can: add Virtual CAN Tunnel driver (vxcan)
RFC PATCH 1/1 can: add Virtual CAN Tunnel driver (vxcan)
Signed-off-by Oliver Hartkopp <[email protected]>
--- drivers/net/can/Kconfig | 18 +++ drivers/net/can/Makefile | 1 + drivers/net/can/vxcan.c | 313 +++++++++++++++++++++++++++++++++++++++++ include/uapi/linux/can/vxcan.h | 12 ++ 4 files changed, 344 insertions(+) create mode 100644 drivers/net/can/vxcan.c create mode 100644 include/uapi/linux/can/vxcan.h
drivers/net/can/Kconfig +18 -0
@@ -9,6 +9,24 @@ config CAN_VCAN This driver can also be built as a module. If so, the module will be called vcan. +config CAN_VXCAN + tristate "Virtual CAN Tunnel (vxcan)" + ---help--- + Similar to the virtual ethernet driver veth, vxcan implements a + local CAN traffic tunnel between two virtual CAN network devices. + When creating a vxcan, two vxcan devices are created as pair. + When one end receives the packet it appears on its pair and vice + versa. The vxcan can be used for cross namespace communication. + + In opposite to vcan loopback devices the vxcan only forwards CAN + frames to its pair and does *not* provide a local echo of sent + CAN frames. To disable a potential echo in af_can.c the vxcan driver + announces IFF_ECHO in the interface flags. To have a clean start + in each namespace the CAN GW hop counter is set to zero. + + This driver can also be built as a module. If so, the module + will be called vxcan. + config CAN_SLCAN tristate "Serial / USB serial CAN Adaptors (slcan)" depends on TTY
drivers/net/can/Makefile +1 -0
@@ -3,6 +3,7 @@ # obj-$(CONFIG_CAN_VCAN) += vcan.o +obj-$(CONFIG_CAN_VXCAN) += vxcan.o obj-$(CONFIG_CAN_SLCAN) += slcan.o obj-$(CONFIG_CAN_DEV) += can-dev.o
drivers/net/can/vxcan.c new file +313 -0
313 lines not on this page · raw message
include/uapi/linux/can/vxcan.h new file +12 -0
@@ -0,0 +1,12 @@ +#ifndef _UAPI_CAN_VXCAN_H +#define _UAPI_CAN_VXCAN_H + +enum { +VXCAN_INFO_UNSPEC , +VXCAN_INFO_PEER , + + __VXCAN_INFO_MAX +#define VXCAN_INFO_MAX (__VXCAN_INFO_MAX - 1) +}; + +#endif
Signature · 1 line
2.11.0
RFC PATCH 1/1 ip: link add vxcan support
Signed-off-by Oliver Hartkopp <[email protected]>
--- include/linux/can/vxcan.h | 12 +++++++ ip/Makefile | 2 +- ip/iplink_vxcan.c | 87 +++++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 100 insertions(+), 1 deletion(-) create mode 100644 include/linux/can/vxcan.h create mode 100644 ip/iplink_vxcan.c
include/linux/can/vxcan.h new file +12 -0
@@ -0,0 +1,12 @@ +#ifndef _UAPI_CAN_VXCAN_H +#define _UAPI_CAN_VXCAN_H + +enum { +VXCAN_INFO_UNSPEC , +VXCAN_INFO_PEER , + + __VXCAN_INFO_MAX +#define VXCAN_INFO_MAX (__VXCAN_INFO_MAX - 1) +}; + +#endif
ip/Makefile +1 -1
@@ -2,7 +2,7 @@ IPOBJ=ip.o ipaddress.o ipaddrlabel.o iproute.o iprule.o ipnetns.o \ rtm_map.o iptunnel.o ip6tunnel.o tunnel.o ipneigh.o ipntable.o iplink.o \ ipmaddr.o ipmonitor.o ipmroute.o ipprefix.o iptuntap.o iptoken.o \ ipxfrm.o xfrm_state.o xfrm_policy.o xfrm_monitor.o iplink_dummy.o \ - iplink_ifb.o iplink_nlmon.o iplink_team.o iplink_vcan.o \ + iplink_ifb.o iplink_nlmon.o iplink_team.o iplink_vcan.o iplink_vxcan.o \ iplink_vlan.o link_veth.o link_gre.o iplink_can.o iplink_xdp.o \ iplink_macvlan.o ipl2tp.o link_vti.o link_vti6.o \ iplink_vxlan.o tcp_metrics.o iplink_ipoib.o ipnetconf.o link_ip6tnl.o \
ip/iplink_vxcan.c new file +87 -0
87 lines not on this page · raw message
Signature · 1 line
2.11.0
Quoted 33 lines · Oliver Hartkopp
Signed-off-by Oliver Hartkopp <[email protected]>drivers/net/can/vxcan.c new file +28
-0@@ -0,0 +1,313 @@ +static netdev_tx_t vxcan_xmit (struct sk_buff *skb,struct net_device *dev) +{ +struct vxcan_priv *priv =netdev_priv (dev); +struct net_device *peer; +struct canfd_frame *cfd = (struct canfd_frame *)skb->data; +struct net_device_stats *stats = &dev->stats; + +if (can_dropped_invalid_skb (dev, skb)) +return NETDEV_TX_OK ; + +rcu_read_lock (); + peer =rcu_dereference (priv->peer); +if (unlikely (!peer)) { +kfree_skb (skb); + dev->stats.tx_dropped++; + goto out_unlock; + } + + skb =can_create_echo_skb (skb); +if (!skb) +return NETDEV_TX_OK ; + + stats->tx_packets++; + stats->tx_bytes += cfd->len; + +/* reset CAN GW hop counter */ + skb->csum_start =0 ; +
drivers/net/can/vxcan.c line 77
Quoted 10 lines
drivers/net/can/vxcan.c new file +10
-0+ stats = &peer->stats; + + stats->rx_packets++; + stats->rx_bytes += cfd->len; + + skb->pkt_type =PACKET_BROADCAST ; + skb->dev = peer; + skb->ip_summed =CHECKSUM_UNNECESSARY ; + +netif_rx_ni (skb);
^^^^^^^^^^
Wouldn't this result in the known out-of-order problem on SMP systems?
Best regards,
Alexander
Hi Alexander,
Quoted 6 lines · Alexander Stein
Review
Quoted 2 lines · Oliver Hartkopp
(quoted diff) +2
-0+ + netif_rx_ni(skb);^^^^^^^^^^Wouldn't this result in the known out-of-order problem on SMP systems?
I don't think so.
We're not dealing with irqs here that could reach different CPUs. This is all in softirq/userspace context.
It's the same thing we have with the virtual CAN vcan.
When there's something to fix, then we could fix up both vcan and vxcan. But I don't think so.
Regards,
Oliver
Hi Oliver,
Quoted 7 lines · Oliver Hartkopp
Quoted 6 lines · Alexander Stein
Review
Quoted 3 lines · Oliver Hartkopp
(quoted diff) +2
-0+ + netif_rx_ni(skb);^^^^^^^^^^
Quoted 6 lines
Quoted 1 line
Wouldn't this result in the known out-of-order problem on SMP systems?
I don't think so.
We're not dealing with irqs here that could reach different CPUs. This is all in softirq/userspace context.
It's the same thing we have with the virtual CAN vcan.
That would be great. I don't know the complete code path but is it possible that this xmit function is executed on different CPUs at the same time? e.g. multiple sockets/threads writing a CAN frame.
Best regards,
Alexander