3. Running the Application
3.1. EAL Command-line Options
3.2. Testpmd Command-line Options
The following are the command-line options for the testpmd applications.
They must be separated from the EAL options, shown in the previous section, with a
sudo ./dpdk-testpmd -l 0-3 -n 4 -- -i --portmask=0x1 --nb-cores=2
The command line options are:
Run testpmd in interactive mode. In this mode, the testpmd starts with a prompt that can be used to start and stop forwarding, configure the application and display stats on the current packet processing session. See Testpmd Runtime Functions for more details.
In non-interactive mode, the application starts with the configuration specified on the command-line and immediately enters forwarding mode.
Display a help message and quit.
Start forwarding on initialization.
Start forwarding, after sending a burst of packets first.
This flag should be only used in non-interactive mode.
Display statistics every PERIOD seconds, if interactive mode is disabled. The default value is 0, which means that the statistics will not be displayed.
Display comma-separated list of extended statistics every PERIOD seconds as specified in
--stats-periodor when used with interactive commands that show Rx/Tx statistics (i.e. ‘show port stats’).
Set the number of forwarding cores, where 1 <= N <= “number of cores” or
RTE_MAX_LCOREfrom the configuration file. The default value is 1.
Set the number of forwarding ports, where 1 <= N <= “number of ports” on the board or
RTE_MAX_ETHPORTSfrom the configuration file. The default value is the number of ports on the board.
Set the hexadecimal bitmask of the cores running the packet forwarding test. The main lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
Set the hexadecimal bitmask of the ports used by the packet forwarding test.
Set the forwarding ports based on the user input used by the packet forwarding test. ‘-‘ denotes a range of ports to set including the two specified port IDs ‘,’ separates multiple port values. Possible examples like –portlist=0,1 or –portlist=0-2 or –portlist=0,1-2 etc
Enable NUMA-aware allocation of RX/TX rings and of RX memory buffers (mbufs). [Default setting]
Disable NUMA-aware allocation of RX/TX rings and of RX memory buffers (mbufs).
Specify the socket on which the memory pool to be used by the port will be allocated.
Specify the socket on which the TX/RX rings for the port will be allocated. Where flag is 1 for RX, 2 for TX, and 3 for RX and TX.
Set the socket from which all memory is allocated in NUMA mode, where 0 <= N < number of sockets on the board.
Set the data size of the mbufs used to N bytes, where N < 65536. The default value is 2048. If multiple mbuf-size values are specified the extra memory pools will be created for allocating mbufs to receive packets with buffer splitting features.
Set the number of mbufs to be allocated in the mbuf pools, where N > 1024.
Set the maximum packet size to N bytes, where N >= 64. The default value is 1518.
Set the maximum LRO aggregated packet size to N bytes, where N >= 64.
Use a configuration file containing the Ethernet addresses of the peer ports. The configuration file should contain the Ethernet addresses on separate lines:
XX:XX:XX:XX:XX:01 XX:XX:XX:XX:XX:02 ...
Set the MAC address
XX:XX:XX:XX:XX:XXof the peer port N, where 0 <= N <
Set the source and destination IP address used when doing transmit only test. The defaults address values are source 198.18.0.1 and destination 198.18.0.2. These are special purpose addresses reserved for benchmarking (RFC 5735).
Set the source and destination UDP port number for transmit test only test. The default port is the port 9 which is defined for the discard protocol (RFC 863).
Set Flow Director mode where mode is either
perfect. See flow_director_mask for more details.
Set Flow Director hash match reporting mode where mode is
match(the default) or
Set Flow Director allocated memory size, where N is 64K, 128K or 256K. Sizes are in kilobytes. The default is 64.
Set the flexbytes offset. The offset is defined in words (not bytes) counted from the first byte of the destination Ethernet MAC address, where N is 0 <= N <= 32. The default value is 0x6.
Set the drop-queue. In perfect filter mode, when a rule is added with queue = -1, the packet will be enqueued into the RX drop-queue. If the drop-queue does not exist, the packet is dropped. The default value is N=127.
Disable hardware CRC stripping.
Enable large receive offload.
Enable hardware RX checksum offload.
Enable scatter (multi-segment) RX.
Enable hardware VLAN.
Enable hardware VLAN filter.
Enable hardware VLAN strip.
Enable hardware VLAN extend.
Enable hardware QINQ strip.
Enable per-queue packet drop for packets with no descriptors.
Disable RSS (Receive Side Scaling).
Set port topology, where mode is
pairedmode, the forwarding is between pairs of ports, for example: (0,1), (2,3), (4,5).
chainedmode, the forwarding is to the next available port in the port mask, for example: (0,1), (1,2), (2,0).
The ordering of the ports can be changed using the portlist testpmd runtime function.
loopmode, ingress traffic is simply transmitted back on the same interface.
Set the forwarding mode where
modeis one of the following:
io (the default) mac macswap flowgen rxonly txonly csum icmpecho ieee1588 tm noisy 5tswap shared-rxq
Set RSS functions for IPv4/IPv6 only.
Set RSS functions for IPv4/IPv6 and UDP.
Set the number of RX queues per port to N, where 1 <= N <= 65535. The default value is 1.
Set the number of descriptors in the RX rings to N, where N > 0. The default value is 128.
Set the number of TX queues per port to N, where 1 <= N <= 65535. The default value is 1.
Set the number of descriptors in the TX rings to N, where N > 0. The default value is 512.
Set the number of hairpin queues per port to N, where 1 <= N <= 65535. The default value is 0. The number of hairpin queues are added to the number of TX queues and to the number of RX queues. then the first RX hairpin is binded to the first TX hairpin, the second RX hairpin is binded to the second TX hairpin and so on. The index of the first RX hairpin queue is the number of RX queues as configured using –rxq. The index of the first TX hairpin queue is the number of TX queues as configured using –txq.
Set the number of packets per burst to N, where 1 <= N <= 512. The default value is 32. If set to 0, driver default is used if defined. Else, if driver default is not defined, default of 32 is used.
Set the number of each packet clones to be sent in flowgen mode. Sending clones reduces host CPU load on creating packets and may help in testing extreme speeds or maxing out Tx packet performance. N should be not zero, but less than ‘burst’ parameter.
Set the number of flows to be generated in flowgen mode, where 1 <= N <= INT32_MAX.
Set the cache of mbuf memory pools to N, where 0 <= N <= 512. The default value is 16.
Set the prefetch threshold register of RX rings to N, where N >= 0. The default value is 8.
Set the host threshold register of RX rings to N, where N >= 0. The default value is 8.
Set the free threshold of RX descriptors to N, where 0 <= N < value of –rxd. The default value is 0.
Set the write-back threshold register of RX rings to N, where N >= 0. The default value is 4.
Set the prefetch threshold register of TX rings to N, where N >= 0. The default value is 36.
Set the host threshold register of TX rings to N, where N >= 0. The default value is 0.
Set the write-back threshold register of TX rings to N, where N >= 0. The default value is 0.
Set the transmit free threshold of TX rings to N, where 0 <= N <= value of
--txd. The default value is 0.
Set the transmit RS bit threshold of TX rings to N, where 0 <= N <= value of
--txd. The default value is 0.
Don’t flush the RX streams before starting forwarding. Used mainly with the PCAP PMD.
Set the offsets of packet segments on receiving if split feature is engaged. Affects only the queues configured with split offloads (currently BUFFER_SPLIT is supported only).
Set the length of segments to scatter packets on receiving if split feature is engaged. Affects only the queues configured with split offloads (currently BUFFER_SPLIT is supported only). Optionally the multiple memory pools can be specified with –mbuf-size command line parameter and the mbufs to receive will be allocated sequentially from these extra memory pools.
Set TX segment sizes or total packet length. Valid for
Generate multiple flows in txonly mode.
Create queues in shared Rx queue mode if device supports. Shared Rx queues are grouped per X ports. X defaults to UINT32_MAX, implies all ports join share group 1. Forwarding engine “shared-rxq” should be used for shared Rx queues. This engine does Rx only and update stream statistics accordingly.
Set a forced link speed to the ethernet port:
10 - 10Mbps (not supported) 100 - 100Mbps (not supported) 1000 - 1Gbps 10000 - 10Gbps 25000 - 25Gbps 40000 - 40Gbps 50000 - 50Gbps 100000 - 100Gbps 200000 - 200Gbps ...
Disable check on link status when starting/stopping ports.
Do not automatically start all ports. This allows testing configuration of rx and tx queues before device is started for the first time.
Disable LSC interrupts for all ports, even those supporting it.
Disable RMV interrupts for all ports, even those supporting it.
Set the logical core N to perform bitrate calculation.
Enable printing the occurrence of the designated event. Using all will enable all of them.
Disable printing the occurrence of the designated event. Using all will disable all of them.
Providing this parameter requests flow API isolated mode on all ports at initialization time. It ensures all traffic is received through the configured flow rules only (see flow command).
Ports that do not support this mode are automatically discarded.
Set the hexadecimal bitmask of TX queue offloads. The default value is 0.
Set the hexadecimal bitmask of RX queue offloads. The default value is 0.
Enable device event monitor mechanism for hotplug.
Set the UDP port number of tunnel VXLAN-GPE to N. The default value is 4790.
Set the UDP port number that is used for parsing the GENEVE protocol to N. HW may be configured with another tunnel Geneve port. The default value is 6081.
Enable locking all memory.
Disable locking all memory.
Select mempool allocation mode:
- native: create and populate mempool using native DPDK memory
- anon: create mempool using native DPDK memory, but populate using anonymous memory
- xmem: create and populate mempool using externally and anonymously allocated area
- xmemhuge: create and populate mempool using externally and anonymously allocated hugepage area
Set the number of maximum elements of the FIFO queue to be created for buffering packets. Only available with the noisy forwarding mode. The default value is 0.
Set the time before packets in the FIFO queue is flushed. Only available with the noisy forwarding mode. The default value is 0.
Set the size of the noisy neighbor simulation memory buffer in MB to N. Only available with the noisy forwarding mode. The default value is 0.
Set the number of reads to be done in noisy neighbor simulation memory buffer to N. Only available with the noisy forwarding mode. The default value is 0.
Set the number of writes to be done in noisy neighbor simulation memory buffer to N. Only available with the noisy forwarding mode. The default value is 0.
Set the number of r/w accesses to be done in noisy neighbor simulation memory buffer to N. Only available with the noisy forwarding mode. The default value is 0.
Enable to create mempool which is not IOVA contiguous. Valid only with –mp-alloc=anon. The default value is 0.
Set the hexadecimal bitmask of RX multi queue mode which can be enabled. The default value is 0x7:
RTE_ETH_MQ_RX_RSS_FLAG | RTE_ETH_MQ_RX_DCB_FLAG | RTE_ETH_MQ_RX_VMDQ_FLAG
Enable measurement of CPU cycles per packet.
Enable display of RX and TX burst stats.
Set the hairpin port mode with bitmask, only valid when hairpin queues number is set:
bit 4 - explicit Tx flow rule bit 1 - two hairpin ports paired bit 0 - two hairpin ports loop
The default value is 0. Hairpin will use single port mode and implicit Tx flow mode.
3.2.1. Testpmd Multi-Process Command-line Options
The following are the command-line options for testpmd multi-process support:
- primary process:
sudo ./dpdk-testpmd -a xxx --proc-type=auto -l 0-1 -- -i --rxq=4 --txq=4 \ --num-procs=2 --proc-id=0
- secondary process:
sudo ./dpdk-testpmd -a xxx --proc-type=auto -l 2-3 -- -i --rxq=4 --txq=4 \ --num-procs=2 --proc-id=1
The command line options are:
The number of processes which will be used.
The ID of the current process (ID < num-procs). ID should be different in primary process and secondary process, which starts from ‘0’.
Calculation rule for queue:
All queues are allocated to different processes based on
Calculation rule for the testpmd to allocate queues to each process:
* start(queue start id) = proc_id * nb_q / num_procs；
- end(queue end id) = start + nb_q / num_procs；
For example, if testpmd is configured to have 4 Tx and Rx queues, queues 0 and 1 will be used by the primary process and queues 2 and 3 will be used by the secondary process.
The number of queues should be a multiple of the number of processes. If not, redundant queues will exist after queues are allocated to processes. If RSS is enabled, packet loss occurs when traffic is sent to all processes at the same time. Some traffic goes to redundant queues and cannot be forwarded.
All the dev ops is supported in primary process. While secondary process is not permitted to allocate or release shared memory, so some ops are not supported as follows:
So, any command from testpmd which calls those APIs will not be supported in secondary process, like:
port config all rxq|txq|rxd|txd <value> port config <port_id> rx_offload xxx on/off port config <port_id> tx_offload xxx on/off
When secondary is running, port in primary is not permitted to be stopped. Reconfigure operation is only valid in primary.
Stats is supported, stats will not change when one quits and starts, as they share the same buffer to store the stats. Flow rules are maintained in process level: primary and secondary has its own flow list (but one flow list in HW). The two can see all the queues, so setting the flow rules for the other is OK. But in the testpmd primary process receiving or transmitting packets from the queue allocated for secondary process is not permitted, and same for secondary process.
Flow API and RSS are supported.