1.. SPDX-License-Identifier: BSD-3-Clause 2 Copyright(c) 2010-2015 Intel Corporation. 3 4.. _Hash_Library: 5 6Hash Library 7============ 8 9The DPDK provides a Hash Library for creating hash table for fast lookup. 10The hash table is a data structure optimized for searching through a set of entries that are each identified by a unique key. 11For increased performance the DPDK Hash requires that all the keys have the same number of bytes which is set at the hash creation time. 12 13Hash API Overview 14----------------- 15 16The main configuration parameters for the hash are: 17 18* Total number of hash entries 19 20* Size of the key in bytes 21 22The hash also allows the configuration of some low-level implementation related parameters such as: 23 24* Hash function to translate the key into a bucket index 25 26The main methods exported by the hash are: 27 28* Add entry with key: The key is provided as input. If a new entry is successfully added to the hash for the specified key, 29 or there is already an entry in the hash for the specified key, then the position of the entry is returned. 30 If the operation was not successful, for example due to lack of free entries in the hash, then a negative value is returned; 31 32* Delete entry with key: The key is provided as input. If an entry with the specified key is found in the hash, 33 then the entry is removed from the hash and the position where the entry was found in the hash is returned. 34 If no entry with the specified key exists in the hash, then a negative value is returned 35 36* Lookup for entry with key: The key is provided as input. If an entry with the specified key is found in the hash (lookup hit), 37 then the position of the entry is returned, otherwise (lookup miss) a negative value is returned. 38 39Apart from these method explained above, the API allows the user three more options: 40 41* Add / lookup / delete with key and precomputed hash: Both the key and its precomputed hash are provided as input. This allows 42 the user to perform these operations faster, as hash is already computed. 43 44* Add / lookup with key and data: A pair of key-value is provided as input. This allows the user to store 45 not only the key, but also data which may be either a 8-byte integer or a pointer to external data (if data size is more than 8 bytes). 46 47* Combination of the two options above: User can provide key, precomputed hash and data. 48 49Also, the API contains a method to allow the user to look up entries in bursts, achieving higher performance 50than looking up individual entries, as the function prefetches next entries at the time it is operating 51with the first ones, which reduces significantly the impact of the necessary memory accesses. 52Notice that this method uses a pipeline of 8 entries (4 stages of 2 entries), so it is highly recommended 53to use at least 8 entries per burst. 54 55The actual data associated with each key can be either managed by the user using a separate table that 56mirrors the hash in terms of number of entries and position of each entry, 57as shown in the Flow Classification use case describes in the following sections, 58or stored in the hash table itself. 59 60The example hash tables in the L2/L3 Forwarding sample applications defines which port to forward a packet to based on a packet flow identified by the five-tuple lookup. 61However, this table could also be used for more sophisticated features and provide many other functions and actions that could be performed on the packets and flows. 62 63Multi-process support 64--------------------- 65 66The hash library can be used in a multi-process environment, minding that only lookups are thread-safe. 67The only function that can only be used in single-process mode is rte_hash_set_cmp_func(), which sets up 68a custom compare function, which is assigned to a function pointer (therefore, it is not supported in 69multi-process mode). 70 71Implementation Details 72---------------------- 73 74The hash table has two main tables: 75 76* First table is an array of entries which is further divided into buckets, 77 with the same number of consecutive array entries in each bucket. Each entry contains the computed primary 78 and secondary hashes of a given key (explained below), and an index to the second table. 79 80* The second table is an array of all the keys stored in the hash table and its data associated to each key. 81 82The hash library uses the cuckoo hash method to resolve collisions. 83For any input key, there are two possible buckets (primary and secondary/alternative location) 84where that key can be stored in the hash, therefore only the entries within those bucket need to be examined 85when the key is looked up. 86The lookup speed is achieved by reducing the number of entries to be scanned from the total 87number of hash entries down to the number of entries in the two hash buckets, 88as opposed to the basic method of linearly scanning all the entries in the array. 89The hash uses a hash function (configurable) to translate the input key into a 4-byte key signature. 90The bucket index is the key signature modulo the number of hash buckets. 91 92Once the buckets are identified, the scope of the hash add, 93delete and lookup operations is reduced to the entries in those buckets (it is very likely that entries are in the primary bucket). 94 95To speed up the search logic within the bucket, each hash entry stores the 4-byte key signature together with the full key for each hash entry. 96For large key sizes, comparing the input key against a key from the bucket can take significantly more time than 97comparing the 4-byte signature of the input key against the signature of a key from the bucket. 98Therefore, the signature comparison is done first and the full key comparison done only when the signatures matches. 99The full key comparison is still necessary, as two input keys from the same bucket can still potentially have the same 4-byte hash signature, 100although this event is relatively rare for hash functions providing good uniform distributions for the set of input keys. 101 102Example of lookup: 103 104First of all, the primary bucket is identified and entry is likely to be stored there. 105If signature was stored there, we compare its key against the one provided and return the position 106where it was stored and/or the data associated to that key if there is a match. 107If signature is not in the primary bucket, the secondary bucket is looked up, where same procedure 108is carried out. If there is no match there either, key is considered not to be in the table. 109 110Example of addition: 111 112Like lookup, the primary and secondary buckets are identified. If there is an empty slot in 113the primary bucket, primary and secondary signatures are stored in that slot, key and data (if any) are added to 114the second table and an index to the position in the second table is stored in the slot of the first table. 115If there is no space in the primary bucket, one of the entries on that bucket is pushed to its alternative location, 116and the key to be added is inserted in its position. 117To know where the alternative bucket of the evicted entry is, the secondary signature is looked up and alternative bucket index 118is calculated from doing the modulo, as seen above. If there is room in the alternative bucket, the evicted entry 119is stored in it. If not, same process is repeated (one of the entries gets pushed) until a non full bucket is found. 120Notice that despite all the entry movement in the first table, the second table is not touched, which would impact 121greatly in performance. 122 123In the very unlikely event that table enters in a loop where same entries are being evicted indefinitely, 124key is considered not able to be stored. 125With random keys, this method allows the user to get around 90% of the table utilization, without 126having to drop any stored entry (LRU) or allocate more memory (extended buckets). 127 128Entry distribution in hash table 129-------------------------------- 130 131As mentioned above, Cuckoo hash implementation pushes elements out of their bucket, 132if there is a new entry to be added which primary location coincides with their current bucket, 133being pushed to their alternative location. 134Therefore, as user adds more entries to the hash table, distribution of the hash values 135in the buckets will change, being most of them in their primary location and a few in 136their secondary location, which the later will increase, as table gets busier. 137This information is quite useful, as performance may be lower as more entries 138are evicted to their secondary location. 139 140See the tables below showing example entry distribution as table utilization increases. 141 142.. _table_hash_lib_1: 143 144.. table:: Entry distribution measured with an example table with 1024 random entries using jhash algorithm 145 146 +--------------+-----------------------+-------------------------+ 147 | % Table used | % In Primary location | % In Secondary location | 148 +==============+=======================+=========================+ 149 | 25 | 100 | 0 | 150 +--------------+-----------------------+-------------------------+ 151 | 50 | 96.1 | 3.9 | 152 +--------------+-----------------------+-------------------------+ 153 | 75 | 88.2 | 11.8 | 154 +--------------+-----------------------+-------------------------+ 155 | 80 | 86.3 | 13.7 | 156 +--------------+-----------------------+-------------------------+ 157 | 85 | 83.1 | 16.9 | 158 +--------------+-----------------------+-------------------------+ 159 | 90 | 77.3 | 22.7 | 160 +--------------+-----------------------+-------------------------+ 161 | 95.8 | 64.5 | 35.5 | 162 +--------------+-----------------------+-------------------------+ 163 164| 165 166.. _table_hash_lib_2: 167 168.. table:: Entry distribution measured with an example table with 1 million random entries using jhash algorithm 169 170 +--------------+-----------------------+-------------------------+ 171 | % Table used | % In Primary location | % In Secondary location | 172 +==============+=======================+=========================+ 173 | 50 | 96 | 4 | 174 +--------------+-----------------------+-------------------------+ 175 | 75 | 86.9 | 13.1 | 176 +--------------+-----------------------+-------------------------+ 177 | 80 | 83.9 | 16.1 | 178 +--------------+-----------------------+-------------------------+ 179 | 85 | 80.1 | 19.9 | 180 +--------------+-----------------------+-------------------------+ 181 | 90 | 74.8 | 25.2 | 182 +--------------+-----------------------+-------------------------+ 183 | 94.5 | 67.4 | 32.6 | 184 +--------------+-----------------------+-------------------------+ 185 186.. note:: 187 188 Last values on the tables above are the average maximum table 189 utilization with random keys and using Jenkins hash function. 190 191Use Case: Flow Classification 192----------------------------- 193 194Flow classification is used to map each input packet to the connection/flow it belongs to. 195This operation is necessary as the processing of each input packet is usually done in the context of their connection, 196so the same set of operations is applied to all the packets from the same flow. 197 198Applications using flow classification typically have a flow table to manage, with each separate flow having an entry associated with it in this table. 199The size of the flow table entry is application specific, with typical values of 4, 16, 32 or 64 bytes. 200 201Each application using flow classification typically has a mechanism defined to uniquely identify a flow based on 202a number of fields read from the input packet that make up the flow key. 203One example is to use the DiffServ 5-tuple made up of the following fields of the IP and transport layer packet headers: 204Source IP Address, Destination IP Address, Protocol, Source Port, Destination Port. 205 206The DPDK hash provides a generic method to implement an application specific flow classification mechanism. 207Given a flow table implemented as an array, the application should create a hash object with the same number of entries as the flow table and 208with the hash key size set to the number of bytes in the selected flow key. 209 210The flow table operations on the application side are described below: 211 212* Add flow: Add the flow key to hash. 213 If the returned position is valid, use it to access the flow entry in the flow table for adding a new flow or 214 updating the information associated with an existing flow. 215 Otherwise, the flow addition failed, for example due to lack of free entries for storing new flows. 216 217* Delete flow: Delete the flow key from the hash. If the returned position is valid, 218 use it to access the flow entry in the flow table to invalidate the information associated with the flow. 219 220* Lookup flow: Lookup for the flow key in the hash. 221 If the returned position is valid (flow lookup hit), use the returned position to access the flow entry in the flow table. 222 Otherwise (flow lookup miss) there is no flow registered for the current packet. 223 224References 225---------- 226 227* Donald E. Knuth, The Art of Computer Programming, Volume 3: Sorting and Searching (2nd Edition), 1998, Addison-Wesley Professional 228