name)
+		return -EINVAL;
+
+	if (!is_valid_ether_addr(mac))
+		return -EADDRNOTAVAIL;
+
+	vp = vic_provinfo_alloc(GFP_KERNEL, oui, VIC_PROVINFO_LINUX_TYPE);
+	if (!vp)
+		return -ENOMEM;
+
+	vic_provinfo_add_tlv(vp,
+		VIC_LINUX_PROV_TLV_PORT_PROFILE_NAME_STR,
+		strlen(name) + 1, name);
+
+	vic_provinfo_add_tlv(vp,
+		VIC_LINUX_PROV_TLV_CLIENT_MAC_ADDR,
+		ETH_ALEN, mac);
+
+	if (instance_uuid) {
+		uuid = (unsigned short *)instance_uuid;
+		sprintf(uuid_str, uuid_fmt,
+			uuid[0], uuid[1], uuid[2], uuid[3],
+			uuid[4], uuid[5], uuid[6], uuid[7]);
+		vic_provinfo_add_tlv(vp,
+			VIC_LINUX_PROV_TLV_CLIENT_UUID_STR,
+			sizeof(uuid_str), uuid_str);
+	}
+
+	if (host_uuid) {
+		uuid = (unsigned short *)host_uuid;
+		sprintf(uuid_str, uuid_fmt,
+			uuid[0], uuid[1], uuid[2], uuid[3],
+			uuid[4], uuid[5], uuid[6], uuid[7]);
+		vic_provinfo_add_tlv(vp,
+			VIC_LINUX_PROV_TLV_HOST_UUID_STR,
+			sizeof(uuid_str), uuid_str);
+	}
+
+	err = enic_vnic_dev_deinit(enic);
+	if (err)
+		goto err_out;
+
+	memset(&enic->pp, 0, sizeof(enic->pp));
+
+	err = enic_dev_init_prov(enic, vp);
+	if (err)
+		goto err_out;
+
+	memcpy(enic->pp.name, name, VF_PORT_PROFILE_MAX);
+	if (instance_uuid)
+		memcpy(enic->pp.instance_uuid,
+			instance_uuid, VF_PORT_UUID_MAX);
+	if (host_uuid)
+		memcpy(enic->pp.host_uuid,
+			host_uuid, VF_PORT_UUID_MAX);
+
+err_out:
+	vic_provinfo_free(vp);
+
+	return err;
+}
+
+static int enic_unset_port_profile(struct enic *enic, int vf)
+{
+	memset(&enic->pp, 0, sizeof(enic->pp));
+	return enic_vnic_dev_deinit(enic);
+}
+
+static int enic_set_vf_port(struct net_device *netdev, int vf,
+	struct nlattr *vf_port[])
+{
+	struct enic *enic = netdev_priv(netdev);
+	char *name = NULL;
+	u8 *instance_uuid = NULL;
+	u8 *host_uuid = NULL;
+	u8 request = VF_PORT_REQUEST_DISASSOCIATE;
+
+	if (!enic_is_dynamic(enic))
+		return -EOPNOTSUPP;
+
+	if (vf_port[IFLA_VF_PORT_REQUEST])
+		request = nla_get_u8(vf_port[IFLA_VF_PORT_REQUEST]);
+
+	switch (request) {
+	case VF_PORT_REQUEST_ASSOCIATE:
+
+		if (vf_port[IFLA_VF_PORT_PROFILE])
+			name = nla_data(vf_port[IFLA_VF_PORT_PROFILE]);
+
+		if (vf_port[IFLA_VF_PORT_INSTANCE_UUID])
+			instance_uuid =
+				nla_data(vf_port[IFLA_VF_PORT_INSTANCE_UUID]);
+
+		if (vf_port[IFLA_VF_PORT_HOST_UUID])
+			host_uuid = nla_data(vf_port[IFLA_VF_PORT_HOST_UUID]);
+
+		return enic_set_port_profile(enic, vf,
+			netdev->dev_addr, name,
+			instance_uuid, host_uuid);
+
+	case VF_PORT_REQUEST_DISASSOCIATE:
+
+		return enic_unset_port_profile(enic, vf);
+
+	default:
+		break;
+	}
+
+	return -EOPNOTSUPP;
+}
+
+static int enic_get_vf_port(struct net_device *netdev, int vf,
+	struct sk_buff *skb)
+{
+	struct enic *enic = netdev_priv(netdev);
+	int err, error, done;
+	u16 response = VF_PORT_PROFILE_RESPONSE_SUCCESS;
+
+	if (!enic_is_dynamic(enic))
+		return -EOPNOTSUPP;
+
+	err = enic_dev_init_done(enic, &done, &error);
+
+	if (err)
+		return err;
+
+	switch (error) {
+	case ERR_SUCCESS:
+		if (!done)
+			response = VF_PORT_PROFILE_RESPONSE_INPROGRESS;
+		break;
+	case ERR_EINVAL:
+		response = VF_PORT_PROFILE_RESPONSE_INVALID;
+		break;
+	case ERR_EBADSTATE:
+		response = VF_PORT_PROFILE_RESPONSE_BADSTATE;
+		break;
+	case ERR_ENOMEM:
+		response = VF_PORT_PROFILE_RESPONSE_INSUFFICIENT_RESOURCES;
+		break;
+	default:
+		response = VF_PORT_PROFILE_RESPONSE_ERROR;
+		break;
+	}
+
+	NLA_PUT_U16(skb, IFLA_VF_PORT_RESPONSE, response);
+	NLA_PUT(skb, IFLA_VF_PORT_PROFILE, VF_PORT_PROFILE_MAX,
+		enic->pp.name);
+	NLA_PUT(skb, IFLA_VF_PORT_INSTANCE_UUID, VF_PORT_UUID_MAX,
+		enic->pp.instance_uuid);
+	NLA_PUT(skb, IFLA_VF_PORT_HOST_UUID, VF_PORT_UUID_MAX,
+		enic->pp.host_uuid);
+
+	return 0;
+
+nla_put_failure:
+	return -EMSGSIZE;
+}
+
 static void enic_free_rq_buf(struct vnic_rq *rq, struct vnic_rq_buf *buf)
 {
 	struct enic *enic = vnic_dev_priv(rq->vdev);
@@ -1440,9 +1711,7 @@ static int enic_open(struct net_device *netdev)
 	for (i = 0; i < enic->rq_count; i++)
 		vnic_rq_enable(&enic->rq[i]);
 
-	spin_lock(&enic->devcmd_lock);
-	enic_add_station_addr(enic);
-	spin_unlock(&enic->devcmd_lock);
+	enic_dev_add_station_addr(enic);
 	enic_set_multicast_list(netdev);
 
 	netif_wake_queue(netdev);
@@ -1489,6 +1758,8 @@ static int enic_stop(struct net_device *netdev)
 	netif_carrier_off(netdev);
 	netif_tx_disable(netdev);
 
+	enic_dev_del_station_addr(enic);
+
 	for (i = 0; i < enic->wq_count; i++) {
 		err = vnic_wq_disable(&enic->wq[i]);
 		if (err)
@@ -1775,20 +2046,22 @@ static void enic_clear_intr_mode(struct enic *enic)
 }
 
 static const struct net_device_ops enic_netdev_ops = {
-	.ndo_open		= enic_open,
-	.ndo_stop		= enic_stop,
-	.ndo_start_xmit		= enic_hard_start_xmit,
-	.ndo_get_stats		= enic_get_stats,
-	.ndo_validate_addr	= eth_validate_addr,
-	.ndo_set_mac_address 	= eth_mac_addr,
-	.ndo_set_multicast_list	= enic_set_multicast_list,
-	.ndo_change_mtu		= enic_change_mtu,
-	.ndo_vlan_rx_register	= enic_vlan_rx_register,
-	.ndo_vlan_rx_add_vid	= enic_vlan_rx_add_vid,
-	.ndo_vlan_rx_kill_vid	= enic_vlan_rx_kill_vid,
-	.ndo_tx_timeout		= enic_tx_timeout,
+	.ndo_open			= enic_open,
+	.ndo_stop			= enic_stop,
+	.ndo_start_xmit			= enic_hard_start_xmit,
+	.ndo_get_stats			= enic_get_stats,
+	.ndo_validate_addr		= eth_validate_addr,
+	.ndo_set_multicast_list		= enic_set_multicast_list,
+	.ndo_set_mac_address		= enic_set_mac_address,
+	.ndo_change_mtu			= enic_change_mtu,
+	.ndo_vlan_rx_register		= enic_vlan_rx_register,
+	.ndo_vlan_rx_add_vid		= enic_vlan_rx_add_vid,
+	.ndo_vlan_rx_kill_vid		= enic_vlan_rx_kill_vid,
+	.ndo_tx_timeout			= enic_tx_timeout,
+	.ndo_set_vf_port		= enic_set_vf_port,
+	.ndo_get_vf_port		= enic_get_vf_port,
 #ifdef CONFIG_NET_POLL_CONTROLLER
-	.ndo_poll_controller	= enic_poll_controller,
+	.ndo_poll_controller		= enic_poll_controller,
 #endif
 };
 
@@ -2010,11 +2283,13 @@ static int __devinit enic_probe(struct pci_dev *pdev,
 
 	netif_carrier_off(netdev);
 
-	err = vnic_dev_init(enic->vdev, 0);
-	if (err) {
-		printk(KERN_ERR PFX
-			"vNIC dev init failed, aborting.\n");
-		goto err_out_dev_close;
+	if (!enic_is_dynamic(enic)) {
+		err = vnic_dev_init(enic->vdev, 0);
+		if (err) {
+			printk(KERN_ERR PFX
+				"vNIC dev init failed, aborting.\n");
+			goto err_out_dev_close;
+		}
 	}
 
 	err = enic_dev_init(enic);
diff --git a/drivers/net/enic/enic_res.c b/drivers/net/enic/enic_res.c
index 02839bf..9b18840 100644
--- a/drivers/net/enic/enic_res.c
+++ b/drivers/net/enic/enic_res.c
@@ -103,11 +103,6 @@ int enic_get_vnic_config(struct enic *enic)
 	return 0;
 }
 
-void enic_add_station_addr(struct enic *enic)
-{
-	vnic_dev_add_addr(enic->vdev, enic->mac_addr);
-}
-
 void enic_add_multicast_addr(struct enic *enic, u8 *addr)
 {
 	vnic_dev_add_addr(enic->vdev, addr);
diff --git a/drivers/net/enic/enic_res.h b/drivers/net/enic/enic_res.h
index abc1974..494664f 100644
--- a/drivers/net/enic/enic_res.h
+++ b/drivers/net/enic/enic_res.h
@@ -131,7 +131,6 @@ static inline void enic_queue_rq_desc(struct vnic_rq *rq,
 struct enic;
 
 int enic_get_vnic_config(struct enic *);
-void enic_add_station_addr(struct enic *enic);
 void enic_add_multicast_addr(struct enic *enic, u8 *addr);
 void enic_del_multicast_addr(struct enic *enic, u8 *addr);
 void enic_add_vlan(struct enic *enic, u16 vlanid);
diff --git a/drivers/net/enic/vnic_dev.c b/drivers/net/enic/vnic_dev.c
index d43a9d4..2b3e16d 100644
--- a/drivers/net/enic/vnic_dev.c
+++ b/drivers/net/enic/vnic_dev.c
@@ -530,7 +530,7 @@ void vnic_dev_packet_filter(struct vnic_dev *vdev, int directed, int multicast,
 		printk(KERN_ERR "Can't set packet filter\n");
 }
 
-void vnic_dev_add_addr(struct vnic_dev *vdev, u8 *addr)
+int vnic_dev_add_addr(struct vnic_dev *vdev, u8 *addr)
 {
 	u64 a0 = 0, a1 = 0;
 	int wait = 1000;
@@ -543,9 +543,11 @@ void vnic_dev_add_addr(struct vnic_dev *vdev, u8 *addr)
 	err = vnic_dev_cmd(vdev, CMD_ADDR_ADD, &a0, &a1, wait);
 	if (err)
 		printk(KERN_ERR "Can't add addr [%pM], %d\n", addr, err);
+
+	return err;
 }
 
-void vnic_dev_del_addr(struct vnic_dev *vdev, u8 *addr)
+int vnic_dev_del_addr(struct vnic_dev *vdev, u8 *addr)
 {
 	u64 a0 = 0, a1 = 0;
 	int wait = 1000;
@@ -558,6 +560,8 @@ void vnic_dev_del_addr(struct vnic_dev *vdev, u8 *addr)
 	err = vnic_dev_cmd(vdev, CMD_ADDR_DEL, &a0, &a1, wait);
 	if (err)
 		printk(KERN_ERR "Can't del addr [%pM], %d\n", addr, err);
+
+	return err;
 }
 
 int vnic_dev_raise_intr(struct vnic_dev *vdev, u16 intr)
@@ -682,6 +686,56 @@ int vnic_dev_init(struct vnic_dev *vdev, int arg)
 	return r;
 }
 
+int vnic_dev_init_done(struct vnic_dev *vdev, int *done, int *err)
+{
+	u64 a0 = 0, a1 = 0;
+	int wait = 1000;
+	int ret;
+
+	*done = 0;
+
+	ret = vnic_dev_cmd(vdev, CMD_INIT_STATUS, &a0, &a1, wait);
+	if (ret)
+		return ret;
+
+	*done = (a0 == 0);
+
+	*err = (a0 == 0) ? a1 : 0;
+
+	return 0;
+}
+
+int vnic_dev_init_prov(struct vnic_dev *vdev, u8 *buf, u32 len)
+{
+	u64 a0, a1 = len;
+	int wait = 1000;
+	u64 prov_pa;
+	void *prov_buf;
+	int ret;
+
+	prov_buf = pci_alloc_consistent(vdev->pdev, len, &prov_pa);
+	if (!prov_buf)
+		return -ENOMEM;
+
+	memcpy(prov_buf, buf, len);
+
+	a0 = prov_pa;
+
+	ret = vnic_dev_cmd(vdev, CMD_INIT_PROV_INFO, &a0, &a1, wait);
+
+	pci_free_consistent(vdev->pdev, len, prov_buf, prov_pa);
+
+	return ret;
+}
+
+int vnic_dev_deinit(struct vnic_dev *vdev)
+{
+	u64 a0 = 0, a1 = 0;
+	int wait = 1000;
+
+	return vnic_dev_cmd(vdev, CMD_DEINIT, &a0, &a1, wait);
+}
+
 int vnic_dev_link_status(struct vnic_dev *vdev)
 {
 	if (vdev->linkstatus)
diff --git a/drivers/net/enic/vnic_dev.h b/drivers/net/enic/vnic_dev.h
index f5be640..caccce3 100644
--- a/drivers/net/enic/vnic_dev.h
+++ b/drivers/net/enic/vnic_dev.h
@@ -103,8 +103,8 @@ int vnic_dev_stats_dump(struct vnic_dev *vdev, struct vnic_stats **stats);
 int vnic_dev_hang_notify(struct vnic_dev *vdev);
 void vnic_dev_packet_filter(struct vnic_dev *vdev, int directed, int multicast,
 	int broadcast, int promisc, int allmulti);
-void vnic_dev_add_addr(struct vnic_dev *vdev, u8 *addr);
-void vnic_dev_del_addr(struct vnic_dev *vdev, u8 *addr);
+int vnic_dev_add_addr(struct vnic_dev *vdev, u8 *addr);
+int vnic_dev_del_addr(struct vnic_dev *vdev, u8 *addr);
 int vnic_dev_mac_addr(struct vnic_dev *vdev, u8 *mac_addr);
 int vnic_dev_raise_intr(struct vnic_dev *vdev, u16 intr);
 int vnic_dev_notify_setcmd(struct vnic_dev *vdev,
@@ -124,6 +124,9 @@ int vnic_dev_disable(struct vnic_dev *vdev);
 int vnic_dev_open(struct vnic_dev *vdev, int arg);
 int vnic_dev_open_done(struct vnic_dev *vdev, int *done);
 int vnic_dev_init(struct vnic_dev *vdev, int arg);
+int vnic_dev_init_done(struct vnic_dev *vdev, int *done, int *err);
+int vnic_dev_init_prov(struct vnic_dev *vdev, u8 *buf, u32 len);
+int vnic_dev_deinit(struct vnic_dev *vdev);
 int vnic_dev_soft_reset(struct vnic_dev *vdev, int arg);
 int vnic_dev_soft_reset_done(struct vnic_dev *vdev, int *done);
 void vnic_dev_set_intr_mode(struct vnic_dev *vdev,
diff --git a/drivers/net/enic/vnic_vic.c b/drivers/net/enic/vnic_vic.c
new file mode 100644
index 0000000..d769772
--- /dev/null
+++ b/drivers/net/enic/vnic_vic.c
@@ -0,0 +1,73 @@
+/*
+ * Copyright 2010 Cisco Systems, Inc.  All rights reserved.
+ *
+ * This program is free software; you may redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; version 2 of the License.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ *
+ */
+
+#include <linux/kernel.h>
+#include <linux/errno.h>
+#include <linux/types.h>
+#include <linux/slab.h>
+
+#include "vnic_vic.h"
+
+struct vic_provinfo *vic_provinfo_alloc(gfp_t flags, u8 *oui, u8 type)
+{
+	struct vic_provinfo *vp = kzalloc(VIC_PROVINFO_MAX_DATA, flags);
+
+	if (!vp || !oui)
+		return NULL;
+
+	memcpy(vp->oui, oui, sizeof(vp->oui));
+	vp->type = type;
+	vp->length = htonl(sizeof(vp->num_tlvs));
+
+	return vp;
+}
+
+void vic_provinfo_free(struct vic_provinfo *vp)
+{
+	kfree(vp);
+}
+
+int vic_provinfo_add_tlv(struct vic_provinfo *vp, u16 type, u16 length,
+	void *value)
+{
+	struct vic_provinfo_tlv *tlv;
+
+	if (!vp || !value)
+		return -EINVAL;
+
+	if (ntohl(vp->length) + sizeof(*tlv) + length >
+		VIC_PROVINFO_MAX_TLV_DATA)
+		return -ENOMEM;
+
+	tlv = (struct vic_provinfo_tlv *)((u8 *)vp->tlv +
+		ntohl(vp->length) - sizeof(vp->num_tlvs));
+
+	tlv->type = htons(type);
+	tlv->length = htons(length);
+	memcpy(tlv->value, value, length);
+
+	vp->num_tlvs = htonl(ntohl(vp->num_tlvs) + 1);
+	vp->length = htonl(ntohl(vp->length) + sizeof(*tlv) + length);
+
+	return 0;
+}
+
+size_t vic_provinfo_size(struct vic_provinfo *vp)
+{
+	return vp ?  ntohl(vp->length) + sizeof(*vp) - sizeof(vp->num_tlvs) : 0;
+}
diff --git a/drivers/net/enic/vnic_vic.h b/drivers/net/enic/vnic_vic.h
new file mode 100644
index 0000000..085c2a2
--- /dev/null
+++ b/drivers/net/enic/vnic_vic.h
@@ -0,0 +1,59 @@
+/*
+ * Copyright 2010 Cisco Systems, Inc.  All rights reserved.
+ *
+ * This program is free software; you may redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; version 2 of the License.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ *
+ */
+
+#ifndef _VNIC_VIC_H_
+#define _VNIC_VIC_H_
+
+/* Note: All integer fields in NETWORK byte order */
+
+/* Note: String field lengths include null char */
+
+#define VIC_PROVINFO_CISCO_OUI		{ 0x00, 0x00, 0x0c }
+#define VIC_PROVINFO_LINUX_TYPE		0x2
+
+enum vic_linux_prov_tlv_type {
+	VIC_LINUX_PROV_TLV_PORT_PROFILE_NAME_STR = 0,
+	VIC_LINUX_PROV_TLV_CLIENT_MAC_ADDR = 1,			/* u8[6] */
+	VIC_LINUX_PROV_TLV_CLIENT_NAME_STR = 2,
+	VIC_LINUX_PROV_TLV_HOST_UUID_STR = 8,
+	VIC_LINUX_PROV_TLV_CLIENT_UUID_STR = 9,
+};
+
+struct vic_provinfo {
+	u8 oui[3];		/* OUI of data provider */
+	u8 type;		/* provider-specific type */
+	u32 length;		/* length of data below */
+	u32 num_tlvs;		/* number of tlvs */
+	struct vic_provinfo_tlv {
+		u16 type;
+		u16 length;
+		u8 value[0];
+	} tlv[0];
+} __attribute__ ((packed));
+
+#define VIC_PROVINFO_MAX_DATA		1385
+#define VIC_PROVINFO_MAX_TLV_DATA (VIC_PROVINFO_MAX_DATA - \
+	sizeof(struct vic_provinfo))
+
+struct vic_provinfo *vic_provinfo_alloc(gfp_t flags, u8 *oui, u8 type);
+void vic_provinfo_free(struct vic_provinfo *vp);
+int vic_provinfo_add_tlv(struct vic_provinfo *vp, u16 type, u16 length,
+	void *value);
+size_t vic_provinfo_size(struct vic_provinfo *vp);
+
+#endif	/* _VNIC_VIC_H_ */

Re: [net-next-2.6 V6 PATCH 1/2] Add netlink support for virtual port management (was iovnl)

[Chris Wright]

* Scott Feldman (scofeldm@cisco.com) wrote:
> From: Scott Feldman <scofeldm@cisco.com>
> 
> Add new netdev ops ndo_{set|get}_vf_port to allow setting of
> port-profile on a netdev interface.  Extends netlink socket RTM_SETLINK/
> RTM_GETLINK with new sub cmd called IFLA_VF_PORT (added to end of
> IFLA_cmd list).
> 
> A port-profile is used to configure/enable the external switch virtual port
> backing the netdev interface, not to configure the host-facing side of the
> netdev.  A port-profile is an identifier known to the switch.  How port-
> profiles are installed on the switch or how available port-profiles are
> made know to the host is outside the scope of this patch.
> 
> There are two types of port-profiles specs in the netlink msg.  The first spec
> is for 802.1Qbg (pre-)standard, VDP protocol.  The second spec is for devices
> that run a similar protocol as VDP but in firmware, thus hiding the protocol
> details.  In either case, the specs have much in common and makes sense to
> define the netlink msg as the union of the two specs.  For example, both specs
> have a notition of associating/deassociating a port-profile.  And both specs
> require some information from the hypervisor manager, such as client port
> instance ID.
> 
> The general flow is the port-profile is applied to a host netdev interface
> using RTM_SETLINK, the receiver of the RTM_SETLINK msg communicates with the
> switch, and the switch virtual port backing the host netdev interface is
> configured/enabled based on the settings defined by the port-profile.  What
> those settings comprise, and how those settings are managed is again
> outside the scope of this patch, since this patch only deals with the
> first step in the flow.
> 
> There is a RTM_GETLINK cmd to to return port-profile setting of an
> interface and to also return the status of the last port-profile
> association.
> 
> IFLA_VF_PORT is modeled after the existing IFLA_VF_* cmd where a
> VF number is passed in to identify the virtual function (VF) of an SR-IOV-
> capable device.  In this case, the target of IFLA_VF_PORT msg is the
> netdev physical function (PF) device.  The PF will apply the port-profile
> to the VF.  IFLA_VF_PORT can also be used for devices that don't
> adhere to SR-IOV and can apply the port-profile directly to the netdev
> target.  In this case, the VF number is ignored.
> 
> Passing in a NULL port-profile is used to delete the port-profile association.
> 
> Acked-by: Arnd Bergmann <arnd@arndb.de>
> Signed-off-by: Scott Feldman <scofeldm@cisco.com>
> Signed-off-by: Roopa Prabhu<roprabhu@cisco.com>

Nice, this looks good to me.

Acked-by: Chris Wright <chrisw@redhat.com>

Re: [net-next-2.6 V6 PATCH 1/2] Add netlink support for virtual port management (was iovnl)

[Patrick McHardy]

Scott Feldman wrote:
> +struct ifla_vf_port_vsi {
> +	__u8 vsi_mgr_id;
> +	__u8 vsi_type_id[3];
> +	__u8 vsi_type_version;
> +	__u8 pad[3];
> +};

Where is this actually used? The only use I could find is in the
size calculation.

> diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
> index 23a71cb..de14d36 100644
> --- a/net/core/rtnetlink.c
> +++ b/net/core/rtnetlink.c
> +static int rtnl_vf_port_fill_nest(struct sk_buff *skb, struct net_device *dev,
> +	int vf)

Please keep the style used in that file consistent and align arguments
to the beginning of the opening '('.

> +{
> +	struct nlattr *data;
> +	int err;
> +
> +	data = nla_nest_start(skb, IFLA_VF_PORT);
> +	if (!data)
> +		return -EMSGSIZE;
> +
> +	if (vf >= 0)
> +		nla_put_u32(skb, IFLA_VF_PORT_VF, vf);
> +
> +	err = dev->netdev_ops->ndo_get_vf_port(dev, vf, skb);
> +	if (err == -EMSGSIZE) {
> +		nla_nest_cancel(skb, data);
> +		return -EMSGSIZE;
> +	} else if (err) {
> +		nla_nest_cancel(skb, data);
> +		return 0;

Why is the error not returned in this case?

> +	}
> +
> +	nla_nest_end(skb, data);
> +
> +	return 0;
> +}
> +
> +static int rtnl_vf_port_fill(struct sk_buff *skb, struct net_device *dev)
> +{
> +	int num_vf;
> +	int err;
> +
> +	if (!dev->netdev_ops->ndo_get_vf_port || !dev->dev.parent)
> +		return 0;
> +
> +	num_vf = dev_num_vf(dev->dev.parent);
> +
> +	if (num_vf) {
> +		int i;
> +
> +		for (i = 0; i < num_vf; i++) {
> +			err = rtnl_vf_port_fill_nest(skb, dev, i);
> +			if (err)
> +				goto nla_put_failure;
> +		}
> +	} else  {
> +		err = rtnl_vf_port_fill_nest(skb, dev, -1);

What does -1 mean?

> +		if (err)
> +			goto nla_put_failure;
> +	}
> +
> +	return 0;
> +
> +nla_put_failure:
> +	return -EMSGSIZE;
> +}
> +
>  static int rtnl_fill_ifinfo(struct sk_buff *skb, struct net_device *dev,
>  			    int type, u32 pid, u32 seq, u32 change,
>  			    unsigned int flags)
> @@ -747,17 +825,23 @@ static int rtnl_fill_ifinfo(struct sk_buff *skb, struct net_device *dev,
>  		goto nla_put_failure;
>  	copy_rtnl_link_stats64(nla_data(attr), stats);
>  
> +	if (dev->dev.parent)
> +		NLA_PUT_U32(skb, IFLA_NUM_VF, dev_num_vf(dev->dev.parent));

Should this attribute really be included even if the number is zero?

> +
>  	if (dev->netdev_ops->ndo_get_vf_config && dev->dev.parent) {
>  		int i;
>  		struct ifla_vf_info ivi;
>  
> -		NLA_PUT_U32(skb, IFLA_NUM_VF, dev_num_vf(dev->dev.parent));
>  		for (i = 0; i < dev_num_vf(dev->dev.parent); i++) {
>  			if (dev->netdev_ops->ndo_get_vf_config(dev, i, &ivi))
>  				break;
>  			NLA_PUT(skb, IFLA_VFINFO, sizeof(ivi), &ivi);
>  		}
>  	}
> +
> +	if (rtnl_vf_port_fill(skb, dev))
> +		goto nla_put_failure;
> +
>  	if (dev->rtnl_link_ops) {
>  		if (rtnl_link_fill(skb, dev) < 0)
>  			goto nla_put_failure;
> @@ -824,6 +908,7 @@ const struct nla_policy ifla_policy[IFLA_MAX+1] = {
>  				    .len = sizeof(struct ifla_vf_vlan) },
>  	[IFLA_VF_TX_RATE]	= { .type = NLA_BINARY,
>  				    .len = sizeof(struct ifla_vf_tx_rate) },
> +	[IFLA_VF_PORT]		= { .type = NLA_NESTED },
>  };
>  EXPORT_SYMBOL(ifla_policy);
>  
> @@ -832,6 +917,20 @@ static const struct nla_policy ifla_info_policy[IFLA_INFO_MAX+1] = {
>  	[IFLA_INFO_DATA]	= { .type = NLA_NESTED },
>  };
>  
> +static const struct nla_policy ifla_vf_port_policy[IFLA_VF_PORT_MAX+1] = {
> +	[IFLA_VF_PORT_VF]		= { .type = NLA_U32 },
> +	[IFLA_VF_PORT_PROFILE]		= { .type = NLA_STRING,
> +				.len = VF_PORT_PROFILE_MAX },

This is oddly indented, please align .len to .type as in the
existing attributes.

> +	[IFLA_VF_PORT_VSI_TYPE]		= { .type = NLA_BINARY,
> +				.len = sizeof(struct ifla_vf_port_vsi)},
> +	[IFLA_VF_PORT_INSTANCE_UUID]	= { .type = NLA_BINARY,
> +				.len = VF_PORT_UUID_MAX },
> +	[IFLA_VF_PORT_HOST_UUID]	= { .type = NLA_STRING,
> +				.len = VF_PORT_UUID_MAX },
> +	[IFLA_VF_PORT_REQUEST]		= { .type = NLA_U8, },
> +	[IFLA_VF_PORT_RESPONSE]		= { .type = NLA_U16, },
> +};
> +
>  struct net *rtnl_link_get_net(struct net *src_net, struct nlattr *tb[])
>  {
>  	struct net *net;
> @@ -1028,6 +1127,27 @@ static int do_setlink(struct net_device *dev, struct ifinfomsg *ifm,
>  	}
>  	err = 0;
>  
> +	if (tb[IFLA_VF_PORT]) {
> +		struct nlattr *vf_port[IFLA_VF_PORT_MAX+1];
> +		int vf = -1;
> +
> +		err = nla_parse_nested(vf_port, IFLA_VF_PORT_MAX,
> +			tb[IFLA_VF_PORT], ifla_vf_port_policy);
> +		if (err < 0)
> +			goto errout;
> +
> +		if (vf_port[IFLA_VF_PORT_VF])
> +			vf = nla_get_u32(vf_port[IFLA_VF_PORT_VF]);
> +		err = -EOPNOTSUPP;
> +		if (ops->ndo_set_vf_port)
> +			err = ops->ndo_set_vf_port(dev, vf, vf_port);

This appears to be addressing a single VF to issue commands.
I already explained this during the last set of VF patches,
messages are supposed to by symetrical, since you're dumping
state for all existing VFs, you also need to accept configuration
for multiple VFs. Basically, the kernel must be able to receive
a message it created during a dump and fully recreate the state.

Re: [net-next-2.6 V6 PATCH 1/2] Add netlink support for virtual port management (was iovnl)

[Chris Wright]

* Patrick McHardy (kaber@trash.net) wrote:
> > +	} else  {
> > +		err = rtnl_vf_port_fill_nest(skb, dev, -1);
> 
> What does -1 mean?

It means no VFs.  Could be made a macro/enum constant

thanks,
-chris

Re: [net-next-2.6 V6 PATCH 1/2] Add netlink support for virtual port management (was iovnl)

[Patrick McHardy]

Chris Wright wrote:
> * Patrick McHardy (kaber@trash.net) wrote:
>>> +	} else  {
>>> +		err = rtnl_vf_port_fill_nest(skb, dev, -1);
>> What does -1 mean?
> 
> It means no VFs.  Could be made a macro/enum constant

Why call rtnl_vg_port_fill_nest at all in that case? It even
calls the ndo_get_vf_port() callback.

Re: [net-next-2.6 V6 PATCH 1/2] Add netlink support for virtual port management (was iovnl)

[Chris Wright]

* Patrick McHardy (kaber@trash.net) wrote:
> Chris Wright wrote:
> > * Patrick McHardy (kaber@trash.net) wrote:
> >>> +	} else  {
> >>> +		err = rtnl_vf_port_fill_nest(skb, dev, -1);
> >> What does -1 mean?
> > 
> > It means no VFs.  Could be made a macro/enum constant
> 
> Why call rtnl_vg_port_fill_nest at all in that case? It even
> calls the ndo_get_vf_port() callback.

For the case where port profile is set on net dev that does not
have VFs (e.g. the enic case in 2/2).

thanks,
-chris

Re: [net-next-2.6 V6 PATCH 1/2] Add netlink support for virtual port management (was iovnl)

[Patrick McHardy]

Chris Wright wrote:
> * Patrick McHardy (kaber@trash.net) wrote:
>> Chris Wright wrote:
>>> * Patrick McHardy (kaber@trash.net) wrote:
>>>>> +	} else  {
>>>>> +		err = rtnl_vf_port_fill_nest(skb, dev, -1);
>>>> What does -1 mean?
>>> It means no VFs.  Could be made a macro/enum constant
>> Why call rtnl_vg_port_fill_nest at all in that case? It even
>> calls the ndo_get_vf_port() callback.
> 
> For the case where port profile is set on net dev that does not
> have VFs (e.g. the enic case in 2/2).

Thanks for the explanation. I guess a enum constant would be nice
to have. But the bigger problem is the asymetrical message
parsing/construction.

BTW:

> +enum {
> +	VF_PORT_REQUEST_PREASSOCIATE = 0,
> +	VF_PORT_REQUEST_PREASSOCIATE_RR,
> +	VF_PORT_REQUEST_ASSOCIATE,
> +	VF_PORT_REQUEST_DISASSOCIATE,
> +};

Do multiple of these commands have to be issued in order to
reach "associated" state? That also wouldn't fit into the
rtnetlink design, which contains state, not commands.

Re: [net-next-2.6 V6 PATCH 1/2] Add netlink support for virtual port management (was iovnl)

[Chris Wright]

* Patrick McHardy (kaber@trash.net) wrote:
> Chris Wright wrote:
> > * Patrick McHardy (kaber@trash.net) wrote:
> >> Chris Wright wrote:
> >>> * Patrick McHardy (kaber@trash.net) wrote:
> >>>>> +	} else  {
> >>>>> +		err = rtnl_vf_port_fill_nest(skb, dev, -1);
> >>>> What does -1 mean?
> >>> It means no VFs.  Could be made a macro/enum constant
> >> Why call rtnl_vg_port_fill_nest at all in that case? It even
> >> calls the ndo_get_vf_port() callback.
> > 
> > For the case where port profile is set on net dev that does not
> > have VFs (e.g. the enic case in 2/2).
> 
> Thanks for the explanation. I guess a enum constant would be nice
> to have. But the bigger problem is the asymetrical message
> parsing/construction.

Yeah, what would you like to do there?  I think we have to keep the
existing, just break out symmtetic set/get?

> BTW:
> 
> > +enum {
> > +	VF_PORT_REQUEST_PREASSOCIATE = 0,
> > +	VF_PORT_REQUEST_PREASSOCIATE_RR,
> > +	VF_PORT_REQUEST_ASSOCIATE,
> > +	VF_PORT_REQUEST_DISASSOCIATE,
> > +};
> 
> Do multiple of these commands have to be issued in order to
> reach "associated" state? That also wouldn't fit into the
> rtnetlink design, which contains state, not commands.

It's optional.  At the very least, you need 1 associate/disassociate for
each logical link up/down.

For VM migration or (perhaps failover modes) you can optionally issue a
preassociate.  Preassociate has 2 flavors.  One which is purely advisory,
another which will reserve resources on the switch.  These all reprsent
state transitions in the switch, but only associate should allow final
logical link up and traffic to flow.

thanks,
-chris

Re: [net-next-2.6 V6 PATCH 1/2] Add netlink support for virtual port management (was iovnl)

[Patrick McHardy]

Chris Wright wrote:
> * Patrick McHardy (kaber@trash.net) wrote:
>> Chris Wright wrote:
>>> * Patrick McHardy (kaber@trash.net) wrote:
>>>> Chris Wright wrote:
>>>>> * Patrick McHardy (kaber@trash.net) wrote:
>>>>>>> +	} else  {
>>>>>>> +		err = rtnl_vf_port_fill_nest(skb, dev, -1);
>>>>>> What does -1 mean?
>>>>> It means no VFs.  Could be made a macro/enum constant
>>>> Why call rtnl_vg_port_fill_nest at all in that case? It even
>>>> calls the ndo_get_vf_port() callback.
>>> For the case where port profile is set on net dev that does not
>>> have VFs (e.g. the enic case in 2/2).
>> Thanks for the explanation. I guess a enum constant would be nice
>> to have. But the bigger problem is the asymetrical message
>> parsing/construction.
> 
> Yeah, what would you like to do there?  I think we have to keep the
> existing, just break out symmtetic set/get?

Sure, that would be fine. I'll have a closer look at the exact
message layout tommorrow, its getting late here.

>> BTW:
>>
>>> +enum {
>>> +	VF_PORT_REQUEST_PREASSOCIATE = 0,
>>> +	VF_PORT_REQUEST_PREASSOCIATE_RR,
>>> +	VF_PORT_REQUEST_ASSOCIATE,
>>> +	VF_PORT_REQUEST_DISASSOCIATE,
>>> +};
>> Do multiple of these commands have to be issued in order to
>> reach "associated" state? That also wouldn't fit into the
>> rtnetlink design, which contains state, not commands.
> 
> It's optional.  At the very least, you need 1 associate/disassociate for
> each logical link up/down.
> 
> For VM migration or (perhaps failover modes) you can optionally issue a
> preassociate.  Preassociate has 2 flavors.  One which is purely advisory,
> another which will reserve resources on the switch.  These all reprsent
> state transitions in the switch, but only associate should allow final
> logical link up and traffic to flow.

I see, thanks. That seems fine.

Re: [net-next-2.6 V6 PATCH 1/2] Add netlink support for virtual port management (was iovnl)

[Scott Feldman]

On 5/13/10 1:40 PM, "Patrick McHardy" <kaber@trash.net> wrote:

> Scott Feldman wrote:
>> +struct ifla_vf_port_vsi {
>> + __u8 vsi_mgr_id;
>> + __u8 vsi_type_id[3];
>> + __u8 vsi_type_version;
>> + __u8 pad[3];
>> +};
> 
> Where is this actually used? The only use I could find is in the
> size calculation.

This is provisioned for VDP use.  The enic implementation (patch 2/2)
doesn't use these members.
 
> Please keep the style used in that file consistent and align arguments
> to the beginning of the opening '('.

I'll fix.
 
>> +{
>> + struct nlattr *data;
>> + int err;
>> +
>> + data = nla_nest_start(skb, IFLA_VF_PORT);
>> + if (!data)
>> +  return -EMSGSIZE;
>> +
>> + if (vf >= 0)
>> +  nla_put_u32(skb, IFLA_VF_PORT_VF, vf);
>> +
>> + err = dev->netdev_ops->ndo_get_vf_port(dev, vf, skb);
>> + if (err == -EMSGSIZE) {
>> +  nla_nest_cancel(skb, data);
>> +  return -EMSGSIZE;
>> + } else if (err) {
>> +  nla_nest_cancel(skb, data);
>> +  return 0;
> 
> Why is the error not returned in this case?

I was think the netdev could fail the call if the operation wasn't
supported.  I better choice would be to not set the netdev->op in the first
place.  Let me fix this.

>> +  if (err)
>> +   goto nla_put_failure;
>> + }
>> +
>> + return 0;
>> +
>> +nla_put_failure:
>> + return -EMSGSIZE;
>> +}
>> +
>>  static int rtnl_fill_ifinfo(struct sk_buff *skb, struct net_device *dev,
>>    int type, u32 pid, u32 seq, u32 change,
>>    unsigned int flags)
>> @@ -747,17 +825,23 @@ static int rtnl_fill_ifinfo(struct sk_buff *skb, struct
>> net_device *dev,
>> goto nla_put_failure;
>> copy_rtnl_link_stats64(nla_data(attr), stats);
>>  
>> + if (dev->dev.parent)
>> +  NLA_PUT_U32(skb, IFLA_NUM_VF, dev_num_vf(dev->dev.parent));
> 
> Should this attribute really be included even if the number is zero?

The previous code would write zero also.  I moved it out of the
get_vf_config check so it could be used for get_vf_port as well.
 
> This is oddly indented, please align .len to .type as in the
> existing attributes.

I'll fix, but bumping into 80 char issues...
 
>> + [IFLA_VF_PORT_VSI_TYPE]  = { .type = NLA_BINARY,
>> +    .len = sizeof(struct ifla_vf_port_vsi)},
>> + [IFLA_VF_PORT_INSTANCE_UUID] = { .type = NLA_BINARY,
>> +    .len = VF_PORT_UUID_MAX },
>> + [IFLA_VF_PORT_HOST_UUID] = { .type = NLA_STRING,
>> +    .len = VF_PORT_UUID_MAX },
>> + [IFLA_VF_PORT_REQUEST]  = { .type = NLA_U8, },
>> + [IFLA_VF_PORT_RESPONSE]  = { .type = NLA_U16, },
>> +};
>> +
>>  struct net *rtnl_link_get_net(struct net *src_net, struct nlattr *tb[])
>>  {
>> struct net *net;
>> @@ -1028,6 +1127,27 @@ static int do_setlink(struct net_device *dev, struct
>> ifinfomsg *ifm,
>> }
>> err = 0;
>>  
>> + if (tb[IFLA_VF_PORT]) {
>> +  struct nlattr *vf_port[IFLA_VF_PORT_MAX+1];
>> +  int vf = -1;
>> +
>> +  err = nla_parse_nested(vf_port, IFLA_VF_PORT_MAX,
>> +   tb[IFLA_VF_PORT], ifla_vf_port_policy);
>> +  if (err < 0)
>> +   goto errout;
>> +
>> +  if (vf_port[IFLA_VF_PORT_VF])
>> +   vf = nla_get_u32(vf_port[IFLA_VF_PORT_VF]);
>> +  err = -EOPNOTSUPP;
>> +  if (ops->ndo_set_vf_port)
>> +   err = ops->ndo_set_vf_port(dev, vf, vf_port);
> 
> This appears to be addressing a single VF to issue commands.
> I already explained this during the last set of VF patches,
> messages are supposed to by symetrical, since you're dumping
> state for all existing VFs, you also need to accept configuration
> for multiple VFs. Basically, the kernel must be able to receive
> a message it created during a dump and fully recreate the state.

This was modeled same as existing IFLA_VF_ cmd where single VF is addressed
on set, but all VFs for PF are dumped on get.

Re: [net-next-2.6 V6 PATCH 1/2] Add netlink support for virtual port management (was iovnl)

[Arnd Bergmann]

On Thursday 13 May 2010, Patrick McHardy wrote:
> > +enum {
> > +     VF_PORT_REQUEST_PREASSOCIATE = 0,
> > +     VF_PORT_REQUEST_PREASSOCIATE_RR,
> > +     VF_PORT_REQUEST_ASSOCIATE,
> > +     VF_PORT_REQUEST_DISASSOCIATE,
> > +};
> 
> Do multiple of these commands have to be issued in order to
> reach "associated" state? That also wouldn't fit into the
> rtnetlink design, which contains state, not commands.

In 802.1Qbg (see http://www.ieee802.org/1/files/public/docs2010/bg-joint-evb-0410v1.pdf)
they are defined both as state and as commands. I suggested
using the same definition so that we can use the exact values
when forming or receiving a TLV for the VDP protocol. We can change
freely between the first three during guest migration, but the port
(identified by the UUID) can only be in associate state on one
machine, and the switches would use that information to redirect
the data flow to the target of the migration.

The disassociate state is a bit strange in netlink because it
essentially means that the other information is now invalid,
but I'm not sure if there is a good alternative. If the switch
tells us that it is setting the link into disassociate state,
that should still be visible in netlink.

	Arnd

Re: [net-next-2.6 V6 PATCH 1/2] Add netlink support for virtual port management (was iovnl)

[Patrick McHardy]

Scott Feldman wrote:
> On 5/13/10 1:40 PM, "Patrick McHardy" <kaber@trash.net> wrote:
> 
>>> +  if (vf_port[IFLA_VF_PORT_VF])
>>> +   vf = nla_get_u32(vf_port[IFLA_VF_PORT_VF]);
>>> +  err = -EOPNOTSUPP;
>>> +  if (ops->ndo_set_vf_port)
>>> +   err = ops->ndo_set_vf_port(dev, vf, vf_port);
>> This appears to be addressing a single VF to issue commands.
>> I already explained this during the last set of VF patches,
>> messages are supposed to by symetrical, since you're dumping
>> state for all existing VFs, you also need to accept configuration
>> for multiple VFs. Basically, the kernel must be able to receive
>> a message it created during a dump and fully recreate the state.
> 
> This was modeled same as existing IFLA_VF_ cmd where single VF is addressed
> on set, but all VFs for PF are dumped on get.

Yes, that one should have been done differently as well,
unfortunately my comments were ignored. So far rtnetlink
had two properties that are now broken:

- messages sent by the kernel could be sent back to the
  kernel to re-create an object in the same state

- the same parsing functions could be used in userspace for
  messages sent by the kernel and netlink error messages,
  which contain the original userspace message

I know at least one program I've written a few years ago which
relies on the second property. Anyways, this is easily fixable
by encapsulating all top-level VF attributes in a list and
invoking the ndo_set_vf_port() callback for each VF configuration.