[PATCH rdma-next 0/4] Get IB port and node GUIDs through rtnetlink

[PATCH rdma-next 0/4] Get IB port and node GUIDs through rtnetlink

[Leon Romanovsky]

From: Leon Romanovsky <leonro@mellanox.com>

Hi,

This series from Danit extends RTNETLINK to provide IB port and node
GUIDs, which were configured for Infiniband VFs.

The functionality to set VF GUIDs already existed for a long time, and here
we are adding the missing "get" so that netlink will be symmetric
and various cloud orchestration tools will be able to manage such
VFs more naturally.

The iproute2 was extended too to present those GUIDs.

- ip link show <device>

For example:
- ip link set ib4 vf 0 node_guid 22:44:33:00:33:11:00:33
- ip link set ib4 vf 0 port_guid 10:21:33:12:00:11:22:10
- ip link show ib4
    ib4: <BROADCAST,MULTICAST> mtu 4092 qdisc noop state DOWN mode DEFAULT group default qlen 256
    link/infiniband 00:00:0a:2d:fe:80:00:00:00:00:00:00:ec:0d:9a:03:00:44:36:8d brd 00:ff:ff:ff:ff:12:40:1b:ff:ff:00:00:00:00:00:00:ff:ff:ff:ff
    vf 0     link/infiniband 00:00:0a:2d:fe:80:00:00:00:00:00:00:ec:0d:9a:03:00:44:36:8d brd 00:ff:ff:ff:ff:12:40:1b:ff:ff:00:00:00:00:00:00:ff:ff:ff:ff,
    spoof checking off, NODE_GUID 22:44:33:00:33:11:00:33, PORT_GUID 10:21:33:12:00:11:22:10, link-state disable, trust off, query_rss off

Due to the fact that this series touches both net and RDMA, we assume
that it needs to be applied to our shared branch (mlx5-next) and pulled
later by Dave and Doug/Jason.

Thanks

Danit Goldberg (4):
  net/core: Add support for getting VF GUIDs
  IB/core: Add interfaces to get VF node and port GUIDs
  IB/ipoib: Add ndo operation for getting VFs GUID attributes
  IB/mlx5: Implement callbacks for getting VFs GUID attributes

 drivers/infiniband/core/device.c          |  1 +
 drivers/infiniband/core/verbs.c           | 10 ++++++++++
 drivers/infiniband/hw/mlx5/ib_virt.c      | 24 +++++++++++++++++++++++
 drivers/infiniband/hw/mlx5/main.c         |  1 +
 drivers/infiniband/hw/mlx5/mlx5_ib.h      |  3 +++
 drivers/infiniband/ulp/ipoib/ipoib_main.c | 10 ++++++++++
 include/linux/netdevice.h                 |  4 ++++
 include/rdma/ib_verbs.h                   |  6 ++++++
 net/core/rtnetlink.c                      | 11 +++++++++++
 9 files changed, 70 insertions(+)

--
2.20.1

[PATCH iproute2-next] ip link: Add support to get SR-IOV VF node GUID and port GUID

[Leon Romanovsky]

From: Danit Goldberg <danitg@mellanox.com>

Extend iplink to show VF GUIDs (IFLA_VF_IB_NODE_GUID, IFLA_VF_IB_PORT_GUID),
giving the ability for user-space application to print GUID values.
This ability is added to the one of setting new node GUID and port GUID values.

Suitable ip link command:
- ip link show <device>

For example:
- ip link set ib4 vf 0 node_guid 22:44:33:00:33:11:00:33
- ip link set ib4 vf 0 port_guid 10:21:33:12:00:11:22:10
- ip link show ib4
ib4: <BROADCAST,MULTICAST> mtu 4092 qdisc noop state DOWN mode DEFAULT group default qlen 256
link/infiniband 00:00:0a:2d:fe:80:00:00:00:00:00:00:ec:0d:9a:03:00:44:36:8d brd 00:ff:ff:ff:ff:12:40:1b:ff:ff:00:00:00:00:00:00:ff:ff:ff:ff
vf 0     link/infiniband 00:00:0a:2d:fe:80:00:00:00:00:00:00:ec:0d:9a:03:00:44:36:8d brd 00:ff:ff:ff:ff:12:40:1b:ff:ff:00:00:00:00:00:00:ff:ff:ff:ff,
spoof checking off, NODE_GUID 22:44:33:00:33:11:00:33, PORT_GUID 10:21:33:12:00:11:22:10, link-state disable, trust off, query_rss off

Signed-off-by: Danit Goldberg <danitg@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
---
 ip/ipaddress.c | 23 +++++++++++++++++++++++
 1 file changed, 23 insertions(+)

diff --git a/ip/ipaddress.c b/ip/ipaddress.c
index b72eb7a1..ed72d0bd 100644
--- a/ip/ipaddress.c
+++ b/ip/ipaddress.c
@@ -484,6 +484,29 @@ static void print_vfinfo(FILE *fp, struct ifinfomsg *ifi, struct rtattr *vfinfo)
 				   vf_spoofchk->setting);
 	}
 
+#define GUID_STR_LEN 24
+	if (vf[IFLA_VF_IB_NODE_GUID]) {
+		char buf[GUID_STR_LEN];
+		struct ifla_vf_guid *guid = RTA_DATA(vf[IFLA_VF_IB_NODE_GUID]);
+		uint64_t node_guid = ntohll(guid->guid);
+
+		print_string(PRINT_ANY, "node guid", ", NODE_GUID %s",
+				ll_addr_n2a((const unsigned char *)&node_guid,
+					 RTA_PAYLOAD(vf[IFLA_VF_IB_NODE_GUID]),
+					 ARPHRD_INFINIBAND,
+					 buf, sizeof(buf)));
+	}
+	if (vf[IFLA_VF_IB_PORT_GUID]) {
+		char buf[GUID_STR_LEN];
+		struct ifla_vf_guid *guid = RTA_DATA(vf[IFLA_VF_IB_PORT_GUID]);
+		uint64_t port_guid = ntohll(guid->guid);
+
+		print_string(PRINT_ANY, "port guid", ", PORT_GUID %s",
+				ll_addr_n2a((const unsigned char *)&port_guid,
+					 RTA_PAYLOAD(vf[IFLA_VF_IB_PORT_GUID]),
+					 ARPHRD_INFINIBAND,
+					 buf, sizeof(buf)));
+	}
 	if (vf[IFLA_VF_LINK_STATE]) {
 		struct ifla_vf_link_state *vf_linkstate =
 			RTA_DATA(vf[IFLA_VF_LINK_STATE]);
-- 
2.20.1

[PATCH rdma-next 1/4] net/core: Add support for getting VF GUIDs

[Leon Romanovsky]

From: Danit Goldberg <danitg@mellanox.com>


Introduce a new ndo: ndo_get_vf_guid, to get from the net
device the port and node GUID.

New applications can choose to use this interface to show
GUIDs with iproute2 with commands such as:

- ip link show ib4
ib4: <BROADCAST,MULTICAST> mtu 4092 qdisc noop state DOWN mode DEFAULT group default qlen 256
link/infiniband 00:00:0a:2d:fe:80:00:00:00:00:00:00:ec:0d:9a:03:00:44:36:8d brd 00:ff:ff:ff:ff:12:40:1b:ff:ff:00:00:00:00:00:00:ff:ff:ff:ff
vf 0     link/infiniband 00:00:0a:2d:fe:80:00:00:00:00:00:00:ec:0d:9a:03:00:44:36:8d brd 00:ff:ff:ff:ff:12:40:1b:ff:ff:00:00:00:00:00:00:ff:ff:ff:ff,
spoof checking off, NODE_GUID 22:44:33:00:33:11:00:33, PORT_GUID 10:21:33:12:00:11:22:10, link-state disable, trust off, query_rss off

Signed-off-by: Danit Goldberg <danitg@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
---
 include/linux/netdevice.h |  4 ++++
 net/core/rtnetlink.c      | 11 +++++++++++
 2 files changed, 15 insertions(+)

diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 9eda1c31d1f7..379338239e49 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -1316,6 +1316,10 @@ struct net_device_ops {
 						   struct nlattr *port[]);
 	int			(*ndo_get_vf_port)(struct net_device *dev,
 						   int vf, struct sk_buff *skb);
+	int			(*ndo_get_vf_guid)(struct net_device *dev,
+						   int vf,
+						   struct ifla_vf_guid *node_guid,
+						   struct ifla_vf_guid *port_guid);
 	int			(*ndo_set_vf_guid)(struct net_device *dev,
 						   int vf, u64 guid,
 						   int guid_type);
diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
index 1ee6460f8275..b8d152f55a21 100644
--- a/net/core/rtnetlink.c
+++ b/net/core/rtnetlink.c
@@ -1204,6 +1204,8 @@ static noinline_for_stack int rtnl_fill_vfinfo(struct sk_buff *skb,
 	struct ifla_vf_mac vf_mac;
 	struct ifla_vf_broadcast vf_broadcast;
 	struct ifla_vf_info ivi;
+	struct ifla_vf_guid node_guid;
+	struct ifla_vf_guid port_guid;

 	memset(&ivi, 0, sizeof(ivi));

@@ -1270,6 +1272,15 @@ static noinline_for_stack int rtnl_fill_vfinfo(struct sk_buff *skb,
 	    nla_put(skb, IFLA_VF_TRUST,
 		    sizeof(vf_trust), &vf_trust))
 		goto nla_put_vf_failure;
+	if (dev->netdev_ops->ndo_get_vf_guid &&
+	    !dev->netdev_ops->ndo_get_vf_guid(dev, vfs_num, &node_guid,
+					      &port_guid)) {
+		if (nla_put(skb, IFLA_VF_IB_NODE_GUID, sizeof(node_guid),
+			    &node_guid) ||
+		    nla_put(skb, IFLA_VF_IB_PORT_GUID, sizeof(port_guid),
+			    &port_guid))
+			goto nla_put_vf_failure;
+	}
 	vfvlanlist = nla_nest_start_noflag(skb, IFLA_VF_VLAN_LIST);
 	if (!vfvlanlist)
 		goto nla_put_vf_failure;
--
2.20.1

[PATCH rdma-next 2/4] IB/core: Add interfaces to get VF node and port GUIDs

[Leon Romanovsky]

From: Danit Goldberg <danitg@mellanox.com>


Provide ability to get node and port GUIDs of VFs to be symmetrical
to already existing set option.

Signed-off-by: Danit Goldberg <danitg@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
---
 drivers/infiniband/core/device.c |  1 +
 drivers/infiniband/core/verbs.c  | 10 ++++++++++
 include/rdma/ib_verbs.h          |  6 ++++++
 3 files changed, 17 insertions(+)

diff --git a/drivers/infiniband/core/device.c b/drivers/infiniband/core/device.c
index 1a789d2f7a38..e2229bab304a 100644
--- a/drivers/infiniband/core/device.c
+++ b/drivers/infiniband/core/device.c
@@ -2618,6 +2618,7 @@ void ib_set_device_ops(struct ib_device *dev, const struct ib_device_ops *ops)
 	SET_DEVICE_OP(dev_ops, get_port_immutable);
 	SET_DEVICE_OP(dev_ops, get_vector_affinity);
 	SET_DEVICE_OP(dev_ops, get_vf_config);
+	SET_DEVICE_OP(dev_ops, get_vf_guid);
 	SET_DEVICE_OP(dev_ops, get_vf_stats);
 	SET_DEVICE_OP(dev_ops, init_port);
 	SET_DEVICE_OP(dev_ops, invalidate_range);
diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c
index 5cf32bf484b7..da77057bfe69 100644
--- a/drivers/infiniband/core/verbs.c
+++ b/drivers/infiniband/core/verbs.c
@@ -2460,6 +2460,16 @@ int ib_set_vf_guid(struct ib_device *device, int vf, u8 port, u64 guid,
 }
 EXPORT_SYMBOL(ib_set_vf_guid);

+int ib_get_vf_guid(struct ib_device *device, int vf, u8 port,
+		   struct ifla_vf_guid *node_guid,
+		   struct ifla_vf_guid *port_guid)
+{
+	if (!device->ops.get_vf_guid)
+		return -EOPNOTSUPP;
+
+	return device->ops.get_vf_guid(device, vf, port, node_guid, port_guid);
+}
+EXPORT_SYMBOL(ib_get_vf_guid);
 /**
  * ib_map_mr_sg_pi() - Map the dma mapped SG lists for PI (protection
  *     information) and set an appropriate memory region for registration.
diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h
index f9e121b03158..d8b73a89ca1a 100644
--- a/include/rdma/ib_verbs.h
+++ b/include/rdma/ib_verbs.h
@@ -2472,6 +2472,9 @@ struct ib_device_ops {
 			     struct ifla_vf_info *ivf);
 	int (*get_vf_stats)(struct ib_device *device, int vf, u8 port,
 			    struct ifla_vf_stats *stats);
+	int (*get_vf_guid)(struct ib_device *device, int vf, u8 port,
+			    struct ifla_vf_guid *node_guid,
+			    struct ifla_vf_guid *port_guid);
 	int (*set_vf_guid)(struct ib_device *device, int vf, u8 port, u64 guid,
 			   int type);
 	struct ib_wq *(*create_wq)(struct ib_pd *pd,
@@ -3342,6 +3345,9 @@ int ib_get_vf_config(struct ib_device *device, int vf, u8 port,
 		     struct ifla_vf_info *info);
 int ib_get_vf_stats(struct ib_device *device, int vf, u8 port,
 		    struct ifla_vf_stats *stats);
+int ib_get_vf_guid(struct ib_device *device, int vf, u8 port,
+		    struct ifla_vf_guid *node_guid,
+		    struct ifla_vf_guid *port_guid);
 int ib_set_vf_guid(struct ib_device *device, int vf, u8 port, u64 guid,
 		   int type);

--
2.20.1

[PATCH rdma-next 3/4] IB/ipoib: Add ndo operation for getting VFs GUID attributes

[Leon Romanovsky]

From: Danit Goldberg <danitg@mellanox.com>

Add ndo operation to the network driver that enables configuring
ipoib_get_vf_guid operation. The operation allows to get a VF port
and node GUIDs.

Signed-off-by: Danit Goldberg <danitg@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
---
 drivers/infiniband/ulp/ipoib/ipoib_main.c | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c
index ac0583ff280d..e5f438ab716c 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_main.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c
@@ -2019,6 +2019,15 @@ static int ipoib_set_vf_guid(struct net_device *dev, int vf, u64 guid, int type)
 	return ib_set_vf_guid(priv->ca, vf, priv->port, guid, type);
 }

+static int ipoib_get_vf_guid(struct net_device *dev, int vf,
+			     struct ifla_vf_guid *node_guid,
+			     struct ifla_vf_guid *port_guid)
+{
+	struct ipoib_dev_priv *priv = ipoib_priv(dev);
+
+	return ib_get_vf_guid(priv->ca, vf, priv->port, node_guid, port_guid);
+}
+
 static int ipoib_get_vf_stats(struct net_device *dev, int vf,
 			      struct ifla_vf_stats *vf_stats)
 {
@@ -2045,6 +2054,7 @@ static const struct net_device_ops ipoib_netdev_ops_pf = {
 	.ndo_set_vf_link_state	 = ipoib_set_vf_link_state,
 	.ndo_get_vf_config	 = ipoib_get_vf_config,
 	.ndo_get_vf_stats	 = ipoib_get_vf_stats,
+	.ndo_get_vf_guid	 = ipoib_get_vf_guid,
 	.ndo_set_vf_guid	 = ipoib_set_vf_guid,
 	.ndo_set_mac_address	 = ipoib_set_mac,
 	.ndo_get_stats64	 = ipoib_get_stats,
--
2.20.1

[PATCH rdma-next 4/4] IB/mlx5: Implement callbacks for getting VFs GUID attributes

[Leon Romanovsky]

From: Danit Goldberg <danitg@mellanox.com>

Implement the IB defined callback mlx5_ib_get_vf_guid used to query FW
for VFs attributes and return node and port GUIDs.

Signed-off-by: Danit Goldberg <danitg@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
---
 drivers/infiniband/hw/mlx5/ib_virt.c | 24 ++++++++++++++++++++++++
 drivers/infiniband/hw/mlx5/main.c    |  1 +
 drivers/infiniband/hw/mlx5/mlx5_ib.h |  3 +++
 3 files changed, 28 insertions(+)

diff --git a/drivers/infiniband/hw/mlx5/ib_virt.c b/drivers/infiniband/hw/mlx5/ib_virt.c
index 649a3364f838..4f0edd4832bd 100644
--- a/drivers/infiniband/hw/mlx5/ib_virt.c
+++ b/drivers/infiniband/hw/mlx5/ib_virt.c
@@ -201,3 +201,27 @@ int mlx5_ib_set_vf_guid(struct ib_device *device, int vf, u8 port,

 	return -EINVAL;
 }
+
+int mlx5_ib_get_vf_guid(struct ib_device *device, int vf, u8 port,
+			struct ifla_vf_guid *node_guid,
+			struct ifla_vf_guid *port_guid)
+{
+	struct mlx5_ib_dev *dev = to_mdev(device);
+	struct mlx5_core_dev *mdev = dev->mdev;
+	struct mlx5_hca_vport_context *rep;
+	int err;
+
+	rep = kzalloc(sizeof(*rep), GFP_KERNEL);
+	if (!rep)
+		return -ENOMEM;
+
+	err = mlx5_query_hca_vport_context(mdev, 1, 1, vf+1, rep);
+	if (err)
+		goto ex;
+
+	port_guid->guid = rep->port_guid;
+	node_guid->guid = rep->node_guid;
+ex:
+	kfree(rep);
+	return err;
+}
diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index 2d83f8c1d7ba..7127cdd313c1 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -6313,6 +6313,7 @@ static const struct ib_device_ops mlx5_ib_dev_ipoib_enhanced_ops = {

 static const struct ib_device_ops mlx5_ib_dev_sriov_ops = {
 	.get_vf_config = mlx5_ib_get_vf_config,
+	.get_vf_guid = mlx5_ib_get_vf_guid,
 	.get_vf_stats = mlx5_ib_get_vf_stats,
 	.set_vf_guid = mlx5_ib_set_vf_guid,
 	.set_vf_link_state = mlx5_ib_set_vf_link_state,
diff --git a/drivers/infiniband/hw/mlx5/mlx5_ib.h b/drivers/infiniband/hw/mlx5/mlx5_ib.h
index 00fd412ef686..3ad20b120bf4 100644
--- a/drivers/infiniband/hw/mlx5/mlx5_ib.h
+++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h
@@ -1315,6 +1315,9 @@ int mlx5_ib_set_vf_link_state(struct ib_device *device, int vf,
 			      u8 port, int state);
 int mlx5_ib_get_vf_stats(struct ib_device *device, int vf,
 			 u8 port, struct ifla_vf_stats *stats);
+int mlx5_ib_get_vf_guid(struct ib_device *device, int vf, u8 port,
+			struct ifla_vf_guid *node_guid,
+			struct ifla_vf_guid *port_guid);
 int mlx5_ib_set_vf_guid(struct ib_device *device, int vf, u8 port,
 			u64 guid, int type);

--
2.20.1

Re: [PATCH iproute2-next] ip link: Add support to get SR-IOV VF node GUID and port GUID

[David Ahern]

On 11/14/19 6:31 AM, Leon Romanovsky wrote:
> diff --git a/ip/ipaddress.c b/ip/ipaddress.c
> index b72eb7a1..ed72d0bd 100644
> --- a/ip/ipaddress.c
> +++ b/ip/ipaddress.c
> @@ -484,6 +484,29 @@ static void print_vfinfo(FILE *fp, struct ifinfomsg *ifi, struct rtattr *vfinfo)
>  				   vf_spoofchk->setting);
>  	}
>  
> +#define GUID_STR_LEN 24
> +	if (vf[IFLA_VF_IB_NODE_GUID]) {
> +		char buf[GUID_STR_LEN];

buf should be declared with SPRINT_BUF; see other users of ll_addr_n2a.
And, print_vfinfo already has b1 declared so you do not need a new one;
just change buf to b1.


> +		struct ifla_vf_guid *guid = RTA_DATA(vf[IFLA_VF_IB_NODE_GUID]);
> +		uint64_t node_guid = ntohll(guid->guid);
> +
> +		print_string(PRINT_ANY, "node guid", ", NODE_GUID %s",
> +				ll_addr_n2a((const unsigned char *)&node_guid,
> +					 RTA_PAYLOAD(vf[IFLA_VF_IB_NODE_GUID]),
> +					 ARPHRD_INFINIBAND,
> +					 buf, sizeof(buf)));
> +	}
> +	if (vf[IFLA_VF_IB_PORT_GUID]) {
> +		char buf[GUID_STR_LEN];
> +		struct ifla_vf_guid *guid = RTA_DATA(vf[IFLA_VF_IB_PORT_GUID]);
> +		uint64_t port_guid = ntohll(guid->guid);
> +
> +		print_string(PRINT_ANY, "port guid", ", PORT_GUID %s",
> +				ll_addr_n2a((const unsigned char *)&port_guid,
> +					 RTA_PAYLOAD(vf[IFLA_VF_IB_PORT_GUID]),
> +					 ARPHRD_INFINIBAND,
> +					 buf, sizeof(buf)));
> +	}
>  	if (vf[IFLA_VF_LINK_STATE]) {
>  		struct ifla_vf_link_state *vf_linkstate =
>  			RTA_DATA(vf[IFLA_VF_LINK_STATE]);
> 

Re: [PATCH iproute2-next] ip link: Add support to get SR-IOV VF node GUID and port GUID

[Leon Romanovsky]

On Thu, Nov 14, 2019 at 01:35:11PM -0700, David Ahern wrote:
> On 11/14/19 6:31 AM, Leon Romanovsky wrote:
> > diff --git a/ip/ipaddress.c b/ip/ipaddress.c
> > index b72eb7a1..ed72d0bd 100644
> > --- a/ip/ipaddress.c
> > +++ b/ip/ipaddress.c
> > @@ -484,6 +484,29 @@ static void print_vfinfo(FILE *fp, struct ifinfomsg *ifi, struct rtattr *vfinfo)
> >  				   vf_spoofchk->setting);
> >  	}
> >
> > +#define GUID_STR_LEN 24
> > +	if (vf[IFLA_VF_IB_NODE_GUID]) {
> > +		char buf[GUID_STR_LEN];
>
> buf should be declared with SPRINT_BUF; see other users of ll_addr_n2a.
> And, print_vfinfo already has b1 declared so you do not need a new one;
> just change buf to b1.

Thanks, I posted v1.
https://lore.kernel.org/linux-rdma/20191115152155.246821-1-leon@kernel.org

Re: [PATCH rdma-next 0/4] Get IB port and node GUIDs through rtnetlink

[Leon Romanovsky]

On Thu, Nov 14, 2019 at 03:31:21PM +0200, Leon Romanovsky wrote:
> From: Leon Romanovsky <leonro@mellanox.com>
>
> Hi,
>
> This series from Danit extends RTNETLINK to provide IB port and node
> GUIDs, which were configured for Infiniband VFs.
>
> The functionality to set VF GUIDs already existed for a long time, and here
> we are adding the missing "get" so that netlink will be symmetric
> and various cloud orchestration tools will be able to manage such
> VFs more naturally.
>
> The iproute2 was extended too to present those GUIDs.
>
> - ip link show <device>
>
> For example:
> - ip link set ib4 vf 0 node_guid 22:44:33:00:33:11:00:33
> - ip link set ib4 vf 0 port_guid 10:21:33:12:00:11:22:10
> - ip link show ib4
>     ib4: <BROADCAST,MULTICAST> mtu 4092 qdisc noop state DOWN mode DEFAULT group default qlen 256
>     link/infiniband 00:00:0a:2d:fe:80:00:00:00:00:00:00:ec:0d:9a:03:00:44:36:8d brd 00:ff:ff:ff:ff:12:40:1b:ff:ff:00:00:00:00:00:00:ff:ff:ff:ff
>     vf 0     link/infiniband 00:00:0a:2d:fe:80:00:00:00:00:00:00:ec:0d:9a:03:00:44:36:8d brd 00:ff:ff:ff:ff:12:40:1b:ff:ff:00:00:00:00:00:00:ff:ff:ff:ff,
>     spoof checking off, NODE_GUID 22:44:33:00:33:11:00:33, PORT_GUID 10:21:33:12:00:11:22:10, link-state disable, trust off, query_rss off
>
> Due to the fact that this series touches both net and RDMA, we assume
> that it needs to be applied to our shared branch (mlx5-next) and pulled
> later by Dave and Doug/Jason.
>
> Thanks
>
> Danit Goldberg (4):
>   net/core: Add support for getting VF GUIDs
>   IB/core: Add interfaces to get VF node and port GUIDs
>   IB/ipoib: Add ndo operation for getting VFs GUID attributes
>   IB/mlx5: Implement callbacks for getting VFs GUID attributes

Dave,

I see that you marked these patches as "not applicable" in patchworks.

Can I assume that you are OK with them and we can take them through RDMA
tree? If yes, can you give your Ack on the first patch?

Thanks

Re: [PATCH rdma-next 0/4] Get IB port and node GUIDs through rtnetlink

[Leon Romanovsky]

On Mon, Nov 18, 2019 at 07:25:02AM +0000, Leon Romanovsky wrote:
> On Thu, Nov 14, 2019 at 03:31:21PM +0200, Leon Romanovsky wrote:
> > From: Leon Romanovsky <leonro@mellanox.com>
> >
> > Hi,
> >
> > This series from Danit extends RTNETLINK to provide IB port and node
> > GUIDs, which were configured for Infiniband VFs.
> >
> > The functionality to set VF GUIDs already existed for a long time, and here
> > we are adding the missing "get" so that netlink will be symmetric
> > and various cloud orchestration tools will be able to manage such
> > VFs more naturally.
> >
> > The iproute2 was extended too to present those GUIDs.
> >
> > - ip link show <device>
> >
> > For example:
> > - ip link set ib4 vf 0 node_guid 22:44:33:00:33:11:00:33
> > - ip link set ib4 vf 0 port_guid 10:21:33:12:00:11:22:10
> > - ip link show ib4
> >     ib4: <BROADCAST,MULTICAST> mtu 4092 qdisc noop state DOWN mode DEFAULT group default qlen 256
> >     link/infiniband 00:00:0a:2d:fe:80:00:00:00:00:00:00:ec:0d:9a:03:00:44:36:8d brd 00:ff:ff:ff:ff:12:40:1b:ff:ff:00:00:00:00:00:00:ff:ff:ff:ff
> >     vf 0     link/infiniband 00:00:0a:2d:fe:80:00:00:00:00:00:00:ec:0d:9a:03:00:44:36:8d brd 00:ff:ff:ff:ff:12:40:1b:ff:ff:00:00:00:00:00:00:ff:ff:ff:ff,
> >     spoof checking off, NODE_GUID 22:44:33:00:33:11:00:33, PORT_GUID 10:21:33:12:00:11:22:10, link-state disable, trust off, query_rss off
> >
> > Due to the fact that this series touches both net and RDMA, we assume
> > that it needs to be applied to our shared branch (mlx5-next) and pulled
> > later by Dave and Doug/Jason.
> >
> > Thanks
> >
> > Danit Goldberg (4):
> >   net/core: Add support for getting VF GUIDs
> >   IB/core: Add interfaces to get VF node and port GUIDs
> >   IB/ipoib: Add ndo operation for getting VFs GUID attributes
> >   IB/mlx5: Implement callbacks for getting VFs GUID attributes
>
> Dave,
>
> I see that you marked these patches as "not applicable" in patchworks.
>
> Can I assume that you are OK with them and we can take them through RDMA
> tree? If yes, can you give your Ack on the first patch?

Kindly reminder.

Thanks

>
> Thanks

Re: [PATCH rdma-next 1/4] net/core: Add support for getting VF GUIDs

[David Ahern]

On 11/14/19 6:31 AM, Leon Romanovsky wrote:
> From: Danit Goldberg <danitg@mellanox.com>
> 
> 
> Introduce a new ndo: ndo_get_vf_guid, to get from the net
> device the port and node GUID.
> 
> New applications can choose to use this interface to show
> GUIDs with iproute2 with commands such as:
> 
> - ip link show ib4
> ib4: <BROADCAST,MULTICAST> mtu 4092 qdisc noop state DOWN mode DEFAULT group default qlen 256
> link/infiniband 00:00:0a:2d:fe:80:00:00:00:00:00:00:ec:0d:9a:03:00:44:36:8d brd 00:ff:ff:ff:ff:12:40:1b:ff:ff:00:00:00:00:00:00:ff:ff:ff:ff
> vf 0     link/infiniband 00:00:0a:2d:fe:80:00:00:00:00:00:00:ec:0d:9a:03:00:44:36:8d brd 00:ff:ff:ff:ff:12:40:1b:ff:ff:00:00:00:00:00:00:ff:ff:ff:ff,
> spoof checking off, NODE_GUID 22:44:33:00:33:11:00:33, PORT_GUID 10:21:33:12:00:11:22:10, link-state disable, trust off, query_rss off
> 
> Signed-off-by: Danit Goldberg <danitg@mellanox.com>
> Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
> ---
>  include/linux/netdevice.h |  4 ++++
>  net/core/rtnetlink.c      | 11 +++++++++++
>  2 files changed, 15 insertions(+)
> 


LGTM

Acked-by: David Ahern <dsahern@gmail.com>

Re: [PATCH rdma-next 1/4] net/core: Add support for getting VF GUIDs

[Leon Romanovsky]

On Thu, Nov 21, 2019 at 05:54:30PM -0700, David Ahern wrote:
> On 11/14/19 6:31 AM, Leon Romanovsky wrote:
> > From: Danit Goldberg <danitg@mellanox.com>
> >
> >
> > Introduce a new ndo: ndo_get_vf_guid, to get from the net
> > device the port and node GUID.
> >
> > New applications can choose to use this interface to show
> > GUIDs with iproute2 with commands such as:
> >
> > - ip link show ib4
> > ib4: <BROADCAST,MULTICAST> mtu 4092 qdisc noop state DOWN mode DEFAULT group default qlen 256
> > link/infiniband 00:00:0a:2d:fe:80:00:00:00:00:00:00:ec:0d:9a:03:00:44:36:8d brd 00:ff:ff:ff:ff:12:40:1b:ff:ff:00:00:00:00:00:00:ff:ff:ff:ff
> > vf 0     link/infiniband 00:00:0a:2d:fe:80:00:00:00:00:00:00:ec:0d:9a:03:00:44:36:8d brd 00:ff:ff:ff:ff:12:40:1b:ff:ff:00:00:00:00:00:00:ff:ff:ff:ff,
> > spoof checking off, NODE_GUID 22:44:33:00:33:11:00:33, PORT_GUID 10:21:33:12:00:11:22:10, link-state disable, trust off, query_rss off
> >
> > Signed-off-by: Danit Goldberg <danitg@mellanox.com>
> > Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
> > ---
> >  include/linux/netdevice.h |  4 ++++
> >  net/core/rtnetlink.c      | 11 +++++++++++
> >  2 files changed, 15 insertions(+)
> >
>
>
> LGTM
>
> Acked-by: David Ahern <dsahern@gmail.com>

Thanks a lot.

>
>

Re: [PATCH rdma-next 0/4] Get IB port and node GUIDs through rtnetlink

[Leon Romanovsky]

On Thu, Nov 14, 2019 at 03:31:21PM +0200, Leon Romanovsky wrote:
> From: Leon Romanovsky <leonro@mellanox.com>
>
> Hi,
>
> This series from Danit extends RTNETLINK to provide IB port and node
> GUIDs, which were configured for Infiniband VFs.
>
> The functionality to set VF GUIDs already existed for a long time, and here
> we are adding the missing "get" so that netlink will be symmetric
> and various cloud orchestration tools will be able to manage such
> VFs more naturally.
>
> The iproute2 was extended too to present those GUIDs.
>
> - ip link show <device>
>
> For example:
> - ip link set ib4 vf 0 node_guid 22:44:33:00:33:11:00:33
> - ip link set ib4 vf 0 port_guid 10:21:33:12:00:11:22:10
> - ip link show ib4
>     ib4: <BROADCAST,MULTICAST> mtu 4092 qdisc noop state DOWN mode DEFAULT group default qlen 256
>     link/infiniband 00:00:0a:2d:fe:80:00:00:00:00:00:00:ec:0d:9a:03:00:44:36:8d brd 00:ff:ff:ff:ff:12:40:1b:ff:ff:00:00:00:00:00:00:ff:ff:ff:ff
>     vf 0     link/infiniband 00:00:0a:2d:fe:80:00:00:00:00:00:00:ec:0d:9a:03:00:44:36:8d brd 00:ff:ff:ff:ff:12:40:1b:ff:ff:00:00:00:00:00:00:ff:ff:ff:ff,
>     spoof checking off, NODE_GUID 22:44:33:00:33:11:00:33, PORT_GUID 10:21:33:12:00:11:22:10, link-state disable, trust off, query_rss off
>
> Due to the fact that this series touches both net and RDMA, we assume
> that it needs to be applied to our shared branch (mlx5-next) and pulled
> later by Dave and Doug/Jason.
>
> Thanks
>
> Danit Goldberg (4):
>   net/core: Add support for getting VF GUIDs
>   IB/core: Add interfaces to get VF node and port GUIDs
>   IB/ipoib: Add ndo operation for getting VFs GUID attributes
>   IB/mlx5: Implement callbacks for getting VFs GUID attributes

Applied to mlx5-next,
Doug, Jason please pull.

9c0015ef0928 IB/mlx5: Implement callbacks for getting VFs GUID attributes
2446887ed226 IB/ipoib: Add ndo operation for getting VFs GUID attributes
bfcb3c5d1485 IB/core: Add interfaces to get VF node and port GUIDs
30aad41721e0 net/core: Add support for getting VF GUIDs

Thanks

Re: [PATCH rdma-next 0/4] Get IB port and node GUIDs through rtnetlink

[Jason Gunthorpe]

On Fri, Nov 22, 2019 at 04:21:07PM +0000, Leon Romanovsky wrote:
> On Thu, Nov 14, 2019 at 03:31:21PM +0200, Leon Romanovsky wrote:
> > From: Leon Romanovsky <leonro@mellanox.com>
> >
> > Hi,
> >
> > This series from Danit extends RTNETLINK to provide IB port and node
> > GUIDs, which were configured for Infiniband VFs.
> >
> > The functionality to set VF GUIDs already existed for a long time, and here
> > we are adding the missing "get" so that netlink will be symmetric
> > and various cloud orchestration tools will be able to manage such
> > VFs more naturally.
> >
> > The iproute2 was extended too to present those GUIDs.
> >
> > - ip link show <device>
> >
> > For example:
> > - ip link set ib4 vf 0 node_guid 22:44:33:00:33:11:00:33
> > - ip link set ib4 vf 0 port_guid 10:21:33:12:00:11:22:10
> > - ip link show ib4
> >     ib4: <BROADCAST,MULTICAST> mtu 4092 qdisc noop state DOWN mode DEFAULT group default qlen 256
> >     link/infiniband 00:00:0a:2d:fe:80:00:00:00:00:00:00:ec:0d:9a:03:00:44:36:8d brd 00:ff:ff:ff:ff:12:40:1b:ff:ff:00:00:00:00:00:00:ff:ff:ff:ff
> >     vf 0     link/infiniband 00:00:0a:2d:fe:80:00:00:00:00:00:00:ec:0d:9a:03:00:44:36:8d brd 00:ff:ff:ff:ff:12:40:1b:ff:ff:00:00:00:00:00:00:ff:ff:ff:ff,
> >     spoof checking off, NODE_GUID 22:44:33:00:33:11:00:33, PORT_GUID 10:21:33:12:00:11:22:10, link-state disable, trust off, query_rss off
> >
> > Due to the fact that this series touches both net and RDMA, we assume
> > that it needs to be applied to our shared branch (mlx5-next) and pulled
> > later by Dave and Doug/Jason.
> >
> > Thanks
> >
> > Danit Goldberg (4):
> >   net/core: Add support for getting VF GUIDs
> >   IB/core: Add interfaces to get VF node and port GUIDs
> >   IB/ipoib: Add ndo operation for getting VFs GUID attributes
> >   IB/mlx5: Implement callbacks for getting VFs GUID attributes
> 
> Applied to mlx5-next,
> Doug, Jason please pull.
> 
> 9c0015ef0928 IB/mlx5: Implement callbacks for getting VFs GUID attributes
> 2446887ed226 IB/ipoib: Add ndo operation for getting VFs GUID attributes
> bfcb3c5d1485 IB/core: Add interfaces to get VF node and port GUIDs
> 30aad41721e0 net/core: Add support for getting VF GUIDs

Okay, done.

Jason