1.. BSD LICENSE 2 Copyright(c) 2010-2014 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.. _Mbuf_Library: 32 33Mbuf Library 34============ 35 36The mbuf library provides the ability to allocate and free buffers (mbufs) 37that may be used by the DPDK application to store message buffers. 38The message buffers are stored in a mempool, using the :ref:`Mempool Library <Mempool_Library>`. 39 40A rte_mbuf struct can carry network packet buffers (type is RTE_MBUF_PKT) 41or generic control buffers (type is RTE_MBUF_CTRL). 42This can be extended to other types. 43The rte_mbuf is kept as small as possible (one cache line if possible). 44 45Design of Packet Buffers 46------------------------ 47 48For the storage of the packet data (including protocol headers), two approaches were considered: 49 50#. Embed metadata within a single memory buffer the structure followed by a fixed size area for the packet data. 51 52#. Use separate memory buffers for the metadata structure and for the packet data. 53 54The advantage of the first method is that it only needs one operation to allocate/free the whole memory representation of a packet. 55On the other hand, the second method is more flexible and allows 56the complete separation of the allocation of metadata structures from the allocation of packet data buffers. 57 58The first method was chosen for the DPDK. 59The metadata contains control information such as message type, length, 60pointer to the start of the data and a pointer for additional mbuf structures allowing buffer chaining. 61 62Message buffers that are used to carry network packets can handle buffer chaining 63where multiple buffers are required to hold the complete packet. 64This is the case for jumbo frames that are composed of many mbufs linked together through their pkt.next field. 65 66For a newly allocated mbuf, the area at which the data begins in the message buffer is 67RTE_PKTMBUF_HEADROOM bytes after the beginning of the buffer, which is cache aligned. 68Message buffers may be used to carry control information, packets, events, 69and so on between different entities in the system. 70Message buffers may also use their data pointers to point to other message buffer data sections or other structures. 71 72Figure 8 and Figure 9 show some of these scenarios. 73 74.. _pg_figure_8: 75 76**Figure 8. An mbuf with One Segment** 77 78.. image22_png has been replaced 79 80|mbuf1| 81 82.. _pg_figure_9: 83 84**Figure 9. An mbuf with Three Segments** 85 86.. image23_png has been replaced 87 88|mbuf2| 89 90The Buffer Manager implements a fairly standard set of buffer access functions to manipulate network packets. 91 92Buffers Stored in Memory Pools 93------------------------------ 94 95The Buffer Manager uses the :ref:`Mempool Library <Mempool_Library>` to allocate buffers. 96Therefore, it ensures that the packet header is interleaved optimally across the channels and ranks for L3 processing. 97An mbuf contains a field indicating the pool that it originated from. 98When calling rte_ctrlmbuf_free(m) or rte_pktmbuf_free(m), the mbuf returns to its original pool. 99 100Constructors 101------------ 102 103Packet and control mbuf constructors are provided by the API. 104The rte_pktmbuf_init() and rte_ctrlmbuf_init() functions initialize some fields in the mbuf structure that 105are not modified by the user once created (mbuf type, origin pool, buffer start address, and so on). 106This function is given as a callback function to the rte_mempool_create() function at pool creation time. 107 108Allocating and Freeing mbufs 109---------------------------- 110 111Allocating a new mbuf requires the user to specify the mempool from which the mbuf should be taken. 112For a packet mbuf, it contains one segment, with a length of 0. 113The pointer to data is initialized to have some bytes of headroom in the buffer (RTE_PKTMBUF_HEADROOM). 114For a control mbuf, it is initialized with data pointing to the beginning of the buffer and a length of zero. 115 116Freeing a mbuf means returning it into its original mempool. 117The content of an mbuf is not modified when it is stored in a pool (as a free mbuf). 118Fields initialized by the constructor do not need to be re-initialized at mbuf allocation. 119 120When freeing a packet mbuf that contains several segments, all of them are freed and returned to their original mempool. 121 122Manipulating mbufs 123------------------ 124 125This library provides some functions for manipulating the data in a packet mbuf. For instance: 126 127 * Get data length 128 129 * Get a pointer to the start of data 130 131 * Prepend data before data 132 133 * Append data after data 134 135 * Remove data at the beginning of the buffer (rte_pktmbuf_adj()) 136 137 * Remove data at the end of the buffer (rte_pktmbuf_trim()) Refer to the *DPDK API Reference* for details. 138 139Meta Information 140---------------- 141 142Some information is retrieved by the network driver and stored in an mbuf to make processing easier. 143For instance, the VLAN, the RSS hash result (see :ref:`Poll Mode Driver <Poll_Mode_Driver>`) 144and a flag indicating that the checksum was computed by hardware. 145 146An mbuf also contains the input port (where it comes from), and the number of segment mbufs in the chain. 147 148For chained buffers, only the first mbuf of the chain stores this meta information. 149 150Direct and Indirect Buffers 151--------------------------- 152 153A direct buffer is a buffer that is completely separate and self-contained. 154An indirect buffer behaves like a direct buffer but for the fact that the data pointer it contains points to data in another direct buffer. 155This is useful in situations where packets need to be duplicated or fragmented, 156since indirect buffers provide the means to reuse the same packet data across multiple buffers. 157 158A buffer becomes indirect when it is "attached" to a direct buffer using the rte_pktmbuf_attach() function. 159Each buffer has a reference counter field and whenever an indirect buffer is attached to the direct buffer, 160the reference counter on the direct buffer is incremented. 161Similarly, whenever the indirect buffer is detached, the reference counter on the direct buffer is decremented. 162If the resulting reference counter is equal to 0, the direct buffer is freed since it is no longer in use. 163 164There are a few things to remember when dealing with indirect buffers. 165First of all, it is not possible to attach an indirect buffer to another indirect buffer. 166Secondly, for a buffer to become indirect, its reference counter must be equal to 1, 167that is, it must not be already referenced by another indirect buffer. 168Finally, it is not possible to reattach an indirect buffer to the direct buffer (unless it is detached first). 169 170While the attach/detach operations can be invoked directly using the recommended rte_pktmbuf_attach() and rte_pktmbuf_detach() functions, 171it is suggested to use the higher-level rte_pktmbuf_clone() function, 172which takes care of the correct initialization of an indirect buffer and can clone buffers with multiple segments. 173 174Since indirect buffers are not supposed to actually hold any data, 175the memory pool for indirect buffers should be configured to indicate the reduced memory consumption. 176Examples of the initialization of a memory pool for indirect buffers (as well as use case examples for indirect buffers) 177can be found in several of the sample applications, for example, the IPv4 Multicast sample application. 178 179Debug 180----- 181 182In debug mode (CONFIG_RTE_MBUF_DEBUG is enabled), 183the functions of the mbuf library perform sanity checks before any operation (such as, buffer corruption, bad type, and so on). 184 185Use Cases 186--------- 187 188All networking application should use mbufs to transport network packets. 189 190.. |mbuf1| image:: img/mbuf1.svg 191 192.. |mbuf2| image:: img/mbuf2.svg 193
