nuttx/net/devif/ipv6_input.c

656 lines
19 KiB
C
Raw Normal View History

/****************************************************************************
* net/devif/ipv6_input.c
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership. The
* ASF licenses this file to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance with the
* License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
* License for the specific language governing permissions and limitations
* under the License.
*
****************************************************************************/
/****************************************************************************
* Included Files
****************************************************************************/
#include <nuttx/config.h>
#ifdef CONFIG_NET_IPv6
#include <sys/ioctl.h>
#include <stdint.h>
#include <stdbool.h>
#include <debug.h>
#include <string.h>
2015-01-24 14:29:43 +01:00
#include <net/if.h>
#include <nuttx/net/netconfig.h>
#include <nuttx/net/netdev.h>
#include <nuttx/net/netstats.h>
#include <nuttx/net/ip.h>
#include <nuttx/net/ipv6ext.h>
#include "neighbor/neighbor.h"
2014-06-25 02:12:49 +02:00
#include "tcp/tcp.h"
#include "udp/udp.h"
#include "sixlowpan/sixlowpan.h"
2014-06-25 18:34:52 +02:00
#include "pkt/pkt.h"
#include "icmpv6/icmpv6.h"
#include "nat/nat.h"
#include "netdev/netdev.h"
#include "ipforward/ipforward.h"
#include "inet/inet.h"
#include "devif/devif.h"
#include "ipfilter/ipfilter.h"
#include "ipfrag/ipfrag.h"
/****************************************************************************
* Private Functions
****************************************************************************/
/****************************************************************************
* Name: check_dev_destipaddr
*
* Description:
* Check if the destination address in the IPv6 is destined for the
* provided network device.
*
* Returned Value:
* 1 - This packet is destined for this network device
* 0 - This packet is NOT destined for this network device
*
****************************************************************************/
static int check_dev_destipaddr(FAR struct net_driver_s *dev, FAR void *arg)
{
FAR struct ipv6_hdr_s *ipv6 = (FAR struct ipv6_hdr_s *)arg;
/* Check if the IPv6 destination address matches the IPv6 address assigned
* to this device.
*/
if (NETDEV_IS_MY_V6ADDR(dev, ipv6->destipaddr))
{
return 1;
}
/* No match, return 0 to keep searching */
return 0;
}
/****************************************************************************
* Name: check_destipaddr
*
* Description:
* Check if the destination address in the IPv6 is destined for us. This
* is typically just a comparison the of the IPv6 destination address in
* the IPv6 packet with the IPv6 address assigned to the receiving device.
*
* Input Parameters:
* dev - The device on which the packet was received and which contains
* the IPv6 packet.
* ipv6 - A convenience pointer to the IPv6 header in within the IPv6
* packet
*
* Returned Value:
* true - This packet is destined for us
* false - This packet is NOT destined for us and may need to be forwarded.
*
****************************************************************************/
static bool check_destipaddr(FAR struct net_driver_s *dev,
FAR struct ipv6_hdr_s *ipv6)
{
int ret;
/* For IPv6, packet reception is a little trickier as we need to make sure
* that we listen to certain multicast addresses (all hosts multicast
* address, and the solicited-node multicast address) as well. However,
* we will cheat here and accept all multicast packets that are sent to
* the ff02::/16 addresses.
*/
if (ipv6->destipaddr[0] == HTONS(0xff02))
{
#ifdef CONFIG_NET_IPFORWARD_BROADCAST
/* Forward multicast packets */
ipv6_forward_broadcast(dev, ipv6);
#endif
return true;
}
/* We will also allow for a perverse case where we receive a packet
* addressed to us, but on a different device. Can that really happen?
*/
ret = netdev_foreach(check_dev_destipaddr, ipv6);
if (ret == 1)
{
/* The traversal of the network devices will return 0 if there is
* no network device with that address or 1 if there is a network
* device with such an address.
*/
return true;
}
return false;
}
/****************************************************************************
net/l2/l3/l4: add support of iob offload 1. Add new config CONFIG_NET_LL_GUARDSIZE to isolation of l2 stack, which will benefit l3(IP) layer for multi-MAC(l2) implementation, especially in some NICs such as celluler net driver. new configuration options: CONFIG_NET_LL_GUARDSIZE CONFIG_NET_LL_GUARDSIZE will reserved l2 buffer header size of network buffer to isolate the L2/L3 (MAC/IP) data on network layer, which will be beneficial to L3 network layer protocol transparent transmission and forwarding ------------------------------------------------------------ Layout of frist iob entry: iob_data (aligned by CONFIG_IOB_ALIGNMENT) | | io_offset(CONFIG_NET_LL_GUARDSIZE) | | ------------------------------------------------- iob | Reserved | io_len | ------------------------------------------------- ------------------------------------------------------------- Layout of different NICs implementation: iob_data (aligned by CONFIG_IOB_ALIGNMENT) | | io_offset(CONFIG_NET_LL_GUARDSIZE) | | ------------------------------------------------- Ethernet | Reserved | ETH_HDRLEN | io_len | ---------------------------------|--------------- 8021Q | Reserved | ETH_8021Q_HDRLEN | io_len | ---------------------------------|--------------- ipforward | Reserved | io_len | ------------------------------------------------- -------------------------------------------------------------------- 2. Support iob offload to l2 driver to avoid unnecessary memory copy Support send/receive iob vectors directly between the NICs and l3/l4 stack to avoid unnecessary memory copies, especially on hardware that supports Scatter/gather, which can greatly improve performance. new interface to support iob offload: ------------------------------------------ | IOB version | original | |----------------------------------------| | devif_iob_poll() | devif_poll() | | ... | ... | ------------------------------------------ -------------------------------------------------------------------- 1> NIC hardware support Scatter/gather transfer TX: tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll("NIC"_txpoll) callback() // "NIC"_txpoll | dev->d_iob: | --------------- --------------- io_data iob1 | | | iob3 | | | \ --------------- --------------- --------------- | --------------- | iob0 | | | | iob2 | | | | --------------- | --------------- | \ | / / \ | / / ---------------------------------------------- NICs io vector | | | | | | | | | | ---------------------------------------------- RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | pkt/ipv[4/6]_input()/... | | NICs io vector receive(iov_base to each iobs) -------------------------------------------------------------------- 2> CONFIG_IOB_BUFSIZE is greater than MTU: TX: "(CONFIG_IOB_BUFSIZE) > (MAX_NETDEV_PKTSIZE + CONFIG_NET_GUARDSIZE + CONFIG_NET_LL_GUARDSIZE)" tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll("NIC"_txpoll) callback() // "NIC"_txpoll | "NIC"_send() (dev->d_iob->io_data[CONFIG_NET_LL_GUARDSIZE - NET_LL_HDRLEN(dev)]) RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | pkt/ipv[4/6]_input()/... | | NICs io vector receive(iov_base to io_data) -------------------------------------------------------------------- 3> Compatible with all old flat buffer NICs TX: tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll(devif_poll_callback()) devif_poll_callback() /* new interface, gather iobs to flat buffer */ / \ / \ devif_poll("NIC"_txpoll) "NIC"_send()(dev->d_buf) RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | netdev_input() /* new interface, Scatter/gather flat/iob buffer */ | | pkt/ipv[4|6]_input()/... | | NICs io vector receive(Orignal flat buffer) 3. Iperf passthrough on NuttX simulator: ------------------------------------------------- | Protocol | Server | Client | | |-----------------------------------------------| | TCP | 813 | 834 | Mbits/sec | | TCP(Offload) | 1720 | 1100 | Mbits/sec | | UDP | 22 | 757 | Mbits/sec | | UDP(Offload) | 25 | 1250 | Mbits/sec | ------------------------------------------------- Signed-off-by: chao an <anchao@xiaomi.com>
2022-11-23 12:39:40 +01:00
* Name: ipv6_in
*
* Description:
* Receive an IPv6 packet from the network device. Verify and forward to
* L3 packet handling logic if the packet is destined for us.
*
net/l2/l3/l4: add support of iob offload 1. Add new config CONFIG_NET_LL_GUARDSIZE to isolation of l2 stack, which will benefit l3(IP) layer for multi-MAC(l2) implementation, especially in some NICs such as celluler net driver. new configuration options: CONFIG_NET_LL_GUARDSIZE CONFIG_NET_LL_GUARDSIZE will reserved l2 buffer header size of network buffer to isolate the L2/L3 (MAC/IP) data on network layer, which will be beneficial to L3 network layer protocol transparent transmission and forwarding ------------------------------------------------------------ Layout of frist iob entry: iob_data (aligned by CONFIG_IOB_ALIGNMENT) | | io_offset(CONFIG_NET_LL_GUARDSIZE) | | ------------------------------------------------- iob | Reserved | io_len | ------------------------------------------------- ------------------------------------------------------------- Layout of different NICs implementation: iob_data (aligned by CONFIG_IOB_ALIGNMENT) | | io_offset(CONFIG_NET_LL_GUARDSIZE) | | ------------------------------------------------- Ethernet | Reserved | ETH_HDRLEN | io_len | ---------------------------------|--------------- 8021Q | Reserved | ETH_8021Q_HDRLEN | io_len | ---------------------------------|--------------- ipforward | Reserved | io_len | ------------------------------------------------- -------------------------------------------------------------------- 2. Support iob offload to l2 driver to avoid unnecessary memory copy Support send/receive iob vectors directly between the NICs and l3/l4 stack to avoid unnecessary memory copies, especially on hardware that supports Scatter/gather, which can greatly improve performance. new interface to support iob offload: ------------------------------------------ | IOB version | original | |----------------------------------------| | devif_iob_poll() | devif_poll() | | ... | ... | ------------------------------------------ -------------------------------------------------------------------- 1> NIC hardware support Scatter/gather transfer TX: tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll("NIC"_txpoll) callback() // "NIC"_txpoll | dev->d_iob: | --------------- --------------- io_data iob1 | | | iob3 | | | \ --------------- --------------- --------------- | --------------- | iob0 | | | | iob2 | | | | --------------- | --------------- | \ | / / \ | / / ---------------------------------------------- NICs io vector | | | | | | | | | | ---------------------------------------------- RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | pkt/ipv[4/6]_input()/... | | NICs io vector receive(iov_base to each iobs) -------------------------------------------------------------------- 2> CONFIG_IOB_BUFSIZE is greater than MTU: TX: "(CONFIG_IOB_BUFSIZE) > (MAX_NETDEV_PKTSIZE + CONFIG_NET_GUARDSIZE + CONFIG_NET_LL_GUARDSIZE)" tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll("NIC"_txpoll) callback() // "NIC"_txpoll | "NIC"_send() (dev->d_iob->io_data[CONFIG_NET_LL_GUARDSIZE - NET_LL_HDRLEN(dev)]) RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | pkt/ipv[4/6]_input()/... | | NICs io vector receive(iov_base to io_data) -------------------------------------------------------------------- 3> Compatible with all old flat buffer NICs TX: tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll(devif_poll_callback()) devif_poll_callback() /* new interface, gather iobs to flat buffer */ / \ / \ devif_poll("NIC"_txpoll) "NIC"_send()(dev->d_buf) RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | netdev_input() /* new interface, Scatter/gather flat/iob buffer */ | | pkt/ipv[4|6]_input()/... | | NICs io vector receive(Orignal flat buffer) 3. Iperf passthrough on NuttX simulator: ------------------------------------------------- | Protocol | Server | Client | | |-----------------------------------------------| | TCP | 813 | 834 | Mbits/sec | | TCP(Offload) | 1720 | 1100 | Mbits/sec | | UDP | 22 | 757 | Mbits/sec | | UDP(Offload) | 25 | 1250 | Mbits/sec | ------------------------------------------------- Signed-off-by: chao an <anchao@xiaomi.com>
2022-11-23 12:39:40 +01:00
* This is the iob buffer version of ipv6_input(),
* this function will support send/receive iob vectors directly between
* the driver and l3/l4 stack to avoid unnecessary memory copies,
* especially on hardware that supports Scatter/gather, which can
* greatly improve performance
* this function will uses d_iob as packets input which used by some
* NICs such as celluler net driver.
*
* Input Parameters:
* dev - The device on which the packet was received and which contains
* the IPv6 packet.
* Returned Value:
* OK - The packet was processed (or dropped) and can be discarded.
2017-07-17 15:14:12 +02:00
* ERROR - Hold the packet and try again later. There is a listening
* socket but no receive in place to catch the packet yet. The
* device's d_len will be set to zero in this case as there is
* no outgoing data.
*
* If this function returns to the network driver with dev->d_len > 0,
* that is an indication to the driver that there is an outgoing response
* to this input.
*
* Assumptions:
* The network is locked.
*
****************************************************************************/
net/l2/l3/l4: add support of iob offload 1. Add new config CONFIG_NET_LL_GUARDSIZE to isolation of l2 stack, which will benefit l3(IP) layer for multi-MAC(l2) implementation, especially in some NICs such as celluler net driver. new configuration options: CONFIG_NET_LL_GUARDSIZE CONFIG_NET_LL_GUARDSIZE will reserved l2 buffer header size of network buffer to isolate the L2/L3 (MAC/IP) data on network layer, which will be beneficial to L3 network layer protocol transparent transmission and forwarding ------------------------------------------------------------ Layout of frist iob entry: iob_data (aligned by CONFIG_IOB_ALIGNMENT) | | io_offset(CONFIG_NET_LL_GUARDSIZE) | | ------------------------------------------------- iob | Reserved | io_len | ------------------------------------------------- ------------------------------------------------------------- Layout of different NICs implementation: iob_data (aligned by CONFIG_IOB_ALIGNMENT) | | io_offset(CONFIG_NET_LL_GUARDSIZE) | | ------------------------------------------------- Ethernet | Reserved | ETH_HDRLEN | io_len | ---------------------------------|--------------- 8021Q | Reserved | ETH_8021Q_HDRLEN | io_len | ---------------------------------|--------------- ipforward | Reserved | io_len | ------------------------------------------------- -------------------------------------------------------------------- 2. Support iob offload to l2 driver to avoid unnecessary memory copy Support send/receive iob vectors directly between the NICs and l3/l4 stack to avoid unnecessary memory copies, especially on hardware that supports Scatter/gather, which can greatly improve performance. new interface to support iob offload: ------------------------------------------ | IOB version | original | |----------------------------------------| | devif_iob_poll() | devif_poll() | | ... | ... | ------------------------------------------ -------------------------------------------------------------------- 1> NIC hardware support Scatter/gather transfer TX: tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll("NIC"_txpoll) callback() // "NIC"_txpoll | dev->d_iob: | --------------- --------------- io_data iob1 | | | iob3 | | | \ --------------- --------------- --------------- | --------------- | iob0 | | | | iob2 | | | | --------------- | --------------- | \ | / / \ | / / ---------------------------------------------- NICs io vector | | | | | | | | | | ---------------------------------------------- RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | pkt/ipv[4/6]_input()/... | | NICs io vector receive(iov_base to each iobs) -------------------------------------------------------------------- 2> CONFIG_IOB_BUFSIZE is greater than MTU: TX: "(CONFIG_IOB_BUFSIZE) > (MAX_NETDEV_PKTSIZE + CONFIG_NET_GUARDSIZE + CONFIG_NET_LL_GUARDSIZE)" tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll("NIC"_txpoll) callback() // "NIC"_txpoll | "NIC"_send() (dev->d_iob->io_data[CONFIG_NET_LL_GUARDSIZE - NET_LL_HDRLEN(dev)]) RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | pkt/ipv[4/6]_input()/... | | NICs io vector receive(iov_base to io_data) -------------------------------------------------------------------- 3> Compatible with all old flat buffer NICs TX: tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll(devif_poll_callback()) devif_poll_callback() /* new interface, gather iobs to flat buffer */ / \ / \ devif_poll("NIC"_txpoll) "NIC"_send()(dev->d_buf) RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | netdev_input() /* new interface, Scatter/gather flat/iob buffer */ | | pkt/ipv[4|6]_input()/... | | NICs io vector receive(Orignal flat buffer) 3. Iperf passthrough on NuttX simulator: ------------------------------------------------- | Protocol | Server | Client | | |-----------------------------------------------| | TCP | 813 | 834 | Mbits/sec | | TCP(Offload) | 1720 | 1100 | Mbits/sec | | UDP | 22 | 757 | Mbits/sec | | UDP(Offload) | 25 | 1250 | Mbits/sec | ------------------------------------------------- Signed-off-by: chao an <anchao@xiaomi.com>
2022-11-23 12:39:40 +01:00
static int ipv6_in(FAR struct net_driver_s *dev)
{
FAR struct ipv6_hdr_s *ipv6 = IPv6BUF;
FAR uint8_t *payload;
uint16_t iphdrlen;
uint16_t paylen;
uint8_t nxthdr;
#ifdef CONFIG_NET_IPFORWARD
int ret;
#endif
#ifdef CONFIG_NET_IPFRAG
bool isfrag = false;
#endif
/* This is where the input processing starts. */
#ifdef CONFIG_NET_STATISTICS
g_netstats.ipv6.recv++;
#endif
/* Start of IP input header processing code.
*
* Check validity of the IP header.
*/
if ((ipv6->vtc & 0xf0) != 0x60)
{
/* IP version and header length. */
nwarn("WARNING: Invalid IPv6 version: %d\n", ipv6->vtc >> 4);
#ifdef CONFIG_NET_STATISTICS
g_netstats.ipv6.vhlerr++;
#endif
goto drop;
}
/* Get the size of the packet minus the size of link layer header */
net/l2/l3/l4: add support of iob offload 1. Add new config CONFIG_NET_LL_GUARDSIZE to isolation of l2 stack, which will benefit l3(IP) layer for multi-MAC(l2) implementation, especially in some NICs such as celluler net driver. new configuration options: CONFIG_NET_LL_GUARDSIZE CONFIG_NET_LL_GUARDSIZE will reserved l2 buffer header size of network buffer to isolate the L2/L3 (MAC/IP) data on network layer, which will be beneficial to L3 network layer protocol transparent transmission and forwarding ------------------------------------------------------------ Layout of frist iob entry: iob_data (aligned by CONFIG_IOB_ALIGNMENT) | | io_offset(CONFIG_NET_LL_GUARDSIZE) | | ------------------------------------------------- iob | Reserved | io_len | ------------------------------------------------- ------------------------------------------------------------- Layout of different NICs implementation: iob_data (aligned by CONFIG_IOB_ALIGNMENT) | | io_offset(CONFIG_NET_LL_GUARDSIZE) | | ------------------------------------------------- Ethernet | Reserved | ETH_HDRLEN | io_len | ---------------------------------|--------------- 8021Q | Reserved | ETH_8021Q_HDRLEN | io_len | ---------------------------------|--------------- ipforward | Reserved | io_len | ------------------------------------------------- -------------------------------------------------------------------- 2. Support iob offload to l2 driver to avoid unnecessary memory copy Support send/receive iob vectors directly between the NICs and l3/l4 stack to avoid unnecessary memory copies, especially on hardware that supports Scatter/gather, which can greatly improve performance. new interface to support iob offload: ------------------------------------------ | IOB version | original | |----------------------------------------| | devif_iob_poll() | devif_poll() | | ... | ... | ------------------------------------------ -------------------------------------------------------------------- 1> NIC hardware support Scatter/gather transfer TX: tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll("NIC"_txpoll) callback() // "NIC"_txpoll | dev->d_iob: | --------------- --------------- io_data iob1 | | | iob3 | | | \ --------------- --------------- --------------- | --------------- | iob0 | | | | iob2 | | | | --------------- | --------------- | \ | / / \ | / / ---------------------------------------------- NICs io vector | | | | | | | | | | ---------------------------------------------- RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | pkt/ipv[4/6]_input()/... | | NICs io vector receive(iov_base to each iobs) -------------------------------------------------------------------- 2> CONFIG_IOB_BUFSIZE is greater than MTU: TX: "(CONFIG_IOB_BUFSIZE) > (MAX_NETDEV_PKTSIZE + CONFIG_NET_GUARDSIZE + CONFIG_NET_LL_GUARDSIZE)" tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll("NIC"_txpoll) callback() // "NIC"_txpoll | "NIC"_send() (dev->d_iob->io_data[CONFIG_NET_LL_GUARDSIZE - NET_LL_HDRLEN(dev)]) RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | pkt/ipv[4/6]_input()/... | | NICs io vector receive(iov_base to io_data) -------------------------------------------------------------------- 3> Compatible with all old flat buffer NICs TX: tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll(devif_poll_callback()) devif_poll_callback() /* new interface, gather iobs to flat buffer */ / \ / \ devif_poll("NIC"_txpoll) "NIC"_send()(dev->d_buf) RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | netdev_input() /* new interface, Scatter/gather flat/iob buffer */ | | pkt/ipv[4|6]_input()/... | | NICs io vector receive(Orignal flat buffer) 3. Iperf passthrough on NuttX simulator: ------------------------------------------------- | Protocol | Server | Client | | |-----------------------------------------------| | TCP | 813 | 834 | Mbits/sec | | TCP(Offload) | 1720 | 1100 | Mbits/sec | | UDP | 22 | 757 | Mbits/sec | | UDP(Offload) | 25 | 1250 | Mbits/sec | ------------------------------------------------- Signed-off-by: chao an <anchao@xiaomi.com>
2022-11-23 12:39:40 +01:00
if (IPv6_HDRLEN > dev->d_len)
{
nwarn("WARNING: Packet shorter than IPv6 header\n");
goto drop;
}
/* Make sure that all packet processing logic knows that there is an IPv6
* packet in the device buffer.
*/
IFF_SET_IPv6(dev->d_flags);
/* Check the size of the packet. If the size reported to us in d_len is
* smaller the size reported in the IP header, we assume that the packet
* has been corrupted in transit. If the size of d_len is larger than the
* size reported in the IP packet header, the packet has been padded and
* we set d_len to the correct value.
*
* The length reported in the IPv6 header is the length of the payload
2017-01-18 14:53:28 +01:00
* that follows the header. The device interface uses the d_len variable
* for holding the size of the entire packet, including the IP header but
* without the link layer header (subtracted out above).
*
* NOTE: The payload length in the includes the size of the Ipv6 extension
* options, but not the size of the IPv6 header.
*
* REVISIT: Length will be set to zero if the extension header carries
* a Jumbo payload option.
*/
paylen = ((uint16_t)ipv6->len[0] << 8) + (uint16_t)ipv6->len[1] +
IPv6_HDRLEN;
net/l2/l3/l4: add support of iob offload 1. Add new config CONFIG_NET_LL_GUARDSIZE to isolation of l2 stack, which will benefit l3(IP) layer for multi-MAC(l2) implementation, especially in some NICs such as celluler net driver. new configuration options: CONFIG_NET_LL_GUARDSIZE CONFIG_NET_LL_GUARDSIZE will reserved l2 buffer header size of network buffer to isolate the L2/L3 (MAC/IP) data on network layer, which will be beneficial to L3 network layer protocol transparent transmission and forwarding ------------------------------------------------------------ Layout of frist iob entry: iob_data (aligned by CONFIG_IOB_ALIGNMENT) | | io_offset(CONFIG_NET_LL_GUARDSIZE) | | ------------------------------------------------- iob | Reserved | io_len | ------------------------------------------------- ------------------------------------------------------------- Layout of different NICs implementation: iob_data (aligned by CONFIG_IOB_ALIGNMENT) | | io_offset(CONFIG_NET_LL_GUARDSIZE) | | ------------------------------------------------- Ethernet | Reserved | ETH_HDRLEN | io_len | ---------------------------------|--------------- 8021Q | Reserved | ETH_8021Q_HDRLEN | io_len | ---------------------------------|--------------- ipforward | Reserved | io_len | ------------------------------------------------- -------------------------------------------------------------------- 2. Support iob offload to l2 driver to avoid unnecessary memory copy Support send/receive iob vectors directly between the NICs and l3/l4 stack to avoid unnecessary memory copies, especially on hardware that supports Scatter/gather, which can greatly improve performance. new interface to support iob offload: ------------------------------------------ | IOB version | original | |----------------------------------------| | devif_iob_poll() | devif_poll() | | ... | ... | ------------------------------------------ -------------------------------------------------------------------- 1> NIC hardware support Scatter/gather transfer TX: tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll("NIC"_txpoll) callback() // "NIC"_txpoll | dev->d_iob: | --------------- --------------- io_data iob1 | | | iob3 | | | \ --------------- --------------- --------------- | --------------- | iob0 | | | | iob2 | | | | --------------- | --------------- | \ | / / \ | / / ---------------------------------------------- NICs io vector | | | | | | | | | | ---------------------------------------------- RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | pkt/ipv[4/6]_input()/... | | NICs io vector receive(iov_base to each iobs) -------------------------------------------------------------------- 2> CONFIG_IOB_BUFSIZE is greater than MTU: TX: "(CONFIG_IOB_BUFSIZE) > (MAX_NETDEV_PKTSIZE + CONFIG_NET_GUARDSIZE + CONFIG_NET_LL_GUARDSIZE)" tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll("NIC"_txpoll) callback() // "NIC"_txpoll | "NIC"_send() (dev->d_iob->io_data[CONFIG_NET_LL_GUARDSIZE - NET_LL_HDRLEN(dev)]) RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | pkt/ipv[4/6]_input()/... | | NICs io vector receive(iov_base to io_data) -------------------------------------------------------------------- 3> Compatible with all old flat buffer NICs TX: tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll(devif_poll_callback()) devif_poll_callback() /* new interface, gather iobs to flat buffer */ / \ / \ devif_poll("NIC"_txpoll) "NIC"_send()(dev->d_buf) RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | netdev_input() /* new interface, Scatter/gather flat/iob buffer */ | | pkt/ipv[4|6]_input()/... | | NICs io vector receive(Orignal flat buffer) 3. Iperf passthrough on NuttX simulator: ------------------------------------------------- | Protocol | Server | Client | | |-----------------------------------------------| | TCP | 813 | 834 | Mbits/sec | | TCP(Offload) | 1720 | 1100 | Mbits/sec | | UDP | 22 | 757 | Mbits/sec | | UDP(Offload) | 25 | 1250 | Mbits/sec | ------------------------------------------------- Signed-off-by: chao an <anchao@xiaomi.com>
2022-11-23 12:39:40 +01:00
if (paylen < dev->d_len)
{
iob_update_pktlen(dev->d_iob, paylen, false);
dev->d_len = paylen;
}
net/l2/l3/l4: add support of iob offload 1. Add new config CONFIG_NET_LL_GUARDSIZE to isolation of l2 stack, which will benefit l3(IP) layer for multi-MAC(l2) implementation, especially in some NICs such as celluler net driver. new configuration options: CONFIG_NET_LL_GUARDSIZE CONFIG_NET_LL_GUARDSIZE will reserved l2 buffer header size of network buffer to isolate the L2/L3 (MAC/IP) data on network layer, which will be beneficial to L3 network layer protocol transparent transmission and forwarding ------------------------------------------------------------ Layout of frist iob entry: iob_data (aligned by CONFIG_IOB_ALIGNMENT) | | io_offset(CONFIG_NET_LL_GUARDSIZE) | | ------------------------------------------------- iob | Reserved | io_len | ------------------------------------------------- ------------------------------------------------------------- Layout of different NICs implementation: iob_data (aligned by CONFIG_IOB_ALIGNMENT) | | io_offset(CONFIG_NET_LL_GUARDSIZE) | | ------------------------------------------------- Ethernet | Reserved | ETH_HDRLEN | io_len | ---------------------------------|--------------- 8021Q | Reserved | ETH_8021Q_HDRLEN | io_len | ---------------------------------|--------------- ipforward | Reserved | io_len | ------------------------------------------------- -------------------------------------------------------------------- 2. Support iob offload to l2 driver to avoid unnecessary memory copy Support send/receive iob vectors directly between the NICs and l3/l4 stack to avoid unnecessary memory copies, especially on hardware that supports Scatter/gather, which can greatly improve performance. new interface to support iob offload: ------------------------------------------ | IOB version | original | |----------------------------------------| | devif_iob_poll() | devif_poll() | | ... | ... | ------------------------------------------ -------------------------------------------------------------------- 1> NIC hardware support Scatter/gather transfer TX: tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll("NIC"_txpoll) callback() // "NIC"_txpoll | dev->d_iob: | --------------- --------------- io_data iob1 | | | iob3 | | | \ --------------- --------------- --------------- | --------------- | iob0 | | | | iob2 | | | | --------------- | --------------- | \ | / / \ | / / ---------------------------------------------- NICs io vector | | | | | | | | | | ---------------------------------------------- RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | pkt/ipv[4/6]_input()/... | | NICs io vector receive(iov_base to each iobs) -------------------------------------------------------------------- 2> CONFIG_IOB_BUFSIZE is greater than MTU: TX: "(CONFIG_IOB_BUFSIZE) > (MAX_NETDEV_PKTSIZE + CONFIG_NET_GUARDSIZE + CONFIG_NET_LL_GUARDSIZE)" tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll("NIC"_txpoll) callback() // "NIC"_txpoll | "NIC"_send() (dev->d_iob->io_data[CONFIG_NET_LL_GUARDSIZE - NET_LL_HDRLEN(dev)]) RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | pkt/ipv[4/6]_input()/... | | NICs io vector receive(iov_base to io_data) -------------------------------------------------------------------- 3> Compatible with all old flat buffer NICs TX: tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll(devif_poll_callback()) devif_poll_callback() /* new interface, gather iobs to flat buffer */ / \ / \ devif_poll("NIC"_txpoll) "NIC"_send()(dev->d_buf) RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | netdev_input() /* new interface, Scatter/gather flat/iob buffer */ | | pkt/ipv[4|6]_input()/... | | NICs io vector receive(Orignal flat buffer) 3. Iperf passthrough on NuttX simulator: ------------------------------------------------- | Protocol | Server | Client | | |-----------------------------------------------| | TCP | 813 | 834 | Mbits/sec | | TCP(Offload) | 1720 | 1100 | Mbits/sec | | UDP | 22 | 757 | Mbits/sec | | UDP(Offload) | 25 | 1250 | Mbits/sec | ------------------------------------------------- Signed-off-by: chao an <anchao@xiaomi.com>
2022-11-23 12:39:40 +01:00
else if (paylen > dev->d_len)
{
nwarn("WARNING: IP packet shorter than length in IP header\n");
goto drop;
}
/* Parse IPv6 extension headers (parsed but ignored) */
payload = IPBUF(IPv6_HDRLEN); /* Assume payload starts right after IPv6 header */
iphdrlen = IPv6_HDRLEN; /* Total length of the IPv6 header */
nxthdr = ipv6->proto; /* Next header determined by IPv6 header prototype */
while (ipv6_exthdr(nxthdr))
{
FAR struct ipv6_extension_s *exthdr;
uint16_t extlen;
/* Just skip over the extension header */
exthdr = (FAR struct ipv6_extension_s *)payload;
if (nxthdr == NEXT_FRAGMENT_EH)
{
extlen = EXTHDR_FRAG_LEN;
#ifdef CONFIG_NET_IPFRAG
isfrag = true;
#endif
}
else
{
extlen = EXTHDR_LEN((unsigned int)exthdr->len);
}
payload += extlen;
iphdrlen += extlen;
nxthdr = exthdr->nxthdr;
}
#ifdef CONFIG_NET_NAT66
/* Try NAT inbound, rule matching will be performed in NAT module. */
ipv6_nat_inbound(dev, ipv6);
#endif
#ifdef CONFIG_NET_BROADCAST
/* Check for a multicast packet, which may be destined to us (even if
* there is no IP address yet assigned to the device). We only expect
* multicast packets destined for sockets that have joined a multicast
* group or for ICMPv6 Autoconfiguration and Neighbor discovery or ICMPv6
* MLD packets.
*
* We should actually pick off certain multicast address (all hosts
* multicast address, and the solicited-node multicast address). We
* will cheat here and accept all multicast packets that are sent to the
* ff00::/8 addresses (see net_is_addr_mcast).
*/
if (net_is_addr_mcast(ipv6->destipaddr))
{
#ifdef CONFIG_NET_IPFORWARD_BROADCAST
/* Packets sent to ffx0 are reserved, ffx1 are interface-local, and
* ffx2 are interface-local, and therefore, should not be forwarded
*/
if (((ipv6->destipaddr[0] & HTONS(0xff0f)) != HTONS(0xff00)) &&
((ipv6->destipaddr[0] & HTONS(0xff0f)) != HTONS(0xff01)) &&
((ipv6->destipaddr[0] & HTONS(0xff0f)) != HTONS(0xff02)))
{
/* Forward broadcast packets */
ipv6_forward_broadcast(dev, ipv6);
}
#endif
/* Fall through with no further address checks and handle the multicast
* address by its IPv6 nexthdr field.
*/
}
else
#endif
{
/* Check if the packet is destined for us. */
if (!check_destipaddr(dev, ipv6))
{
#ifdef CONFIG_NET_IPFORWARD
/* Not destined for us, try to forward the packet */
ret = ipv6_forward(dev, ipv6);
if (ret >= 0)
{
/* The packet was forwarded. Return success; d_len will
* be set appropriately by the forwarding logic: Cleared
* if the packet is forward via another device or non-
* zero if it will be forwarded by the same device that
* it was received on.
*/
goto done;
}
else
#endif
#if defined(NET_UDP_HAVE_STACK) && defined(CONFIG_NET_BINDTODEVICE)
/* If the protocol specific socket option NET_BINDTODEVICE
* is selected, then we must forward all UDP packets to the bound
* socket.
*/
if (nxthdr != IP_PROTO_UDP)
#endif
{
/* Not destined for us and not forwardable...
* drop the packet.
*/
ninfo("WARNING: Not destined for us... Dropping!\n");
goto drop;
}
}
}
#ifdef CONFIG_NET_IPFORWARD
/* Return success if the packet was forwarded. */
if (dev->d_len == 0)
{
goto done;
}
#endif
#ifdef CONFIG_NET_ICMPv6
/* In other cases, the device must be assigned a non-zero IP address
* (the all zero address is the "unspecified" address.
*/
if (!NETDEV_HAS_V6ADDR(dev))
{
nwarn("WARNING: No IP address assigned\n");
goto drop;
}
#endif
#ifdef CONFIG_NET_IPFRAG
if (isfrag)
{
if (ipv6_fragin(dev) == OK)
{
return OK;
}
else
{
#ifdef CONFIG_NET_STATISTICS
g_netstats.ipv6.fragerr++;
#endif
goto drop;
}
}
#endif
#ifdef CONFIG_NET_IPFILTER
if (ipv6_filter_in(dev) != IPFILTER_TARGET_ACCEPT)
{
ninfo("Drop/Reject INPUT packet due to filter.\n");
goto done;
}
#endif
/* Now process the incoming packet according to the protocol specified in
* the next header IPv6 field.
*/
2015-01-24 14:29:43 +01:00
switch (nxthdr)
{
#ifdef NET_TCP_HAVE_STACK
2014-07-05 00:38:51 +02:00
case IP_PROTO_TCP: /* TCP input */
2019-10-25 19:31:42 +02:00
2017-04-02 19:15:46 +02:00
/* Forward the IPv6 TCP packet */
tcp_ipv6_input(dev, iphdrlen);
2017-04-02 19:15:46 +02:00
#ifdef CONFIG_NET_6LOWPAN
/* TCP output comes through three different mechanisms. Either from:
2017-04-02 19:15:46 +02:00
*
* 1. TCP socket output. For the case of TCP output to an
* IEEE802.15.4, the TCP output is caught in the socket
2017-06-19 00:00:08 +02:00
* send()/sendto() logic and and redirected to 6LoWPAN logic.
* 2. TCP output from the TCP state machine. That will occur
* during TCP packet processing by the TCP state machine.
* 3. TCP output resulting from TX or timer polling
*
* Case 3 is handled here. Logic here detects if (1) an attempt
* to return with d_len > 0 and (2) that the device is an
2017-06-19 00:00:08 +02:00
* IEEE802.15.4 MAC network driver. Under those conditions, 6LoWPAN
* logic will be called to create the IEEE80215.4 frames.
2017-04-02 19:15:46 +02:00
*/
if (dev->d_len > 0 && dev->d_lltype == CONFIG_NET_6LOWPAN)
{
2017-06-19 00:00:08 +02:00
/* Let 6LoWPAN handle the TCP output */
2017-04-02 19:15:46 +02:00
sixlowpan_tcp_send(dev, dev, ipv6);
2017-04-02 19:15:46 +02:00
/* Drop the packet in the d_buf */
goto drop;
}
2017-04-02 19:15:46 +02:00
#endif /* CONFIG_NET_6LOWPAN */
break;
#endif /* NET_TCP_HAVE_STACK */
#ifdef NET_UDP_HAVE_STACK
2014-07-05 00:38:51 +02:00
case IP_PROTO_UDP: /* UDP input */
2019-10-25 19:31:42 +02:00
2017-04-02 19:15:46 +02:00
/* Forward the IPv6 UDP packet */
udp_ipv6_input(dev, iphdrlen);
break;
#endif
/* Check for ICMP input */
#ifdef NET_ICMPv6_HAVE_STACK
2014-07-05 00:38:51 +02:00
case IP_PROTO_ICMP6: /* ICMP6 input */
2019-10-25 19:31:42 +02:00
2017-04-02 19:15:46 +02:00
/* Forward the ICMPv6 packet */
icmpv6_input(dev, iphdrlen);
#ifdef CONFIG_NET_6LOWPAN
/* All outgoing ICMPv6 messages come through one of two mechanisms:
*
* 1. The output from internal ICMPv6 message passing. These
* outgoing messages will use device polling and will be
* handled elsewhere.
* 2. ICMPv6 output resulting from TX or timer polling.
*
* Case 2 is handled here. Logic here detects if (1) an attempt
* to return with d_len > 0 and (2) that the device is an
* IEEE802.15.4 MAC network driver. Under those conditions, 6LoWPAN
* logic will be called to create the IEEE80215.4 frames.
*/
if (dev->d_len > 0 && dev->d_lltype == CONFIG_NET_6LOWPAN)
{
/* Let 6LoWPAN handle the ICMPv6 output */
sixlowpan_icmpv6_send(dev, dev, ipv6);
/* Drop the packet in the d_buf */
goto drop;
}
#endif /* CONFIG_NET_6LOWPAN */
break;
#endif /* NET_ICMPv6_HAVE_STACK */
default: /* Unrecognized/unsupported protocol */
nwarn("WARNING: Unrecognized IP protocol: %04x\n", ipv6->proto);
#ifdef CONFIG_NET_STATISTICS
g_netstats.ipv6.protoerr++;
#endif
goto drop;
}
#ifdef CONFIG_NET_IPFILTER
ipfilter_out(dev);
#endif
#if defined(CONFIG_NET_IPFORWARD) || defined(CONFIG_NET_IPFILTER)
done:
#endif
#ifdef CONFIG_NET_IPFRAG
ip_fragout(dev);
#endif
devif_out(dev);
/* Return and let the caller do any pending transmission. */
return OK;
/* Drop the packet. NOTE that OK is returned meaning that the
* packet has been processed (although processed unsuccessfully).
*/
drop:
#ifdef CONFIG_NET_STATISTICS
g_netstats.ipv6.drop++;
#endif
dev->d_len = 0;
return OK;
}
net/l2/l3/l4: add support of iob offload 1. Add new config CONFIG_NET_LL_GUARDSIZE to isolation of l2 stack, which will benefit l3(IP) layer for multi-MAC(l2) implementation, especially in some NICs such as celluler net driver. new configuration options: CONFIG_NET_LL_GUARDSIZE CONFIG_NET_LL_GUARDSIZE will reserved l2 buffer header size of network buffer to isolate the L2/L3 (MAC/IP) data on network layer, which will be beneficial to L3 network layer protocol transparent transmission and forwarding ------------------------------------------------------------ Layout of frist iob entry: iob_data (aligned by CONFIG_IOB_ALIGNMENT) | | io_offset(CONFIG_NET_LL_GUARDSIZE) | | ------------------------------------------------- iob | Reserved | io_len | ------------------------------------------------- ------------------------------------------------------------- Layout of different NICs implementation: iob_data (aligned by CONFIG_IOB_ALIGNMENT) | | io_offset(CONFIG_NET_LL_GUARDSIZE) | | ------------------------------------------------- Ethernet | Reserved | ETH_HDRLEN | io_len | ---------------------------------|--------------- 8021Q | Reserved | ETH_8021Q_HDRLEN | io_len | ---------------------------------|--------------- ipforward | Reserved | io_len | ------------------------------------------------- -------------------------------------------------------------------- 2. Support iob offload to l2 driver to avoid unnecessary memory copy Support send/receive iob vectors directly between the NICs and l3/l4 stack to avoid unnecessary memory copies, especially on hardware that supports Scatter/gather, which can greatly improve performance. new interface to support iob offload: ------------------------------------------ | IOB version | original | |----------------------------------------| | devif_iob_poll() | devif_poll() | | ... | ... | ------------------------------------------ -------------------------------------------------------------------- 1> NIC hardware support Scatter/gather transfer TX: tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll("NIC"_txpoll) callback() // "NIC"_txpoll | dev->d_iob: | --------------- --------------- io_data iob1 | | | iob3 | | | \ --------------- --------------- --------------- | --------------- | iob0 | | | | iob2 | | | | --------------- | --------------- | \ | / / \ | / / ---------------------------------------------- NICs io vector | | | | | | | | | | ---------------------------------------------- RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | pkt/ipv[4/6]_input()/... | | NICs io vector receive(iov_base to each iobs) -------------------------------------------------------------------- 2> CONFIG_IOB_BUFSIZE is greater than MTU: TX: "(CONFIG_IOB_BUFSIZE) > (MAX_NETDEV_PKTSIZE + CONFIG_NET_GUARDSIZE + CONFIG_NET_LL_GUARDSIZE)" tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll("NIC"_txpoll) callback() // "NIC"_txpoll | "NIC"_send() (dev->d_iob->io_data[CONFIG_NET_LL_GUARDSIZE - NET_LL_HDRLEN(dev)]) RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | pkt/ipv[4/6]_input()/... | | NICs io vector receive(iov_base to io_data) -------------------------------------------------------------------- 3> Compatible with all old flat buffer NICs TX: tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll(devif_poll_callback()) devif_poll_callback() /* new interface, gather iobs to flat buffer */ / \ / \ devif_poll("NIC"_txpoll) "NIC"_send()(dev->d_buf) RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | netdev_input() /* new interface, Scatter/gather flat/iob buffer */ | | pkt/ipv[4|6]_input()/... | | NICs io vector receive(Orignal flat buffer) 3. Iperf passthrough on NuttX simulator: ------------------------------------------------- | Protocol | Server | Client | | |-----------------------------------------------| | TCP | 813 | 834 | Mbits/sec | | TCP(Offload) | 1720 | 1100 | Mbits/sec | | UDP | 22 | 757 | Mbits/sec | | UDP(Offload) | 25 | 1250 | Mbits/sec | ------------------------------------------------- Signed-off-by: chao an <anchao@xiaomi.com>
2022-11-23 12:39:40 +01:00
/****************************************************************************
* Public Functions
****************************************************************************/
/****************************************************************************
* Name: ipv6_exthdr
*
* Description:
* Check whether it is an IPv6 extension header.
*
* Input Parameters:
* The next header value extracted from an IPv6 frame.
*
* Returned Value:
* Return true if the next header value is an IPv6 extension header.
*
****************************************************************************/
bool ipv6_exthdr(uint8_t nxthdr)
{
switch (nxthdr)
{
case NEXT_HOPBYBOT_EH: /* Hop-by-Hop Options Header */
case NEXT_ENCAP_EH: /* Encapsulated IPv6 Header */
case NEXT_ROUTING_EH: /* Routing Header */
case NEXT_FRAGMENT_EH: /* Fragment Header */
case NEXT_RRSVP_EH: /* Resource ReSerVation Protocol */
case NEXT_ENCAPSEC_EH: /* Encapsulating Security Payload */
case NEXT_AUTH_EH: /* Authentication Header */
case NEXT_DESTOPT_EH: /* Destination Options Header */
case NEXT_MOBILITY_EH: /* Mobility */
case NEXT_HOSTID_EH: /* Host Identity Protocol */
case NEXT_SHIM6_EH: /* Shim6 Protocol */
return true;
case NEXT_NOHEADER: /* No next header */
default:
return false;
}
}
net/l2/l3/l4: add support of iob offload 1. Add new config CONFIG_NET_LL_GUARDSIZE to isolation of l2 stack, which will benefit l3(IP) layer for multi-MAC(l2) implementation, especially in some NICs such as celluler net driver. new configuration options: CONFIG_NET_LL_GUARDSIZE CONFIG_NET_LL_GUARDSIZE will reserved l2 buffer header size of network buffer to isolate the L2/L3 (MAC/IP) data on network layer, which will be beneficial to L3 network layer protocol transparent transmission and forwarding ------------------------------------------------------------ Layout of frist iob entry: iob_data (aligned by CONFIG_IOB_ALIGNMENT) | | io_offset(CONFIG_NET_LL_GUARDSIZE) | | ------------------------------------------------- iob | Reserved | io_len | ------------------------------------------------- ------------------------------------------------------------- Layout of different NICs implementation: iob_data (aligned by CONFIG_IOB_ALIGNMENT) | | io_offset(CONFIG_NET_LL_GUARDSIZE) | | ------------------------------------------------- Ethernet | Reserved | ETH_HDRLEN | io_len | ---------------------------------|--------------- 8021Q | Reserved | ETH_8021Q_HDRLEN | io_len | ---------------------------------|--------------- ipforward | Reserved | io_len | ------------------------------------------------- -------------------------------------------------------------------- 2. Support iob offload to l2 driver to avoid unnecessary memory copy Support send/receive iob vectors directly between the NICs and l3/l4 stack to avoid unnecessary memory copies, especially on hardware that supports Scatter/gather, which can greatly improve performance. new interface to support iob offload: ------------------------------------------ | IOB version | original | |----------------------------------------| | devif_iob_poll() | devif_poll() | | ... | ... | ------------------------------------------ -------------------------------------------------------------------- 1> NIC hardware support Scatter/gather transfer TX: tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll("NIC"_txpoll) callback() // "NIC"_txpoll | dev->d_iob: | --------------- --------------- io_data iob1 | | | iob3 | | | \ --------------- --------------- --------------- | --------------- | iob0 | | | | iob2 | | | | --------------- | --------------- | \ | / / \ | / / ---------------------------------------------- NICs io vector | | | | | | | | | | ---------------------------------------------- RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | pkt/ipv[4/6]_input()/... | | NICs io vector receive(iov_base to each iobs) -------------------------------------------------------------------- 2> CONFIG_IOB_BUFSIZE is greater than MTU: TX: "(CONFIG_IOB_BUFSIZE) > (MAX_NETDEV_PKTSIZE + CONFIG_NET_GUARDSIZE + CONFIG_NET_LL_GUARDSIZE)" tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll("NIC"_txpoll) callback() // "NIC"_txpoll | "NIC"_send() (dev->d_iob->io_data[CONFIG_NET_LL_GUARDSIZE - NET_LL_HDRLEN(dev)]) RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | pkt/ipv[4/6]_input()/... | | NICs io vector receive(iov_base to io_data) -------------------------------------------------------------------- 3> Compatible with all old flat buffer NICs TX: tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll(devif_poll_callback()) devif_poll_callback() /* new interface, gather iobs to flat buffer */ / \ / \ devif_poll("NIC"_txpoll) "NIC"_send()(dev->d_buf) RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | netdev_input() /* new interface, Scatter/gather flat/iob buffer */ | | pkt/ipv[4|6]_input()/... | | NICs io vector receive(Orignal flat buffer) 3. Iperf passthrough on NuttX simulator: ------------------------------------------------- | Protocol | Server | Client | | |-----------------------------------------------| | TCP | 813 | 834 | Mbits/sec | | TCP(Offload) | 1720 | 1100 | Mbits/sec | | UDP | 22 | 757 | Mbits/sec | | UDP(Offload) | 25 | 1250 | Mbits/sec | ------------------------------------------------- Signed-off-by: chao an <anchao@xiaomi.com>
2022-11-23 12:39:40 +01:00
/****************************************************************************
* Name: ipv6_input
*
* Description:
* Receive an IPv6 packet from the network device. Verify and forward to
* L3 packet handling logic if the packet is destined for us.
*
* Input Parameters:
* dev - The device on which the packet was received and which contains
* the IPv6 packet.
* Returned Value:
* OK - The packet was processed (or dropped) and can be discarded.
* ERROR - Hold the packet and try again later. There is a listening
* socket but no receive in place to catch the packet yet. The
* device's d_len will be set to zero in this case as there is
* no outgoing data.
*
* If this function returns to the network driver with dev->d_len > 0,
* that is an indication to the driver that there is an outgoing response
* to this input.
*
* Assumptions:
* The network is locked.
*
****************************************************************************/
int ipv6_input(FAR struct net_driver_s *dev)
{
FAR uint8_t *buf;
int ret;
/* Store reception timestamp if enabled and not provided by hardware. */
#if defined(CONFIG_NET_TIMESTAMP) && !defined(CONFIG_ARCH_HAVE_NETDEV_TIMESTAMP)
clock_gettime(CLOCK_REALTIME, &dev->d_rxtime);
#endif
net/l2/l3/l4: add support of iob offload 1. Add new config CONFIG_NET_LL_GUARDSIZE to isolation of l2 stack, which will benefit l3(IP) layer for multi-MAC(l2) implementation, especially in some NICs such as celluler net driver. new configuration options: CONFIG_NET_LL_GUARDSIZE CONFIG_NET_LL_GUARDSIZE will reserved l2 buffer header size of network buffer to isolate the L2/L3 (MAC/IP) data on network layer, which will be beneficial to L3 network layer protocol transparent transmission and forwarding ------------------------------------------------------------ Layout of frist iob entry: iob_data (aligned by CONFIG_IOB_ALIGNMENT) | | io_offset(CONFIG_NET_LL_GUARDSIZE) | | ------------------------------------------------- iob | Reserved | io_len | ------------------------------------------------- ------------------------------------------------------------- Layout of different NICs implementation: iob_data (aligned by CONFIG_IOB_ALIGNMENT) | | io_offset(CONFIG_NET_LL_GUARDSIZE) | | ------------------------------------------------- Ethernet | Reserved | ETH_HDRLEN | io_len | ---------------------------------|--------------- 8021Q | Reserved | ETH_8021Q_HDRLEN | io_len | ---------------------------------|--------------- ipforward | Reserved | io_len | ------------------------------------------------- -------------------------------------------------------------------- 2. Support iob offload to l2 driver to avoid unnecessary memory copy Support send/receive iob vectors directly between the NICs and l3/l4 stack to avoid unnecessary memory copies, especially on hardware that supports Scatter/gather, which can greatly improve performance. new interface to support iob offload: ------------------------------------------ | IOB version | original | |----------------------------------------| | devif_iob_poll() | devif_poll() | | ... | ... | ------------------------------------------ -------------------------------------------------------------------- 1> NIC hardware support Scatter/gather transfer TX: tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll("NIC"_txpoll) callback() // "NIC"_txpoll | dev->d_iob: | --------------- --------------- io_data iob1 | | | iob3 | | | \ --------------- --------------- --------------- | --------------- | iob0 | | | | iob2 | | | | --------------- | --------------- | \ | / / \ | / / ---------------------------------------------- NICs io vector | | | | | | | | | | ---------------------------------------------- RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | pkt/ipv[4/6]_input()/... | | NICs io vector receive(iov_base to each iobs) -------------------------------------------------------------------- 2> CONFIG_IOB_BUFSIZE is greater than MTU: TX: "(CONFIG_IOB_BUFSIZE) > (MAX_NETDEV_PKTSIZE + CONFIG_NET_GUARDSIZE + CONFIG_NET_LL_GUARDSIZE)" tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll("NIC"_txpoll) callback() // "NIC"_txpoll | "NIC"_send() (dev->d_iob->io_data[CONFIG_NET_LL_GUARDSIZE - NET_LL_HDRLEN(dev)]) RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | pkt/ipv[4/6]_input()/... | | NICs io vector receive(iov_base to io_data) -------------------------------------------------------------------- 3> Compatible with all old flat buffer NICs TX: tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll(devif_poll_callback()) devif_poll_callback() /* new interface, gather iobs to flat buffer */ / \ / \ devif_poll("NIC"_txpoll) "NIC"_send()(dev->d_buf) RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | netdev_input() /* new interface, Scatter/gather flat/iob buffer */ | | pkt/ipv[4|6]_input()/... | | NICs io vector receive(Orignal flat buffer) 3. Iperf passthrough on NuttX simulator: ------------------------------------------------- | Protocol | Server | Client | | |-----------------------------------------------| | TCP | 813 | 834 | Mbits/sec | | TCP(Offload) | 1720 | 1100 | Mbits/sec | | UDP | 22 | 757 | Mbits/sec | | UDP(Offload) | 25 | 1250 | Mbits/sec | ------------------------------------------------- Signed-off-by: chao an <anchao@xiaomi.com>
2022-11-23 12:39:40 +01:00
if (dev->d_iob != NULL)
{
buf = dev->d_buf;
/* Set the device buffer to l2 */
dev->d_buf = NETLLBUF;
net/l2/l3/l4: add support of iob offload 1. Add new config CONFIG_NET_LL_GUARDSIZE to isolation of l2 stack, which will benefit l3(IP) layer for multi-MAC(l2) implementation, especially in some NICs such as celluler net driver. new configuration options: CONFIG_NET_LL_GUARDSIZE CONFIG_NET_LL_GUARDSIZE will reserved l2 buffer header size of network buffer to isolate the L2/L3 (MAC/IP) data on network layer, which will be beneficial to L3 network layer protocol transparent transmission and forwarding ------------------------------------------------------------ Layout of frist iob entry: iob_data (aligned by CONFIG_IOB_ALIGNMENT) | | io_offset(CONFIG_NET_LL_GUARDSIZE) | | ------------------------------------------------- iob | Reserved | io_len | ------------------------------------------------- ------------------------------------------------------------- Layout of different NICs implementation: iob_data (aligned by CONFIG_IOB_ALIGNMENT) | | io_offset(CONFIG_NET_LL_GUARDSIZE) | | ------------------------------------------------- Ethernet | Reserved | ETH_HDRLEN | io_len | ---------------------------------|--------------- 8021Q | Reserved | ETH_8021Q_HDRLEN | io_len | ---------------------------------|--------------- ipforward | Reserved | io_len | ------------------------------------------------- -------------------------------------------------------------------- 2. Support iob offload to l2 driver to avoid unnecessary memory copy Support send/receive iob vectors directly between the NICs and l3/l4 stack to avoid unnecessary memory copies, especially on hardware that supports Scatter/gather, which can greatly improve performance. new interface to support iob offload: ------------------------------------------ | IOB version | original | |----------------------------------------| | devif_iob_poll() | devif_poll() | | ... | ... | ------------------------------------------ -------------------------------------------------------------------- 1> NIC hardware support Scatter/gather transfer TX: tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll("NIC"_txpoll) callback() // "NIC"_txpoll | dev->d_iob: | --------------- --------------- io_data iob1 | | | iob3 | | | \ --------------- --------------- --------------- | --------------- | iob0 | | | | iob2 | | | | --------------- | --------------- | \ | / / \ | / / ---------------------------------------------- NICs io vector | | | | | | | | | | ---------------------------------------------- RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | pkt/ipv[4/6]_input()/... | | NICs io vector receive(iov_base to each iobs) -------------------------------------------------------------------- 2> CONFIG_IOB_BUFSIZE is greater than MTU: TX: "(CONFIG_IOB_BUFSIZE) > (MAX_NETDEV_PKTSIZE + CONFIG_NET_GUARDSIZE + CONFIG_NET_LL_GUARDSIZE)" tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll("NIC"_txpoll) callback() // "NIC"_txpoll | "NIC"_send() (dev->d_iob->io_data[CONFIG_NET_LL_GUARDSIZE - NET_LL_HDRLEN(dev)]) RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | pkt/ipv[4/6]_input()/... | | NICs io vector receive(iov_base to io_data) -------------------------------------------------------------------- 3> Compatible with all old flat buffer NICs TX: tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll(devif_poll_callback()) devif_poll_callback() /* new interface, gather iobs to flat buffer */ / \ / \ devif_poll("NIC"_txpoll) "NIC"_send()(dev->d_buf) RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | netdev_input() /* new interface, Scatter/gather flat/iob buffer */ | | pkt/ipv[4|6]_input()/... | | NICs io vector receive(Orignal flat buffer) 3. Iperf passthrough on NuttX simulator: ------------------------------------------------- | Protocol | Server | Client | | |-----------------------------------------------| | TCP | 813 | 834 | Mbits/sec | | TCP(Offload) | 1720 | 1100 | Mbits/sec | | UDP | 22 | 757 | Mbits/sec | | UDP(Offload) | 25 | 1250 | Mbits/sec | ------------------------------------------------- Signed-off-by: chao an <anchao@xiaomi.com>
2022-11-23 12:39:40 +01:00
ret = ipv6_in(dev);
dev->d_buf = buf;
return ret;
}
return netdev_input(dev, ipv6_in, true);
}
#endif /* CONFIG_NET_IPv6 */