2 Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
5 Redistribution and use in source and binary forms, with or without
6 modification, are permitted provided that the following conditions
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
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.
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.
32 Vhost Sample Application
33 ========================
35 The vhost sample application demonstrates integration of the Data Plane Development Kit (DPDK)
36 with the Linux* KVM hypervisor by implementing the vhost-net offload API.
37 The sample application performs simple packet switching between virtual machines based on Media Access Control
38 (MAC) address or Virtual Local Area Network (VLAN) tag.
39 The splitting of ethernet traffic from an external switch is performed in hardware by the Virtual Machine Device Queues
40 (VMDQ) and Data Center Bridging (DCB) features of the IntelĀ® 82599 10 Gigabit Ethernet Controller.
45 Virtio networking (virtio-net) was developed as the Linux* KVM para-virtualized method for communicating network packets
46 between host and guest.
47 It was found that virtio-net performance was poor due to context switching and packet copying between host, guest, and QEMU.
48 The following figure shows the system architecture for a virtio- based networking (virtio-net).
52 **Figure16. QEMU Virtio-net (prior to vhost-net)**
54 .. image19_png has been renamed
58 The Linux* Kernel vhost-net module was developed as an offload mechanism for virtio-net.
59 The vhost-net module enables KVM (QEMU) to offload the servicing of virtio-net devices to the vhost-net kernel module,
60 reducing the context switching and packet copies in the virtual dataplane.
62 This is achieved by QEMU sharing the following information with the vhost-net module through the vhost-net API:
64 * The layout of the guest memory space, to enable the vhost-net module to translate addresses.
66 * The locations of virtual queues in QEMU virtual address space,
67 to enable the vhost module to read/write directly to and from the virtqueues.
69 * An event file descriptor (eventfd) configured in KVM to send interrupts to the virtio- net device driver in the guest.
70 This enables the vhost-net module to notify (call) the guest.
72 * An eventfd configured in KVM to be triggered on writes to the virtio-net device's
73 Peripheral Component Interconnect (PCI) config space.
74 This enables the vhost-net module to receive notifications (kicks) from the guest.
76 The following figure shows the system architecture for virtio-net networking with vhost-net offload.
80 **Figure 17. Virtio with Linux* Kernel Vhost**
82 .. image20_png has been renamed
89 The DPDK vhost-net sample code demonstrates KVM (QEMU) offloading the servicing of a Virtual Machine's (VM's)
90 virtio-net devices to a DPDK-based application in place of the kernel's vhost-net module.
92 The DPDK vhost-net sample code is a simple packet switching application with the following features:
94 * Management of virtio-net device creation/destruction events.
96 * Mapping of the VM's physical memory into the DPDK vhost-net sample code's address space.
98 * Triggering/receiving notifications to/from VMs via eventfds.
100 * A virtio-net back-end implementation providing a subset of virtio-net features.
102 * Packet switching between virtio-net devices and the network interface card,
103 including using VMDQs to reduce the switching that needs to be performed in software.
105 The following figure shows the architecture of the Vhost sample application.
109 **Figure 18. Vhost-net Architectural Overview**
111 .. image21_png has been renamed
115 The following figure shows the flow of packets through the vhost-net sample application.
119 **Figure 19. Packet Flow Through the vhost-net Sample Application**
121 .. image22_png has been renamed
123 |vhost_net_sample_app|
125 Supported Distributions
126 -----------------------
128 The example in this section have been validated with the following distributions:
137 This section lists prerequisite packages that must be installed.
139 Installing Packages on the Host
140 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
142 The vhost sample code uses the following packages; fuse, fuse-devel, and kernel- modules-extra.
144 #. Install Fuse Development Libraries and headers:
146 .. code-block:: console
148 yum -y install fuse fuse-devel
150 #. Install the Cuse Kernel Module:
152 .. code-block:: console
154 yum -y install kernel-modules-extra
156 Setting up the Execution Environment
157 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
159 The vhost sample code requires that QEMU allocates a VM's memory on the hugetlbfs file system.
160 As the vhost sample code requires hugepages,
161 the best practice is to partition the system into separate hugepage mount points for the VMs and the vhost sample code.
165 This is best-practice only and is not mandatory.
166 For systems that only support 2 MB page sizes,
167 both QEMU and vhost sample code can use the same hugetlbfs mount point without issue.
171 VMs with gigabytes of memory can benefit from having QEMU allocate their memory from 1 GB huge pages.
172 1 GB huge pages must be allocated at boot time by passing kernel parameters through the grub boot loader.
174 #. Calculate the maximum memory usage of all VMs to be run on the system.
175 Then, round this value up to the nearest Gigabyte the execution environment will require.
177 #. Edit the /etc/default/grub file, and add the following to the GRUB_CMDLINE_LINUX entry:
179 .. code-block:: console
181 GRUB_CMDLINE_LINUX="... hugepagesz=1G hugepages=<Number of hugepages required> default_hugepagesz=1G"
183 #. Update the grub boot loader:
185 .. code-block:: console
187 grub2-mkconfig -o /boot/grub2/grub.cfg
189 #. Reboot the system.
191 #. The hugetlbfs mount point (/dev/hugepages) should now default to allocating gigabyte pages.
195 Making the above modification will change the system default hugepage size to 1 GB for all applications.
197 **Vhost Sample Code**
199 In this section, we create a second hugetlbs mount point to allocate hugepages for the DPDK vhost sample code.
201 #. Allocate sufficient 2 MB pages for the DPDK vhost sample code:
203 .. code-block:: console
205 echo 256 > /sys/kernel/mm/hugepages/hugepages-2048kB/ nr_hugepages
207 #. Mount hugetlbs at a separate mount point for 2 MB pages:
209 .. code-block:: console
211 mount -t hugetlbfs nodev /mnt/huge -o pagesize=2M
213 The above steps can be automated by doing the following:
215 #. Edit /etc/fstab to add an entry to automatically mount the second hugetlbfs mount point:
219 hugetlbfs <tab> /mnt/huge <tab> hugetlbfs defaults,pagesize=1G 0 0
221 #. Edit the /etc/default/grub file, and add the following to the GRUB_CMDLINE_LINUX entry:
225 GRUB_CMDLINE_LINUX="... hugepagesz=2M hugepages=256 ... default_hugepagesz=1G"
227 #. Update the grub bootloader:
229 .. code-block:: console
231 grub2-mkconfig -o /boot/grub2/grub.cfg
233 #. Reboot the system.
237 Ensure that the default hugepage size after this setup is 1 GB.
239 Setting up the Guest Execution Environment
240 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
242 It is recommended for testing purposes that the DPDK testpmd sample application is used in the guest to forward packets,
243 the reasons for this are discussed in Section 22.7, "Running the Virtual Machine (QEMU)".
245 The testpmd application forwards packets between pairs of Ethernet devices,
246 it requires an even number of Ethernet devices (virtio or otherwise) to execute.
247 It is therefore recommended to create multiples of two virtio-net devices for each Virtual Machine either through libvirt or
248 at the command line as follows.
252 Observe that in the example, "-device" and "-netdev" are repeated for two virtio-net devices.
254 .. code-block:: console
256 user@target:~$ qemu-system-x86_64 ... \
257 -netdev tap,id=hostnet1,vhost=on,vhostfd=<open fd> \
258 -device virtio-net-pci, netdev=hostnet1,id=net1 \
259 -netdev tap,id=hostnet2,vhost=on,vhostfd=<open fd> \
260 -device virtio-net-pci, netdev=hostnet2,id=net1
263 Compiling the Sample Code
264 -------------------------
266 #. Go to the examples directory:
268 .. code-block:: console
270 export RTE_SDK=/path/to/rte_sdk cd ${RTE_SDK}/examples/vhost-net
272 #. Set the target (a default target is used if not specified). For example:
274 .. code-block:: console
276 export RTE_TARGET=x86_64-native-linuxapp-gcc
278 See the DPDK Getting Started Guide for possible RTE_TARGET values.
280 #. Build the application:
282 .. code-block:: console
288 Note For zero copy, need firstly disable CONFIG_RTE_MBUF_SCATTER_GATHER,
289 CONFIG_RTE_LIBRTE_IP_FRAG and CONFIG_RTE_LIBRTE_DISTRIBUTOR
290 in the config file and then re-configure and compile the core lib, and then build the application:
292 .. code-block:: console
294 vi ${RTE_SDK}/config/common_linuxapp
296 change it as follows:
300 CONFIG_RTE_MBUF_SCATTER_GATHER=n
301 CONFIG_RTE_LIBRTE_IP_FRAG=n
302 CONFIG_RTE_LIBRTE_DISTRIBUTOR=n
304 .. code-block:: console
307 make config ${RTE_TARGET}
308 make install ${RTE_TARGET}
309 cd ${RTE_SDK}/examples/vhost
312 #. Go to the eventfd_link directory:
314 .. code-block:: console
316 cd ${RTE_SDK}/examples/vhost-net/eventfd_link
318 #. Build the eventfd_link kernel module:
320 .. code-block:: console
324 Running the Sample Code
325 -----------------------
327 #. Install the cuse kernel module:
329 .. code-block:: console
333 #. Go to the eventfd_link directory:
335 .. code-block:: console
337 export RTE_SDK=/path/to/rte_sdk
338 cd ${RTE_SDK}/examples/vhost-net/eventfd_link
340 #. Install the eventfd_link module:
342 .. code-block:: console
344 insmod ./eventfd_link.ko
346 #. Go to the examples directory:
348 .. code-block:: console
350 export RTE_SDK=/path/to/rte_sdk
351 cd ${RTE_SDK}/examples/vhost-net
353 #. Run the vhost-switch sample code:
355 .. code-block:: console
357 user@target:~$ ./build/app/vhost-switch -c f -n 4 --huge-dir / mnt/huge -- -p 0x1 --dev-basename usvhost --dev-index 1
361 Please note the huge-dir parameter instructs the DPDK to allocate its memory from the 2 MB page hugetlbfs.
366 **Basename and Index.**
367 The DPDK vhost-net sample code uses a Linux* character device to communicate with QEMU.
368 The basename and the index are used to generate the character devices name.
370 /dev/<basename>-<index>
372 The index parameter is provided for a situation where multiple instances of the virtual switch is required.
374 For compatibility with the QEMU wrapper script, a base name of "usvhost" and an index of "1" should be used:
376 .. code-block:: console
378 user@target:~$ ./build/app/vhost-switch -c f -n 4 --huge-dir / mnt/huge -- -p 0x1 --dev-basename usvhost --dev-index 1
381 The vm2vm parameter disable/set mode of packet switching between guests in the host.
382 Value of "0" means disabling vm2vm implies that on virtual machine packet transmission will always go to the Ethernet port;
383 Value of "1" means software mode packet forwarding between guests, it needs packets copy in vHOST,
384 so valid only in one-copy implementation, and invalid for zero copy implementation;
385 value of "2" means hardware mode packet forwarding between guests, it allows packets go to the Ethernet port,
386 hardware L2 switch will determine which guest the packet should forward to or need send to external,
387 which bases on the packet destination MAC address and VLAN tag.
389 .. code-block:: console
391 user@target:~$ ./build/app/vhost-switch -c f -n 4 --huge-dir /mnt/huge -- --vm2vm [0,1,2]
393 **Mergeable Buffers.**
394 The mergeable buffers parameter controls how virtio-net descriptors are used for virtio-net headers.
395 In a disabled state, one virtio-net header is used per packet buffer;
396 in an enabled state one virtio-net header is used for multiple packets.
397 The default value is 0 or disabled since recent kernels virtio-net drivers show performance degradation with this feature is enabled.
399 .. code-block:: console
401 user@target:~$ ./build/app/vhost-switch -c f -n 4 --huge-dir / mnt/huge -- --mergeable [0,1]
404 The stats parameter controls the printing of virtio-net device statistics.
405 The parameter specifies an interval second to print statistics, with an interval of 0 seconds disabling statistics.
407 .. code-block:: console
409 user@target:~$ ./build/app/vhost-switch -c f -n 4 --huge-dir / mnt/huge -- --stats [0,n]
412 The rx-retry option enables/disables enqueue retries when the guests RX queue is full.
413 This feature resolves a packet loss that is observed at high data-rates,
414 by allowing it to delay and retry in the receive path.
415 This option is enabled by default.
417 .. code-block:: console
419 user@target:~$ ./build/app/vhost-switch -c f -n 4 --huge-dir / mnt/huge -- --rx-retry [0,1]
422 The rx-retry-num option specifies the number of retries on an RX burst,
423 it takes effect only when rx retry is enabled.
424 The default value is 4.
426 .. code-block:: console
428 user@target:~$ ./build/app/vhost-switch -c f -n 4 --huge-dir / mnt/huge -- --rx-retry 1 --rx-retry-num 5
430 **RX Retry Delay Time.**
431 The rx-retry-delay option specifies the timeout (in micro seconds) between retries on an RX burst,
432 it takes effect only when rx retry is enabled.
433 The default value is 15.
435 .. code-block:: console
437 user@target:~$ ./build/app/vhost-switch -c f -n 4 --huge-dir / mnt/huge -- --rx-retry 1 --rx-retry-delay 20
440 The zero copy option enables/disables the zero copy mode for RX/TX packet,
441 in the zero copy mode the packet buffer address from guest translate into host physical address
442 and then set directly as DMA address.
443 If the zero copy mode is disabled, then one copy mode is utilized in the sample.
444 This option is disabled by default.
446 .. code-block:: console
448 user@target:~$ ./build/app/vhost-switch -c f -n 4 --huge-dir /mnt/huge -- --zero-copy [0,1]
450 **RX descriptor number.**
451 The RX descriptor number option specify the Ethernet RX descriptor number,
452 Linux legacy virtio-net has different behaviour in how to use the vring descriptor from DPDK based virtio-net PMD,
453 the former likely allocate half for virtio header, another half for frame buffer,
454 while the latter allocate all for frame buffer,
455 this lead to different number for available frame buffer in vring,
456 and then lead to different Ethernet RX descriptor number could be used in zero copy mode.
457 So it is valid only in zero copy mode is enabled. The value is 32 by default.
459 .. code-block:: console
461 user@target:~$ ./build/app/vhost-switch -c f -n 4 --huge-dir /mnt/huge -- --zero-copy 1 --rx-desc-num [0, n]
463 **TX descriptornumber.**
464 The TX descriptor number option specify the Ethernet TX descriptor number, it is valid only in zero copy mode is enabled.
465 The value is 64 by default.
467 .. code-block:: console
469 user@target:~$ ./build/app/vhost-switch -c f -n 4 --huge-dir /mnt/huge -- --zero-copy 1 --tx-desc-num [0, n]
472 The VLAN strip option enable/disable the VLAN strip on host, if disabled, the guest will receive the packets with VLAN tag.
473 It is enabled by default.
475 .. code-block:: console
477 user@target:~$ ./build/app/vhost-switch -c f -n 4 --huge-dir /mnt/huge -- --vlan-strip [0, 1]
479 Running the Virtual Machine (QEMU)
480 ----------------------------------
482 QEMU must be executed with specific parameters to:
484 * Ensure the guest is configured to use virtio-net network adapters.
486 .. code-block:: console
488 user@target:~$ qemu-system-x86_64 ... -device virtio-net-pci, netdev=hostnet1,id=net1 ...
490 * Ensure the guest's virtio-net network adapter is configured with offloads disabled.
492 .. code-block:: console
494 user@target:~$ qemu-system-x86_64 ... -device virtio-net-pci, netdev=hostnet1,id=net1,csum=off,gso=off,guest_tso4=off,guest_ tso6=off,guest_ecn=off
496 * Redirect QEMU to communicate with the DPDK vhost-net sample code in place of the vhost-net kernel module.
498 .. code-block:: console
500 user@target:~$ qemu-system-x86_64 ... -netdev tap,id=hostnet1,vhost=on,vhostfd=<open fd> ...
502 * Enable the vhost-net sample code to map the VM's memory into its own process address space.
504 .. code-block:: console
506 user@target:~$ qemu-system-x86_64 ... -mem-prealloc -mem-path / dev/hugepages ...
510 The QEMU wrapper (qemu-wrap.py) is a Python script designed to automate the QEMU configuration described above.
511 It also facilitates integration with libvirt, although the script may also be used standalone without libvirt.
513 Redirecting QEMU to vhost-net Sample Code
514 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
516 To redirect QEMU to the vhost-net sample code implementation of the vhost-net API,
517 an open file descriptor must be passed to QEMU running as a child process.
519 .. code-block:: python
522 fd = os.open("/dev/usvhost-1", os.O_RDWR)
523 subprocess.call("qemu-system-x86_64 ... . -netdev tap,id=vhostnet0,vhost=on,vhostfd=" + fd +"...", shell=True)
527 This process is automated in the QEMU wrapper script discussed in Section 22.7.3.
529 Mapping the Virtual Machine's Memory
530 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
532 For the DPDK vhost-net sample code to be run correctly, QEMU must allocate the VM's memory on hugetlbfs.
533 This is done by specifying mem-prealloc and mem-path when executing QEMU.
534 The vhost-net sample code accesses the virtio-net device's virtual rings and packet buffers
535 by finding and mapping the VM's physical memory on hugetlbfs.
536 In this case, the path passed to the guest should be that of the 1 GB page hugetlbfs:
538 .. code-block:: console
540 user@target:~$ qemu-system-x86_64 ... -mem-prealloc -mem-path / dev/hugepages ...
544 This process is automated in the QEMU wrapper script discussed in Section 22.7.3.
549 The QEMU wrapper script automatically detects and calls QEMU with the necessary parameters required
550 to integrate with the vhost sample code.
551 It performs the following actions:
553 * Automatically detects the location of the hugetlbfs and inserts this into the command line parameters.
555 * Automatically open file descriptors for each virtio-net device and inserts this into the command line parameters.
557 * Disables offloads on each virtio-net device.
559 * Calls Qemu passing both the command line parameters passed to the script itself and those it has auto-detected.
561 The QEMU wrapper script will automatically configure calls to QEMU:
563 .. code-block:: console
565 user@target:~$ qemu-wrap.py -machine pc-i440fx-1.4,accel=kvm,usb=off -cpu SandyBridge -smp 4,sockets=4,cores=1,threads=1
566 -netdev tap,id=hostnet1,vhost=on -device virtio-net-pci,netdev=hostnet1,id=net1 -hda <disk img> -m 4096
568 which will become the following call to QEMU:
570 .. code-block:: console
572 /usr/local/bin/qemu-system-x86_64 -machine pc-i440fx-1.4,accel=kvm,usb=off -cpu SandyBridge -smp 4,sockets=4,cores=1,threads=1
573 -netdev tap,id=hostnet1,vhost=on,vhostfd=<open fd> -device virtio-net- pci,netdev=hostnet1,id=net1,
574 csum=off,gso=off,guest_tso4=off,gu est_tso6=off,guest_ecn=off -hda <disk img> -m 4096 -mem-path /dev/hugepages -mem-prealloc
579 The QEMU wrapper script (qemu-wrap.py) "wraps" libvirt calls to QEMU,
580 such that QEMU is called with the correct parameters described above.
581 To call the QEMU wrapper automatically from libvirt, the following configuration changes must be made:
583 * Place the QEMU wrapper script in libvirt's binary search PATH ($PATH).
584 A good location is in the directory that contains the QEMU binary.
586 * Ensure that the script has the same owner/group and file permissions as the QEMU binary.
588 * Update the VM xml file using virsh edit <vm name>:
590 * Set the VM to use the launch script
592 * Set the emulator path contained in the #<emulator><emulator/> tags For example,
593 replace <emulator>/usr/bin/qemu-kvm<emulator/> with <emulator>/usr/bin/qemu-wrap.py<emulator/>
595 * Set the VM's virtio-net device's to use vhost-net offload:
599 <interface type="network">
600 <model type="virtio"/>
601 <driver name="vhost"/>
604 * Enable libvirt to access the DPDK Vhost sample code's character device file by adding it
605 to controllers cgroup for libvirtd using the following steps:
609 cgroup_controllers = [ ... "devices", ... ] clear_emulator_capabilities = 0
610 user = "root" group = "root"
611 cgroup_device_acl = [
612 "/dev/null", "/dev/full", "/dev/zero",
613 "/dev/random", "/dev/urandom",
614 "/dev/ptmx", "/dev/kvm", "/dev/kqemu",
615 "/dev/rtc", "/dev/hpet", "/dev/net/tun",
616 "/dev/<devbase-name>-<index>",
619 * Disable SELinux or set to permissive mode.
622 * Mount cgroup device controller:
624 .. code-block:: console
626 user@target:~$ mkdir /dev/cgroup
627 user@target:~$ mount -t cgroup none /dev/cgroup -o devices
629 * Restart the libvirtd system process
631 For example, on Fedora* "systemctl restart libvirtd.service"
633 * Edit the configuration parameters section of the script:
635 * Configure the "emul_path" variable to point to the QEMU emulator.
639 emul_path = "/usr/local/bin/qemu-system-x86_64"
641 * Configure the "us_vhost_path" variable to point to the DPDK vhost- net sample code's character devices name.
642 DPDK vhost-net sample code's character device will be in the format "/dev/<basename>-<index>".
646 us_vhost_path = "/dev/usvhost-1"
651 **QEMU failing to allocate memory on hugetlbfs.**
653 file_ram_alloc: can't mmap RAM pages: Cannot allocate memory
655 When running QEMU the above error implies that it has failed to allocate memory for the Virtual Machine on the hugetlbfs.
656 This is typically due to insufficient hugepages being free to support the allocation request.
657 The number of free hugepages can be checked as follows:
659 .. code-block:: console
661 user@target:cat /sys/kernel/mm/hugepages/hugepages-<pagesize> / nr_hugepages
663 The command above indicates how many hugepages are free to support QEMU's allocation request.
665 Running DPDK in the Virtual Machine
666 -----------------------------------
668 For the DPDK vhost-net sample code to switch packets into the VM,
669 the sample code must first learn the MAC address of the VM's virtio-net device.
670 The sample code detects the address from packets being transmitted from the VM, similar to a learning switch.
672 This behavior requires no special action or configuration with the Linux* virtio-net driver in the VM
673 as the Linux* Kernel will automatically transmit packets during device initialization.
674 However, DPDK-based applications must be modified to automatically transmit packets during initialization
675 to facilitate the DPDK vhost- net sample code's MAC learning.
677 The DPDK testpmd application can be configured to automatically transmit packets during initialization
678 and to act as an L2 forwarding switch.
680 Testpmd MAC Forwarding
681 ~~~~~~~~~~~~~~~~~~~~~~
683 At high packet rates, a minor packet loss may be observed.
684 To resolve this issue, a "wait and retry" mode is implemented in the testpmd and vhost sample code.
685 In the "wait and retry" mode if the virtqueue is found to be full, then testpmd waits for a period of time before retrying to enqueue packets.
687 The "wait and retry" algorithm is implemented in DPDK testpmd as a forwarding method call "mac_retry".
688 The following sequence diagram describes the algorithm in detail.
692 **Figure 20. Packet Flow on TX in DPDK-testpmd**
694 .. image23_png has been renamed
701 The testpmd application is automatically built when DPDK is installed.
702 Run the testpmd application as follows:
704 .. code-block:: console
706 user@target:~$ x86_64-native-linuxapp-gcc/app/testpmd -c 0x3 -- n 4 -socket-mem 128 -- --burst=64 -i
708 The destination MAC address for packets transmitted on each port can be set at the command line:
710 .. code-block:: console
712 user@target:~$ x86_64-native-linuxapp-gcc/app/testpmd -c 0x3 -- n 4 -socket-mem 128 -- --burst=64 -i --eth- peer=0,aa:bb:cc:dd:ee:ff --eth-peer=1,ff,ee,dd,cc,bb,aa
714 * Packets received on port 1 will be forwarded on port 0 to MAC address
718 * Packets received on port 0 will be forwarded on port 1 to MAC address
722 The testpmd application can then be configured to act as an L2 forwarding application:
724 .. code-block:: console
726 testpmd> set fwd mac_retry
728 The testpmd can then be configured to start processing packets,
729 transmitting packets first so the DPDK vhost sample code on the host can learn the MAC address:
731 .. code-block:: console
733 testpmd> start tx_first
737 Please note "set fwd mac_retry" is used in place of "set fwd mac_fwd" to ensure the retry feature is activated.
739 Passing Traffic to the Virtual Machine Device
740 ---------------------------------------------
742 For a virtio-net device to receive traffic,
743 the traffic's Layer 2 header must include both the virtio-net device's MAC address and VLAN tag.
744 The DPDK sample code behaves in a similar manner to a learning switch in that
745 it learns the MAC address of the virtio-net devices from the first transmitted packet.
746 On learning the MAC address,
747 the DPDK vhost sample code prints a message with the MAC address and VLAN tag virtio-net device.
750 .. code-block:: console
752 DATA: (0) MAC_ADDRESS cc:bb:bb:bb:bb:bb and VLAN_TAG 1000 registered
754 The above message indicates that device 0 has been registered with MAC address cc:bb:bb:bb:bb:bb and VLAN tag 1000.
755 Any packets received on the NIC with these values is placed on the devices receive queue.
756 When a virtio-net device transmits packets, the VLAN tag is added to the packet by the DPDK vhost sample code.
758 .. |vhost_net_arch| image:: img/vhost_net_arch.*
760 .. |qemu_virtio_net| image:: img/qemu_virtio_net.*
762 .. |tx_dpdk_testpmd| image:: img/tx_dpdk_testpmd.*
764 .. |vhost_net_sample_app| image:: img/vhost_net_sample_app.*
766 .. |virtio_linux_vhost| image:: img/virtio_linux_vhost.*