xref: /dpdk/doc/guides/sample_app_ug/ip_frag.rst (revision 8750576fb2a9a067ffbcce4bab6481f3bfa47097)
1..  SPDX-License-Identifier: BSD-3-Clause
2    Copyright(c) 2010-2014 Intel Corporation.
3
4IP Fragmentation Sample Application
5===================================
6
7The IPv4 Fragmentation application is a simple example of packet processing
8using the Data Plane Development Kit (DPDK).
9The application does L3 forwarding with IPv4 and IPv6 packet fragmentation.
10
11Overview
12--------
13
14The application demonstrates the use of zero-copy buffers for packet fragmentation.
15The initialization and run-time paths are very similar to those of the :doc:`l2_forward_real_virtual`.
16This guide highlights the differences between the two applications.
17
18There are three key differences from the L2 Forwarding sample application:
19
20*   The first difference is that the IP Fragmentation sample application makes use of indirect buffers.
21
22*   The second difference is that the forwarding decision is taken
23    based on information read from the input packet's IP header.
24
25*   The third difference is that the application differentiates between
26    IP and non-IP traffic by means of offload flags.
27
28The Longest Prefix Match (LPM for IPv4, LPM6 for IPv6) table
29is used to store/lookup an outgoing port number associated with that IP address.
30Any unmatched packets are forwarded to the originating port.
31
32By default, input frame sizes up to 9.5 KB are supported.
33Before forwarding, the input IP packet is fragmented
34to fit into the "standard" Ethernet* v2 MTU (1500 bytes).
35
36Compiling the Application
37-------------------------
38
39To compile the sample application, see :doc:`compiling`.
40
41The application is located in the ``ip_fragmentation`` sub-directory.
42
43Running the Application
44-----------------------
45
46The LPM object is created and loaded with the pre-configured entries read from
47global l3fwd_ipv4_route_array and l3fwd_ipv6_route_array tables.
48For each input packet, the packet forwarding decision
49(that is, the identification of the output interface for the packet) is taken as a result of LPM lookup.
50If the IP packet size is greater than the default output MTU,
51then the input packet is fragmented and several fragments are sent via the output interface.
52
53Application usage:
54
55.. code-block:: console
56
57    ./<build_dir>/examples/dpdk-ip_fragmentation [EAL options] -- -p PORTMASK [-q NQ]
58
59where:
60
61*   -p PORTMASK is a hexadecimal bitmask of ports to configure
62
63*   -q NQ: Maximum number of queues per lcore (default is 1)
64
65To run the example in linux environment with 2 lcores (2,4) over 2 ports(0,2) with 1 RX queue per lcore:
66
67.. code-block:: console
68
69    ./<build_dir>/examples/dpdk-ip_fragmentation -l 2,4 -n 3 -- -p 5
70    EAL: coremask set to 14
71    EAL: Detected lcore 0 on socket 0
72    EAL: Detected lcore 1 on socket 1
73    EAL: Detected lcore 2 on socket 0
74    EAL: Detected lcore 3 on socket 1
75    EAL: Detected lcore 4 on socket 0
76    ...
77
78    Initializing port 0 on lcore 2... Address:00:1B:21:76:FA:2C, rxq=0 txq=2,0 txq=4,1
79    done: Link Up - speed 10000 Mbps - full-duplex
80    Skipping disabled port 1
81    Initializing port 2 on lcore 4... Address:00:1B:21:5C:FF:54, rxq=0 txq=2,0 txq=4,1
82    done: Link Up - speed 10000 Mbps - full-duplex
83    Skipping disabled port 3IP_FRAG: Socket 0: adding route 100.10.0.0/16 (port 0)
84    IP_FRAG: Socket 0: adding route 100.20.0.0/16 (port 1)
85    ...
86    IP_FRAG: Socket 0: adding route 0101:0101:0101:0101:0101:0101:0101:0101/48 (port 0)
87    IP_FRAG: Socket 0: adding route 0201:0101:0101:0101:0101:0101:0101:0101/48 (port 1)
88    ...
89    IP_FRAG: entering main loop on lcore 4
90    IP_FRAG: -- lcoreid=4 portid=2
91    IP_FRAG: entering main loop on lcore 2
92    IP_FRAG: -- lcoreid=2 portid=0
93
94To run the example in linux environment with 1 lcore (4) over 2 ports(0,2) with 2 RX queues per lcore:
95
96.. code-block:: console
97
98    ./<build_dir>/examples/dpdk-ip_fragmentation -l 4 -n 3 -- -p 5 -q 2
99
100To test the application, flows should be set up in the flow generator that match the values in the
101l3fwd_ipv4_route_array and/or l3fwd_ipv6_route_array table.
102
103The default l3fwd_ipv4_route_array table is:
104
105.. literalinclude:: ../../../examples/ip_fragmentation/main.c
106    :language: c
107    :start-after: Default l3fwd_ipv4_route_array table. 8<
108    :end-before: >8 End of default l3fwd_ipv4_route_array table
109
110The default l3fwd_ipv6_route_array table is:
111
112.. literalinclude:: ../../../examples/ip_fragmentation/main.c
113    :language: c
114    :start-after: Default l3fwd_ipv6_route_array table. 8<
115    :end-before: >8 End of default l3fwd_ipv6_route_array table.
116
117For example, for the input IPv4 packet with destination address: 100.10.1.1 and packet length 9198 bytes,
118seven IPv4 packets will be sent out from port #0 to the destination address 100.10.1.1:
119six of those packets will have length 1500 bytes and one packet will have length 318 bytes.
120IP Fragmentation sample application provides basic NUMA support
121in that all the memory structures are allocated on all sockets that have active lcores on them.
122
123
124Refer to the *DPDK Getting Started Guide* for general information on running applications
125and the Environment Abstraction Layer (EAL) options.
126