Implement get_link_ksettings for VXLAN and bridge
Subsystems (3)
v2 cover Implement get_link_ksettings for VXLAN and bridge
Mesh routing protocol batman-adv (in particular the new BATMAN_V algorithm) uses the link speed reported by get_link_ksettings to determine a path metric for wired links. In the mesh framework Gluon [1], we layer VXLAN and sometimes bridge interfaces on our Ethernet links.
These patches implement get_link_ksettings for these two interface types. While this is obviously not accurate for bridges with multiple active ports, it's much better than having no estimate at all (and in the particular setup of Gluon, bridges with a single port aren't completely uncommon).
[1] https://github.com/freifunk-gluon/gluon
Matthias Schiffer (2):
vxlan: implement get_link_ksettings ethtool method
bridge: implement get_link_ksettings ethtool method
drivers/net/vxlan.c | 24 ++++++++++++++++++++++-- net/bridge/br_device.c | 36 ++++++++++++++++++++++++++++++++++-- 2 files changed, 56 insertions(+), 4 deletions(-)
Signature · 1 line
2.24.0
PATCH v2 1/2 vxlan: implement get_link_ksettings ethtool method
Similar to VLAN and similar drivers, we can forward get_link_ksettings to the lower dev if we have one to get meaningful speed/duplex data.
---
v2: no changes
drivers/net/vxlan.c | 24 ++++++++++++++++++++++-- 1 file changed, 22 insertions(+), 2 deletions(-)
drivers/net/vxlan.c +22 -2
@@ -3175,9 +3175,29 @@ static void vxlan_get_drvinfo(struct net_device *netdev,strlcpy (drvinfo->driver,"vxlan" ,sizeof (drvinfo->driver)); } +static int vxlan_get_link_ksettings (struct net_device *dev, +struct ethtool_link_ksettings *cmd) +{ +struct vxlan_dev *vxlan =netdev_priv (dev); +struct vxlan_rdst *dst = &vxlan->default_dst; +struct net_device *lowerdev =__dev_get_by_index (vxlan->net, + dst->remote_ifindex); + +if (!lowerdev) { + cmd->base.duplex =DUPLEX_UNKNOWN ; + cmd->base.port =PORT_OTHER ; + cmd->base.speed =SPEED_UNKNOWN ; + +return 0 ; + } + +return __ethtool_get_link_ksettings (lowerdev, cmd); +} +static const struct ethtool_ops vxlan_ethtool_ops = { - .get_drvinfo = vxlan_get_drvinfo, - .get_link = ethtool_op_get_link, + .get_drvinfo = vxlan_get_drvinfo, + .get_link = ethtool_op_get_link, + .get_link_ksettings = vxlan_get_link_ksettings, }; static struct socket *vxlan_create_sock(struct net *net, bool ipv6,
Signature · 1 line
2.24.0
PATCH v2 2/2 bridge: implement get_link_ksettings ethtool method
In reply to Matthias SchifferWe return the maximum speed of all active ports. This matches how the link speed would give an upper limit for traffic to/from any single peer if the bridge were replaced with a hardware switch.
---
v2: code style
net/bridge/br_device.c | 36 ++++++++++++++++++++++++++++++++++-- 1 file changed, 34 insertions(+), 2 deletions(-)
net/bridge/br_device.c +34 -2
@@ -263,6 +263,37 @@ static void br_getinfo(struct net_device *dev, struct ethtool_drvinfo *info)strlcpy (info->bus_info,"N/A" ,sizeof (info->bus_info)); } +static int br_get_link_ksettings (struct net_device *dev, +struct ethtool_link_ksettings *cmd) +{ +struct net_bridge *br =netdev_priv (dev); +struct net_bridge_port *p; + + cmd->base.duplex =DUPLEX_UNKNOWN ; + cmd->base.port =PORT_OTHER ; + cmd->base.speed =SPEED_UNKNOWN ; + +list_for_each_entry (p, &br->port_list, list) { +struct ethtool_link_ksettings ecmd; +struct net_device *pdev = p->dev; + +if (!netif_running (pdev) || !netif_oper_up (pdev)) +continue ; + +if (__ethtool_get_link_ksettings (pdev, &ecmd)) +continue ; + +if (ecmd.base.speed == (__u32 )SPEED_UNKNOWN ) +continue ; + +if (cmd->base.speed == (__u32 )SPEED_UNKNOWN || + cmd->base.speed < ecmd.base.speed) + cmd->base.speed = ecmd.base.speed; + } + +return 0 ; +} +static netdev_features_t br_fix_features (struct net_device *dev,netdev_features_t features) {
@@ -365,8 +396,9 @@ static int br_del_slave(struct net_device *dev, struct net_device *slave_dev) }static const struct ethtool_ops br_ethtool_ops = { - .get_drvinfo = br_getinfo, - .get_link = ethtool_op_get_link, + .get_drvinfo = br_getinfo, + .get_link = ethtool_op_get_link, + .get_link_ksettings = br_get_link_ksettings, };static const struct net_device_ops br_netdev_ops = {
Signature · 1 line
2.24.0
David Miller
maintainerFrom: Matthias Schiffer <[email protected]>
Date: Tue, 12 Nov 2019 22:12:23 +0100
Quoted 5 lines
Mesh routing protocol batman-adv (in particular the new
BATMAN_Valgorithm) uses the link speed reported byget_link_ksettingsto determine a path metric for wired links. In the mesh framework Gluon [1], we layer VXLAN and sometimes bridge interfaces on our Ethernet links.
Quoted 7 lines
These patches implement
get_link_ksettingsfor these two interface types. While this is obviously not accurate for bridges with multiple active ports, it's much better than having no estimate at all (and in the particular setup of Gluon, bridges with a single port aren't completely uncommon).
Series applied, thanks.