1 .. SPDX-License-Identifier: BSD-3-Clause
2 Copyright(c) 2010-2014 Intel Corporation.
4 Intel Virtual Function Driver
5 =============================
7 Supported Intel® Ethernet Controllers (see the *DPDK Release Notes* for details)
8 support the following modes of operation in a virtualized environment:
10 * **SR-IOV mode**: Involves direct assignment of part of the port resources to different guest operating systems
11 using the PCI-SIG Single Root I/O Virtualization (SR IOV) standard,
12 also known as "native mode" or "pass-through" mode.
13 In this chapter, this mode is referred to as IOV mode.
15 * **VMDq mode**: Involves central management of the networking resources by an IO Virtual Machine (IOVM) or
16 a Virtual Machine Monitor (VMM), also known as software switch acceleration mode.
17 In this chapter, this mode is referred to as the Next Generation VMDq mode.
19 SR-IOV Mode Utilization in a DPDK Environment
20 ---------------------------------------------
22 The DPDK uses the SR-IOV feature for hardware-based I/O sharing in IOV mode.
23 Therefore, it is possible to partition SR-IOV capability on Ethernet controller NIC resources logically and
24 expose them to a virtual machine as a separate PCI function called a "Virtual Function".
25 Refer to :numref:`figure_single_port_nic`.
27 Therefore, a NIC is logically distributed among multiple virtual machines (as shown in :numref:`figure_single_port_nic`),
28 while still having global data in common to share with the Physical Function and other Virtual Functions.
29 The DPDK fm10kvf, i40evf, igbvf or ixgbevf as a Poll Mode Driver (PMD) serves for the Intel® 82576 Gigabit Ethernet Controller,
30 Intel® Ethernet Controller I350 family, Intel® 82599 10 Gigabit Ethernet Controller NIC,
31 Intel® Fortville 10/40 Gigabit Ethernet Controller NIC's virtual PCI function, or PCIe host-interface of the Intel Ethernet Switch
33 Meanwhile the DPDK Poll Mode Driver (PMD) also supports "Physical Function" of such NIC's on the host.
35 The DPDK PF/VF Poll Mode Driver (PMD) supports the Layer 2 switch on Intel® 82576 Gigabit Ethernet Controller,
36 Intel® Ethernet Controller I350 family, Intel® 82599 10 Gigabit Ethernet Controller,
37 and Intel® Fortville 10/40 Gigabit Ethernet Controller NICs so that guest can choose it for inter virtual machine traffic in SR-IOV mode.
39 For more detail on SR-IOV, please refer to the following documents:
41 * `SR-IOV provides hardware based I/O sharing <http://www.intel.com/network/connectivity/solutions/vmdc.htm>`_
43 * `PCI-SIG-Single Root I/O Virtualization Support on IA
44 <http://www.intel.com/content/www/us/en/pci-express/pci-sig-single-root-io-virtualization-support-in-virtualization-technology-for-connectivity-paper.html>`_
46 * `Scalable I/O Virtualized Servers <http://www.intel.com/content/www/us/en/virtualization/server-virtualization/scalable-i-o-virtualized-servers-paper.html>`_
48 .. _figure_single_port_nic:
50 .. figure:: img/single_port_nic.*
52 Virtualization for a Single Port NIC in SR-IOV Mode
55 Physical and Virtual Function Infrastructure
56 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
58 The following describes the Physical Function and Virtual Functions infrastructure for the supported Ethernet Controller NICs.
60 Virtual Functions operate under the respective Physical Function on the same NIC Port and therefore have no access
61 to the global NIC resources that are shared between other functions for the same NIC port.
63 A Virtual Function has basic access to the queue resources and control structures of the queues assigned to it.
64 For global resource access, a Virtual Function has to send a request to the Physical Function for that port,
65 and the Physical Function operates on the global resources on behalf of the Virtual Function.
66 For this out-of-band communication, an SR-IOV enabled NIC provides a memory buffer for each Virtual Function,
67 which is called a "Mailbox".
69 Intel® Ethernet Adaptive Virtual Function
70 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
71 Adaptive Virtual Function (IAVF) is a SR-IOV Virtual Function with the same device id (8086:1889) on different Intel Ethernet Controller.
72 IAVF Driver is VF driver which supports for all future Intel devices without requiring a VM update. And since this happens to be an adaptive VF driver,
73 every new drop of the VF driver would add more and more advanced features that can be turned on in the VM if the underlying HW device supports those
74 advanced features based on a device agnostic way without ever compromising on the base functionality. IAVF provides generic hardware interface and
75 interface between IAVF driver and a compliant PF driver is specified.
77 Intel products starting Ethernet Controller 700 Series to support Adaptive Virtual Function.
79 The way to generate Virtual Function is like normal, and the resource of VF assignment depends on the NIC Infrastructure.
81 For more detail on SR-IOV, please refer to the following documents:
83 * `Intel® IAVF HAS <https://www.intel.com/content/dam/www/public/us/en/documents/product-specifications/ethernet-adaptive-virtual-function-hardware-spec.pdf>`_
87 To use DPDK IAVF PMD on Intel® 700 Series Ethernet Controller, the device id (0x1889) need to specified during device
88 assignment in hypervisor. Take qemu for example, the device assignment should carry the IAVF device id (0x1889) like
89 ``-device vfio-pci,x-pci-device-id=0x1889,host=03:0a.0``.
91 The PCIE host-interface of Intel Ethernet Switch FM10000 Series VF infrastructure
92 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
94 In a virtualized environment, the programmer can enable a maximum of *64 Virtual Functions (VF)*
95 globally per PCIE host-interface of the Intel Ethernet Switch FM10000 Series device.
96 Each VF can have a maximum of 16 queue pairs.
97 The Physical Function in host could be only configured by the Linux* fm10k driver
98 (in the case of the Linux Kernel-based Virtual Machine [KVM]), DPDK PMD PF driver doesn't support it yet.
102 * Using Linux* fm10k driver:
104 .. code-block:: console
106 rmmod fm10k (To remove the fm10k module)
107 insmod fm0k.ko max_vfs=2,2 (To enable two Virtual Functions per port)
109 Virtual Function enumeration is performed in the following sequence by the Linux* pci driver for a dual-port NIC.
110 When you enable the four Virtual Functions with the above command, the four enabled functions have a Function#
111 represented by (Bus#, Device#, Function#) in sequence starting from 0 to 3.
114 * Virtual Functions 0 and 2 belong to Physical Function 0
116 * Virtual Functions 1 and 3 belong to Physical Function 1
120 The above is an important consideration to take into account when targeting specific packets to a selected port.
122 Intel® X710/XL710 Gigabit Ethernet Controller VF Infrastructure
123 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
125 In a virtualized environment, the programmer can enable a maximum of *128 Virtual Functions (VF)*
126 globally per Intel® X710/XL710 Gigabit Ethernet Controller NIC device.
127 The Physical Function in host could be either configured by the Linux* i40e driver
128 (in the case of the Linux Kernel-based Virtual Machine [KVM]) or by DPDK PMD PF driver.
129 When using both DPDK PMD PF/VF drivers, the whole NIC will be taken over by DPDK based application.
133 * Using Linux* i40e driver:
135 .. code-block:: console
137 rmmod i40e (To remove the i40e module)
138 insmod i40e.ko max_vfs=2,2 (To enable two Virtual Functions per port)
140 * Using the DPDK PMD PF i40e driver:
142 Kernel Params: iommu=pt, intel_iommu=on
144 .. code-block:: console
148 ./dpdk-devbind.py -b igb_uio bb:ss.f
149 echo 2 > /sys/bus/pci/devices/0000\:bb\:ss.f/max_vfs (To enable two VFs on a specific PCI device)
151 Launch the DPDK testpmd/example or your own host daemon application using the DPDK PMD library.
153 Virtual Function enumeration is performed in the following sequence by the Linux* pci driver for a dual-port NIC.
154 When you enable the four Virtual Functions with the above command, the four enabled functions have a Function#
155 represented by (Bus#, Device#, Function#) in sequence starting from 0 to 3.
158 * Virtual Functions 0 and 2 belong to Physical Function 0
160 * Virtual Functions 1 and 3 belong to Physical Function 1
164 The above is an important consideration to take into account when targeting specific packets to a selected port.
166 For Intel® X710/XL710 Gigabit Ethernet Controller, queues are in pairs. One queue pair means one receive queue and
167 one transmit queue. The default number of queue pairs per VF is 4, and can be 16 in maximum.
169 Intel® 82599 10 Gigabit Ethernet Controller VF Infrastructure
170 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
172 The programmer can enable a maximum of *63 Virtual Functions* and there must be *one Physical Function* per Intel® 82599
173 10 Gigabit Ethernet Controller NIC port.
174 The reason for this is that the device allows for a maximum of 128 queues per port and a virtual/physical function has to
175 have at least one queue pair (RX/TX).
176 The current implementation of the DPDK ixgbevf driver supports a single queue pair (RX/TX) per Virtual Function.
177 The Physical Function in host could be either configured by the Linux* ixgbe driver
178 (in the case of the Linux Kernel-based Virtual Machine [KVM]) or by DPDK PMD PF driver.
179 When using both DPDK PMD PF/VF drivers, the whole NIC will be taken over by DPDK based application.
183 * Using Linux* ixgbe driver:
185 .. code-block:: console
187 rmmod ixgbe (To remove the ixgbe module)
188 insmod ixgbe max_vfs=2,2 (To enable two Virtual Functions per port)
190 * Using the DPDK PMD PF ixgbe driver:
192 Kernel Params: iommu=pt, intel_iommu=on
194 .. code-block:: console
198 ./dpdk-devbind.py -b igb_uio bb:ss.f
199 echo 2 > /sys/bus/pci/devices/0000\:bb\:ss.f/max_vfs (To enable two VFs on a specific PCI device)
201 Launch the DPDK testpmd/example or your own host daemon application using the DPDK PMD library.
203 * Using the DPDK PMD PF ixgbe driver to enable VF RSS:
205 Same steps as above to install the modules of uio, igb_uio, specify max_vfs for PCI device, and
206 launch the DPDK testpmd/example or your own host daemon application using the DPDK PMD library.
208 The available queue number (at most 4) per VF depends on the total number of pool, which is
209 determined by the max number of VF at PF initialization stage and the number of queue specified
212 * If the max number of VFs (max_vfs) is set in the range of 1 to 32:
214 If the number of Rx queues is specified as 4 (``--rxq=4`` in testpmd), then there are totally 32
215 pools (ETH_32_POOLS), and each VF could have 4 Rx queues;
217 If the number of Rx queues is specified as 2 (``--rxq=2`` in testpmd), then there are totally 32
218 pools (ETH_32_POOLS), and each VF could have 2 Rx queues;
220 * If the max number of VFs (max_vfs) is in the range of 33 to 64:
222 If the number of Rx queues in specified as 4 (``--rxq=4`` in testpmd), then error message is expected
223 as ``rxq`` is not correct at this case;
225 If the number of rxq is 2 (``--rxq=2`` in testpmd), then there is totally 64 pools (ETH_64_POOLS),
226 and each VF have 2 Rx queues;
228 On host, to enable VF RSS functionality, rx mq mode should be set as ETH_MQ_RX_VMDQ_RSS
229 or ETH_MQ_RX_RSS mode, and SRIOV mode should be activated (max_vfs >= 1).
230 It also needs config VF RSS information like hash function, RSS key, RSS key length.
234 The limitation for VF RSS on Intel® 82599 10 Gigabit Ethernet Controller is:
235 The hash and key are shared among PF and all VF, the RETA table with 128 entries is also shared
236 among PF and all VF; So it could not to provide a method to query the hash and reta content per
237 VF on guest, while, if possible, please query them on host for the shared RETA information.
239 Virtual Function enumeration is performed in the following sequence by the Linux* pci driver for a dual-port NIC.
240 When you enable the four Virtual Functions with the above command, the four enabled functions have a Function#
241 represented by (Bus#, Device#, Function#) in sequence starting from 0 to 3.
244 * Virtual Functions 0 and 2 belong to Physical Function 0
246 * Virtual Functions 1 and 3 belong to Physical Function 1
250 The above is an important consideration to take into account when targeting specific packets to a selected port.
252 Intel® 82576 Gigabit Ethernet Controller and Intel® Ethernet Controller I350 Family VF Infrastructure
253 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
255 In a virtualized environment, an Intel® 82576 Gigabit Ethernet Controller serves up to eight virtual machines (VMs).
256 The controller has 16 TX and 16 RX queues.
257 They are generally referred to (or thought of) as queue pairs (one TX and one RX queue).
258 This gives the controller 16 queue pairs.
260 A pool is a group of queue pairs for assignment to the same VF, used for transmit and receive operations.
261 The controller has eight pools, with each pool containing two queue pairs, that is, two TX and two RX queues assigned to each VF.
263 In a virtualized environment, an Intel® Ethernet Controller I350 family device serves up to eight virtual machines (VMs) per port.
264 The eight queues can be accessed by eight different VMs if configured correctly (the i350 has 4x1GbE ports each with 8T X and 8 RX queues),
265 that means, one Transmit and one Receive queue assigned to each VF.
269 * Using Linux* igb driver:
271 .. code-block:: console
273 rmmod igb (To remove the igb module)
274 insmod igb max_vfs=2,2 (To enable two Virtual Functions per port)
276 * Using DPDK PMD PF igb driver:
278 Kernel Params: iommu=pt, intel_iommu=on modprobe uio
280 .. code-block:: console
283 ./dpdk-devbind.py -b igb_uio bb:ss.f
284 echo 2 > /sys/bus/pci/devices/0000\:bb\:ss.f/max_vfs (To enable two VFs on a specific pci device)
286 Launch DPDK testpmd/example or your own host daemon application using the DPDK PMD library.
288 Virtual Function enumeration is performed in the following sequence by the Linux* pci driver for a four-port NIC.
289 When you enable the four Virtual Functions with the above command, the four enabled functions have a Function#
290 represented by (Bus#, Device#, Function#) in sequence, starting from 0 to 7.
293 * Virtual Functions 0 and 4 belong to Physical Function 0
295 * Virtual Functions 1 and 5 belong to Physical Function 1
297 * Virtual Functions 2 and 6 belong to Physical Function 2
299 * Virtual Functions 3 and 7 belong to Physical Function 3
303 The above is an important consideration to take into account when targeting specific packets to a selected port.
305 Validated Hypervisors
306 ~~~~~~~~~~~~~~~~~~~~~
308 The validated hypervisor is:
310 * KVM (Kernel Virtual Machine) with Qemu, version 0.14.0
312 However, the hypervisor is bypassed to configure the Virtual Function devices using the Mailbox interface,
313 the solution is hypervisor-agnostic.
314 Xen* and VMware* (when SR- IOV is supported) will also be able to support the DPDK with Virtual Function driver support.
316 Expected Guest Operating System in Virtual Machine
317 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
319 The expected guest operating systems in a virtualized environment are:
321 * Fedora* 14 (64-bit)
323 * Ubuntu* 10.04 (64-bit)
325 For supported kernel versions, refer to the *DPDK Release Notes*.
327 Setting Up a KVM Virtual Machine Monitor
328 ----------------------------------------
330 The following describes a target environment:
332 * Host Operating System: Fedora 14
334 * Hypervisor: KVM (Kernel Virtual Machine) with Qemu version 0.14.0
336 * Guest Operating System: Fedora 14
338 * Linux Kernel Version: Refer to the *DPDK Getting Started Guide*
340 * Target Applications: l2fwd, l3fwd-vf
342 The setup procedure is as follows:
344 #. Before booting the Host OS, open **BIOS setup** and enable **Intel® VT features**.
346 #. While booting the Host OS kernel, pass the intel_iommu=on kernel command line argument using GRUB.
347 When using DPDK PF driver on host, pass the iommu=pt kernel command line argument in GRUB.
349 #. Download qemu-kvm-0.14.0 from
350 `http://sourceforge.net/projects/kvm/files/qemu-kvm/ <http://sourceforge.net/projects/kvm/files/qemu-kvm/>`_
351 and install it in the Host OS using the following steps:
353 When using a recent kernel (2.6.25+) with kvm modules included:
355 .. code-block:: console
357 tar xzf qemu-kvm-release.tar.gz
359 ./configure --prefix=/usr/local/kvm
362 sudo /sbin/modprobe kvm-intel
364 When using an older kernel, or a kernel from a distribution without the kvm modules,
365 you must download (from the same link), compile and install the modules yourself:
367 .. code-block:: console
369 tar xjf kvm-kmod-release.tar.bz2
374 sudo /sbin/modprobe kvm-intel
376 qemu-kvm installs in the /usr/local/bin directory.
378 For more details about KVM configuration and usage, please refer to:
380 `http://www.linux-kvm.org/page/HOWTO1 <http://www.linux-kvm.org/page/HOWTO1>`_.
382 #. Create a Virtual Machine and install Fedora 14 on the Virtual Machine.
383 This is referred to as the Guest Operating System (Guest OS).
385 #. Download and install the latest ixgbe driver from:
387 `http://downloadcenter.intel.com/Detail_Desc.aspx?agr=Y&DwnldID=14687 <http://downloadcenter.intel.com/Detail_Desc.aspx?agr=Y&DwnldID=14687>`_
391 When using Linux kernel ixgbe driver, unload the Linux ixgbe driver and reload it with the max_vfs=2,2 argument:
393 .. code-block:: console
396 modprobe ixgbe max_vfs=2,2
398 When using DPDK PMD PF driver, insert DPDK kernel module igb_uio and set the number of VF by sysfs max_vfs:
400 .. code-block:: console
404 ./dpdk-devbind.py -b igb_uio 02:00.0 02:00.1 0e:00.0 0e:00.1
405 echo 2 > /sys/bus/pci/devices/0000\:02\:00.0/max_vfs
406 echo 2 > /sys/bus/pci/devices/0000\:02\:00.1/max_vfs
407 echo 2 > /sys/bus/pci/devices/0000\:0e\:00.0/max_vfs
408 echo 2 > /sys/bus/pci/devices/0000\:0e\:00.1/max_vfs
412 You need to explicitly specify number of vfs for each port, for example,
413 in the command above, it creates two vfs for the first two ixgbe ports.
415 Let say we have a machine with four physical ixgbe ports:
426 The command above creates two vfs for device 0000:02:00.0:
428 .. code-block:: console
430 ls -alrt /sys/bus/pci/devices/0000\:02\:00.0/virt*
431 lrwxrwxrwx. 1 root root 0 Apr 13 05:40 /sys/bus/pci/devices/0000:02:00.0/virtfn1 -> ../0000:02:10.2
432 lrwxrwxrwx. 1 root root 0 Apr 13 05:40 /sys/bus/pci/devices/0000:02:00.0/virtfn0 -> ../0000:02:10.0
434 It also creates two vfs for device 0000:02:00.1:
436 .. code-block:: console
438 ls -alrt /sys/bus/pci/devices/0000\:02\:00.1/virt*
439 lrwxrwxrwx. 1 root root 0 Apr 13 05:51 /sys/bus/pci/devices/0000:02:00.1/virtfn1 -> ../0000:02:10.3
440 lrwxrwxrwx. 1 root root 0 Apr 13 05:51 /sys/bus/pci/devices/0000:02:00.1/virtfn0 -> ../0000:02:10.1
442 #. List the PCI devices connected and notice that the Host OS shows two Physical Functions (traditional ports)
443 and four Virtual Functions (two for each port).
444 This is the result of the previous step.
446 #. Insert the pci_stub module to hold the PCI devices that are freed from the default driver using the following command
447 (see http://www.linux-kvm.org/page/How_to_assign_devices_with_VT-d_in_KVM Section 4 for more information):
449 .. code-block:: console
451 sudo /sbin/modprobe pci-stub
453 Unbind the default driver from the PCI devices representing the Virtual Functions.
454 A script to perform this action is as follows:
456 .. code-block:: console
458 echo "8086 10ed" > /sys/bus/pci/drivers/pci-stub/new_id
459 echo 0000:08:10.0 > /sys/bus/pci/devices/0000:08:10.0/driver/unbind
460 echo 0000:08:10.0 > /sys/bus/pci/drivers/pci-stub/bind
462 where, 0000:08:10.0 belongs to the Virtual Function visible in the Host OS.
464 #. Now, start the Virtual Machine by running the following command:
466 .. code-block:: console
468 /usr/local/kvm/bin/qemu-system-x86_64 -m 4096 -smp 4 -boot c -hda lucid.qcow2 -device pci-assign,host=08:10.0
472 — -m = memory to assign
474 — -smp = number of smp cores
476 — -boot = boot option
478 — -hda = virtual disk image
480 — -device = device to attach
484 — The pci-assign,host=08:10.0 value indicates that you want to attach a PCI device
485 to a Virtual Machine and the respective (Bus:Device.Function)
486 numbers should be passed for the Virtual Function to be attached.
488 — qemu-kvm-0.14.0 allows a maximum of four PCI devices assigned to a VM,
489 but this is qemu-kvm version dependent since qemu-kvm-0.14.1 allows a maximum of five PCI devices.
491 — qemu-system-x86_64 also has a -cpu command line option that is used to select the cpu_model
492 to emulate in a Virtual Machine. Therefore, it can be used as:
494 .. code-block:: console
496 /usr/local/kvm/bin/qemu-system-x86_64 -cpu ?
498 (to list all available cpu_models)
500 /usr/local/kvm/bin/qemu-system-x86_64 -m 4096 -cpu host -smp 4 -boot c -hda lucid.qcow2 -device pci-assign,host=08:10.0
502 (to use the same cpu_model equivalent to the host cpu)
504 For more information, please refer to: `http://wiki.qemu.org/Features/CPUModels <http://wiki.qemu.org/Features/CPUModels>`_.
506 #. If use vfio-pci to pass through device instead of pci-assign, steps 8 and 9 need to be updated to bind device to vfio-pci and
507 replace pci-assign with vfio-pci when start virtual machine.
509 .. code-block:: console
511 sudo /sbin/modprobe vfio-pci
513 echo "8086 10ed" > /sys/bus/pci/drivers/vfio-pci/new_id
514 echo 0000:08:10.0 > /sys/bus/pci/devices/0000:08:10.0/driver/unbind
515 echo 0000:08:10.0 > /sys/bus/pci/drivers/vfio-pci/bind
517 /usr/local/kvm/bin/qemu-system-x86_64 -m 4096 -smp 4 -boot c -hda lucid.qcow2 -device vfio-pci,host=08:10.0
519 #. Install and run DPDK host app to take over the Physical Function. Eg.
521 .. code-block:: console
523 ./<build_dir>/app/dpdk-testpmd -l 0-3 -n 4 -- -i
525 #. Finally, access the Guest OS using vncviewer with the localhost:5900 port and check the lspci command output in the Guest OS.
526 The virtual functions will be listed as available for use.
528 #. Configure and install the DPDK on the Guest OS as normal, that is, there is no change to the normal installation procedure.
532 If you are unable to compile the DPDK and you are getting "error: CPU you selected does not support x86-64 instruction set",
533 power off the Guest OS and start the virtual machine with the correct -cpu option in the qemu- system-x86_64 command as shown in step 9.
534 You must select the best x86_64 cpu_model to emulate or you can select host option if available.
538 Run the DPDK l2fwd sample application in the Guest OS with Hugepages enabled.
539 For the expected benchmark performance, you must pin the cores from the Guest OS to the Host OS (taskset can be used to do this) and
540 you must also look at the PCI Bus layout on the board to ensure you are not running the traffic over the QPI Interface.
544 * The Virtual Machine Manager (the Fedora package name is virt-manager) is a utility for virtual machine management
545 that can also be used to create, start, stop and delete virtual machines.
546 If this option is used, step 2 and 6 in the instructions provided will be different.
548 * virsh, a command line utility for virtual machine management,
549 can also be used to bind and unbind devices to a virtual machine in Ubuntu.
550 If this option is used, step 6 in the instructions provided will be different.
552 * The Virtual Machine Monitor (see :numref:`figure_perf_benchmark`) is equivalent to a Host OS with KVM installed as described in the instructions.
554 .. _figure_perf_benchmark:
556 .. figure:: img/perf_benchmark.*
558 Performance Benchmark Setup
561 DPDK SR-IOV PMD PF/VF Driver Usage Model
562 ----------------------------------------
564 Fast Host-based Packet Processing
565 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
567 Software Defined Network (SDN) trends are demanding fast host-based packet handling.
568 In a virtualization environment,
569 the DPDK VF PMD driver performs the same throughput result as a non-VT native environment.
571 With such host instance fast packet processing, lots of services such as filtering, QoS,
572 DPI can be offloaded on the host fast path.
574 :numref:`figure_fast_pkt_proc` shows the scenario where some VMs directly communicate externally via a VFs,
575 while others connect to a virtual switch and share the same uplink bandwidth.
577 .. _figure_fast_pkt_proc:
579 .. figure:: img/fast_pkt_proc.*
581 Fast Host-based Packet Processing
584 SR-IOV (PF/VF) Approach for Inter-VM Communication
585 --------------------------------------------------
587 Inter-VM data communication is one of the traffic bottle necks in virtualization platforms.
588 SR-IOV device assignment helps a VM to attach the real device, taking advantage of the bridge in the NIC.
589 So VF-to-VF traffic within the same physical port (VM0<->VM1) have hardware acceleration.
590 However, when VF crosses physical ports (VM0<->VM2), there is no such hardware bridge.
591 In this case, the DPDK PMD PF driver provides host forwarding between such VMs.
593 :numref:`figure_inter_vm_comms` shows an example.
594 In this case an update of the MAC address lookup tables in both the NIC and host DPDK application is required.
596 In the NIC, writing the destination of a MAC address belongs to another cross device VM to the PF specific pool.
597 So when a packet comes in, its destination MAC address will match and forward to the host DPDK PMD application.
599 In the host DPDK application, the behavior is similar to L2 forwarding,
600 that is, the packet is forwarded to the correct PF pool.
601 The SR-IOV NIC switch forwards the packet to a specific VM according to the MAC destination address
602 which belongs to the destination VF on the VM.
604 .. _figure_inter_vm_comms:
606 .. figure:: img/inter_vm_comms.*
608 Inter-VM Communication