1.. BSD LICENSE 2 Copyright(c) 2010-2015 Intel Corporation. All rights reserved. 3 All rights reserved. 4 5 Redistribution and use in source and binary forms, with or without 6 modification, are permitted provided that the following conditions 7 are met: 8 9 * Redistributions of source code must retain the above copyright 10 notice, this list of conditions and the following disclaimer. 11 * Redistributions in binary form must reproduce the above copyright 12 notice, this list of conditions and the following disclaimer in 13 the documentation and/or other materials provided with the 14 distribution. 15 * Neither the name of Intel Corporation nor the names of its 16 contributors may be used to endorse or promote products derived 17 from this software without specific prior written permission. 18 19 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 20 "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT 21 LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR 22 A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT 23 OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, 24 SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT 25 LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, 26 DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY 27 THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT 28 (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE 29 OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 30 31.. _kni: 32 33Kernel NIC Interface 34==================== 35 36The DPDK Kernel NIC Interface (KNI) allows userspace applications access to the Linux* control plane. 37 38The benefits of using the DPDK KNI are: 39 40* Faster than existing Linux TUN/TAP interfaces 41 (by eliminating system calls and copy_to_user()/copy_from_user() operations. 42 43* Allows management of DPDK ports using standard Linux net tools such as ethtool, ifconfig and tcpdump. 44 45* Allows an interface with the kernel network stack. 46 47The components of an application using the DPDK Kernel NIC Interface are shown in :numref:`figure_kernel_nic_intf`. 48 49.. _figure_kernel_nic_intf: 50 51.. figure:: img/kernel_nic_intf.* 52 53 Components of a DPDK KNI Application 54 55 56The DPDK KNI Kernel Module 57-------------------------- 58 59The KNI kernel loadable module provides support for two types of devices: 60 61* A Miscellaneous device (/dev/kni) that: 62 63 * Creates net devices (via ioctl calls). 64 65 * Maintains a kernel thread context shared by all KNI instances 66 (simulating the RX side of the net driver). 67 68 * For single kernel thread mode, maintains a kernel thread context shared by all KNI instances 69 (simulating the RX side of the net driver). 70 71 * For multiple kernel thread mode, maintains a kernel thread context for each KNI instance 72 (simulating the RX side of the new driver). 73 74* Net device: 75 76 * Net functionality provided by implementing several operations such as netdev_ops, 77 header_ops, ethtool_ops that are defined by struct net_device, 78 including support for DPDK mbufs and FIFOs. 79 80 * The interface name is provided from userspace. 81 82 * The MAC address can be the real NIC MAC address or random. 83 84KNI Creation and Deletion 85------------------------- 86 87The KNI interfaces are created by a DPDK application dynamically. 88The interface name and FIFO details are provided by the application through an ioctl call 89using the rte_kni_device_info struct which contains: 90 91* The interface name. 92 93* Physical addresses of the corresponding memzones for the relevant FIFOs. 94 95* Mbuf mempool details, both physical and virtual (to calculate the offset for mbuf pointers). 96 97* PCI information. 98 99* Core affinity. 100 101Refer to rte_kni_common.h in the DPDK source code for more details. 102 103The physical addresses will be re-mapped into the kernel address space and stored in separate KNI contexts. 104 105The KNI interfaces can be deleted by a DPDK application dynamically after being created. 106Furthermore, all those KNI interfaces not deleted will be deleted on the release operation 107of the miscellaneous device (when the DPDK application is closed). 108 109DPDK mbuf Flow 110-------------- 111 112To minimize the amount of DPDK code running in kernel space, the mbuf mempool is managed in userspace only. 113The kernel module will be aware of mbufs, 114but all mbuf allocation and free operations will be handled by the DPDK application only. 115 116:numref:`figure_pkt_flow_kni` shows a typical scenario with packets sent in both directions. 117 118.. _figure_pkt_flow_kni: 119 120.. figure:: img/pkt_flow_kni.* 121 122 Packet Flow via mbufs in the DPDK KNI 123 124 125Use Case: Ingress 126----------------- 127 128On the DPDK RX side, the mbuf is allocated by the PMD in the RX thread context. 129This thread will enqueue the mbuf in the rx_q FIFO. 130The KNI thread will poll all KNI active devices for the rx_q. 131If an mbuf is dequeued, it will be converted to a sk_buff and sent to the net stack via netif_rx(). 132The dequeued mbuf must be freed, so the same pointer is sent back in the free_q FIFO. 133 134The RX thread, in the same main loop, polls this FIFO and frees the mbuf after dequeuing it. 135 136Use Case: Egress 137---------------- 138 139For packet egress the DPDK application must first enqueue several mbufs to create an mbuf cache on the kernel side. 140 141The packet is received from the Linux net stack, by calling the kni_net_tx() callback. 142The mbuf is dequeued (without waiting due the cache) and filled with data from sk_buff. 143The sk_buff is then freed and the mbuf sent in the tx_q FIFO. 144 145The DPDK TX thread dequeues the mbuf and sends it to the PMD (via rte_eth_tx_burst()). 146It then puts the mbuf back in the cache. 147 148Ethtool 149------- 150 151Ethtool is a Linux-specific tool with corresponding support in the kernel 152where each net device must register its own callbacks for the supported operations. 153The current implementation uses the igb/ixgbe modified Linux drivers for ethtool support. 154Ethtool is not supported in i40e and VMs (VF or EM devices). 155 156Link state and MTU change 157------------------------- 158 159Link state and MTU change are network interface specific operations usually done via ifconfig. 160The request is initiated from the kernel side (in the context of the ifconfig process) 161and handled by the user space DPDK application. 162The application polls the request, calls the application handler and returns the response back into the kernel space. 163 164The application handlers can be registered upon interface creation or explicitly registered/unregistered in runtime. 165This provides flexibility in multiprocess scenarios 166(where the KNI is created in the primary process but the callbacks are handled in the secondary one). 167The constraint is that a single process can register and handle the requests. 168 169KNI Working as a Kernel vHost Backend 170------------------------------------- 171 172vHost is a kernel module usually working as the backend of virtio (a para- virtualization driver framework) 173to accelerate the traffic from the guest to the host. 174The DPDK Kernel NIC interface provides the ability to hookup vHost traffic into userspace DPDK application. 175Together with the DPDK PMD virtio, it significantly improves the throughput between guest and host. 176In the scenario where DPDK is running as fast path in the host, kni-vhost is an efficient path for the traffic. 177 178Overview 179~~~~~~~~ 180 181vHost-net has three kinds of real backend implementations. They are: 1) tap, 2) macvtap and 3) RAW socket. 182The main idea behind kni-vhost is making the KNI work as a RAW socket, attaching it as the backend instance of vHost-net. 183It is using the existing interface with vHost-net, so it does not require any kernel hacking, 184and is fully-compatible with the kernel vhost module. 185As vHost is still taking responsibility for communicating with the front-end virtio, 186it naturally supports both legacy virtio -net and the DPDK PMD virtio. 187There is a little penalty that comes from the non-polling mode of vhost. 188However, it scales throughput well when using KNI in multi-thread mode. 189 190.. _figure_vhost_net_arch2: 191 192.. figure:: img/vhost_net_arch.* 193 194 vHost-net Architecture Overview 195 196 197Packet Flow 198~~~~~~~~~~~ 199 200There is only a minor difference from the original KNI traffic flows. 201On transmit side, vhost kthread calls the RAW socket's ops sendmsg and it puts the packets into the KNI transmit FIFO. 202On the receive side, the kni kthread gets packets from the KNI receive FIFO, puts them into the queue of the raw socket, 203and wakes up the task in vhost kthread to begin receiving. 204All the packet copying, irrespective of whether it is on the transmit or receive side, 205happens in the context of vhost kthread. 206Every vhost-net device is exposed to a front end virtio device in the guest. 207 208.. _figure_kni_traffic_flow: 209 210.. figure:: img/kni_traffic_flow.* 211 212 KNI Traffic Flow 213 214 215Sample Usage 216~~~~~~~~~~~~ 217 218Before starting to use KNI as the backend of vhost, the CONFIG_RTE_KNI_VHOST configuration option must be turned on. 219Otherwise, by default, KNI will not enable its backend support capability. 220 221Of course, as a prerequisite, the vhost/vhost-net kernel CONFIG should be chosen before compiling the kernel. 222 223#. Compile the DPDK and insert uio_pci_generic/igb_uio kernel modules as normal. 224 225#. Insert the KNI kernel module: 226 227 .. code-block:: console 228 229 insmod ./rte_kni.ko 230 231 If using KNI in multi-thread mode, use the following command line: 232 233 .. code-block:: console 234 235 insmod ./rte_kni.ko kthread_mode=multiple 236 237#. Running the KNI sample application: 238 239 .. code-block:: console 240 241 examples/kni/build/app/kni -c -0xf0 -n 4 -- -p 0x3 -P --config="(0,4,6),(1,5,7)" 242 243 This command runs the kni sample application with two physical ports. 244 Each port pins two forwarding cores (ingress/egress) in user space. 245 246#. Assign a raw socket to vhost-net during qemu-kvm startup. 247 The DPDK does not provide a script to do this since it is easy for the user to customize. 248 The following shows the key steps to launch qemu-kvm with kni-vhost: 249 250 .. code-block:: bash 251 252 #!/bin/bash 253 echo 1 > /sys/class/net/vEth0/sock_en 254 fd=`cat /sys/class/net/vEth0/sock_fd` 255 qemu-kvm \ 256 -name vm1 -cpu host -m 2048 -smp 1 -hda /opt/vm-fc16.img \ 257 -netdev tap,fd=$fd,id=hostnet1,vhost=on \ 258 -device virti-net-pci,netdev=hostnet1,id=net1,bus=pci.0,addr=0x4 259 260It is simple to enable raw socket using sysfs sock_en and get raw socket fd using sock_fd under the KNI device node. 261 262Then, using the qemu-kvm command with the -netdev option to assign such raw socket fd as vhost's backend. 263 264.. note:: 265 266 The key word tap must exist as qemu-kvm now only supports vhost with a tap backend, so here we cheat qemu-kvm by an existing fd. 267 268Compatibility Configure Option 269~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 270 271There is a CONFIG_RTE_KNI_VHOST_VNET_HDR_EN configuration option in DPDK configuration file. 272By default, it set to n, which means do not turn on the virtio net header, 273which is used to support additional features (such as, csum offload, vlan offload, generic-segmentation and so on), 274since the kni-vhost does not yet support those features. 275 276Even if the option is turned on, kni-vhost will ignore the information that the header contains. 277When working with legacy virtio on the guest, it is better to turn off unsupported offload features using ethtool -K. 278Otherwise, there may be problems such as an incorrect L4 checksum error. 279