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 22* An extra flag used to describe additional settings, for example the multithreading mode of operation (as will be described later) 23 24The hash also allows the configuration of some low-level implementation related parameters such as: 25 26* Hash function to translate the key into a bucket index 27 28The main methods exported by the hash are: 29 30* Add entry with key: The key is provided as input. If a new entry is successfully added to the hash for the specified key, 31 or there is already an entry in the hash for the specified key, then the position of the entry is returned. 32 If the operation was not successful, for example due to lack of free entries in the hash, then a negative value is returned; 33 34* Delete entry with key: The key is provided as input. If an entry with the specified key is found in the hash, 35 then the entry is removed from the hash and the position where the entry was found in the hash is returned. 36 If no entry with the specified key exists in the hash, then a negative value is returned 37 38* 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), 39 then the position of the entry is returned, otherwise (lookup miss) a negative value is returned. 40 41Apart from these method explained above, the API allows the user three more options: 42 43* Add / lookup / delete with key and precomputed hash: Both the key and its precomputed hash are provided as input. This allows 44 the user to perform these operations faster, as hash is already computed. 45 46* Add / lookup with key and data: A pair of key-value is provided as input. This allows the user to store 47 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). 48 49* Combination of the two options above: User can provide key, precomputed hash and data. 50 51Also, the API contains a method to allow the user to look up entries in bursts, achieving higher performance 52than looking up individual entries, as the function prefetches next entries at the time it is operating 53with the first ones, which reduces significantly the impact of the necessary memory accesses. 54 55 56The actual data associated with each key can be either managed by the user using a separate table that 57mirrors the hash in terms of number of entries and position of each entry, 58as shown in the Flow Classification use case describes in the following sections, 59or stored in the hash table itself. 60 61The 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. 62However, 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. 63 64Multi-process support 65--------------------- 66 67The hash library can be used in a multi-process environment. 68The only function that can only be used in single-process mode is rte_hash_set_cmp_func(), which sets up 69a custom compare function, which is assigned to a function pointer (therefore, it is not supported in 70multi-process mode). 71 72 73Multi-thread support 74--------------------- 75 76The hash library supports multithreading, and the user specifies the needed mode of operation at the creation time of the hash table 77by appropriately setting the flag. In all modes of operation lookups are thread-safe meaning lookups can be called from multiple 78threads concurrently. 79 80For concurrent writes, and concurrent reads and writes the following flag values define the corresponding modes of operation: 81 82* If the multi-writer flag (RTE_HASH_EXTRA_FLAGS_MULTI_WRITER_ADD) is set, multiple threads writing to the table is allowed. 83 Key add, delete, and table reset are protected from other writer threads. With only this flag set, readers are not protected from ongoing writes. 84 85* If the read/write concurrency (RTE_HASH_EXTRA_FLAGS_RW_CONCURRENCY) is set, multithread read/write operation is safe 86 (i.e., no need to stop the readers from accessing the hash table until writers finish their updates. Reads and writes can operate table concurrently). 87 88* In addition to these two flag values, if the transactional memory flag (RTE_HASH_EXTRA_FLAGS_TRANS_MEM_SUPPORT) is also set, 89 hardware transactional memory will be used to guarantee the thread safety as long as it is supported by the hardware (for example the Intel® TSX support). 90 91If the platform supports Intel® TSX, it is advised to set the transactional memory flag, as this will speed up concurrent table operations. 92Otherwise concurrent operations will be slower because of the overhead associated with the software locking mechanisms. 93 94Implementation Details 95---------------------- 96 97The hash table has two main tables: 98 99* First table is an array of entries which is further divided into buckets, 100 with the same number of consecutive array entries in each bucket. Each entry contains the computed primary 101 and secondary hashes of a given key (explained below), and an index to the second table. 102 103* The second table is an array of all the keys stored in the hash table and its data associated to each key. 104 105The hash library uses the cuckoo hash method to resolve collisions. 106For any input key, there are two possible buckets (primary and secondary/alternative location) 107where that key can be stored in the hash, therefore only the entries within those bucket need to be examined 108when the key is looked up. 109The lookup speed is achieved by reducing the number of entries to be scanned from the total 110number of hash entries down to the number of entries in the two hash buckets, 111as opposed to the basic method of linearly scanning all the entries in the array. 112The hash uses a hash function (configurable) to translate the input key into a 4-byte key signature. 113The bucket index is the key signature modulo the number of hash buckets. 114 115Once the buckets are identified, the scope of the hash add, 116delete and lookup operations is reduced to the entries in those buckets (it is very likely that entries are in the primary bucket). 117 118To 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. 119For large key sizes, comparing the input key against a key from the bucket can take significantly more time than 120comparing the 4-byte signature of the input key against the signature of a key from the bucket. 121Therefore, the signature comparison is done first and the full key comparison done only when the signatures matches. 122The full key comparison is still necessary, as two input keys from the same bucket can still potentially have the same 4-byte hash signature, 123although this event is relatively rare for hash functions providing good uniform distributions for the set of input keys. 124 125Example of lookup: 126 127First of all, the primary bucket is identified and entry is likely to be stored there. 128If signature was stored there, we compare its key against the one provided and return the position 129where it was stored and/or the data associated to that key if there is a match. 130If signature is not in the primary bucket, the secondary bucket is looked up, where same procedure 131is carried out. If there is no match there either, key is considered not to be in the table. 132 133Example of addition: 134 135Like lookup, the primary and secondary buckets are identified. If there is an empty slot in 136the primary bucket, primary and secondary signatures are stored in that slot, key and data (if any) are added to 137the second table and an index to the position in the second table is stored in the slot of the first table. 138If there is no space in the primary bucket, one of the entries on that bucket is pushed to its alternative location, 139and the key to be added is inserted in its position. 140To know where the alternative bucket of the evicted entry is, the secondary signature is looked up and alternative bucket index 141is calculated from doing the modulo, as seen above. If there is room in the alternative bucket, the evicted entry 142is stored in it. If not, same process is repeated (one of the entries gets pushed) until a non full bucket is found. 143Notice that despite all the entry movement in the first table, the second table is not touched, which would impact 144greatly in performance. 145 146In the very unlikely event that table enters in a loop where same entries are being evicted indefinitely, 147key is considered not able to be stored. 148With random keys, this method allows the user to get around 90% of the table utilization, without 149having to drop any stored entry (LRU) or allocate more memory (extended buckets). 150 151Entry distribution in hash table 152-------------------------------- 153 154As mentioned above, Cuckoo hash implementation pushes elements out of their bucket, 155if there is a new entry to be added which primary location coincides with their current bucket, 156being pushed to their alternative location. 157Therefore, as user adds more entries to the hash table, distribution of the hash values 158in the buckets will change, being most of them in their primary location and a few in 159their secondary location, which the later will increase, as table gets busier. 160This information is quite useful, as performance may be lower as more entries 161are evicted to their secondary location. 162 163See the tables below showing example entry distribution as table utilization increases. 164 165.. _table_hash_lib_1: 166 167.. table:: Entry distribution measured with an example table with 1024 random entries using jhash algorithm 168 169 +--------------+-----------------------+-------------------------+ 170 | % Table used | % In Primary location | % In Secondary location | 171 +==============+=======================+=========================+ 172 | 25 | 100 | 0 | 173 +--------------+-----------------------+-------------------------+ 174 | 50 | 96.1 | 3.9 | 175 +--------------+-----------------------+-------------------------+ 176 | 75 | 88.2 | 11.8 | 177 +--------------+-----------------------+-------------------------+ 178 | 80 | 86.3 | 13.7 | 179 +--------------+-----------------------+-------------------------+ 180 | 85 | 83.1 | 16.9 | 181 +--------------+-----------------------+-------------------------+ 182 | 90 | 77.3 | 22.7 | 183 +--------------+-----------------------+-------------------------+ 184 | 95.8 | 64.5 | 35.5 | 185 +--------------+-----------------------+-------------------------+ 186 187| 188 189.. _table_hash_lib_2: 190 191.. table:: Entry distribution measured with an example table with 1 million random entries using jhash algorithm 192 193 +--------------+-----------------------+-------------------------+ 194 | % Table used | % In Primary location | % In Secondary location | 195 +==============+=======================+=========================+ 196 | 50 | 96 | 4 | 197 +--------------+-----------------------+-------------------------+ 198 | 75 | 86.9 | 13.1 | 199 +--------------+-----------------------+-------------------------+ 200 | 80 | 83.9 | 16.1 | 201 +--------------+-----------------------+-------------------------+ 202 | 85 | 80.1 | 19.9 | 203 +--------------+-----------------------+-------------------------+ 204 | 90 | 74.8 | 25.2 | 205 +--------------+-----------------------+-------------------------+ 206 | 94.5 | 67.4 | 32.6 | 207 +--------------+-----------------------+-------------------------+ 208 209.. note:: 210 211 Last values on the tables above are the average maximum table 212 utilization with random keys and using Jenkins hash function. 213 214Use Case: Flow Classification 215----------------------------- 216 217Flow classification is used to map each input packet to the connection/flow it belongs to. 218This operation is necessary as the processing of each input packet is usually done in the context of their connection, 219so the same set of operations is applied to all the packets from the same flow. 220 221Applications using flow classification typically have a flow table to manage, with each separate flow having an entry associated with it in this table. 222The size of the flow table entry is application specific, with typical values of 4, 16, 32 or 64 bytes. 223 224Each application using flow classification typically has a mechanism defined to uniquely identify a flow based on 225a number of fields read from the input packet that make up the flow key. 226One example is to use the DiffServ 5-tuple made up of the following fields of the IP and transport layer packet headers: 227Source IP Address, Destination IP Address, Protocol, Source Port, Destination Port. 228 229The DPDK hash provides a generic method to implement an application specific flow classification mechanism. 230Given 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 231with the hash key size set to the number of bytes in the selected flow key. 232 233The flow table operations on the application side are described below: 234 235* Add flow: Add the flow key to hash. 236 If the returned position is valid, use it to access the flow entry in the flow table for adding a new flow or 237 updating the information associated with an existing flow. 238 Otherwise, the flow addition failed, for example due to lack of free entries for storing new flows. 239 240* Delete flow: Delete the flow key from the hash. If the returned position is valid, 241 use it to access the flow entry in the flow table to invalidate the information associated with the flow. 242 243* Lookup flow: Lookup for the flow key in the hash. 244 If the returned position is valid (flow lookup hit), use the returned position to access the flow entry in the flow table. 245 Otherwise (flow lookup miss) there is no flow registered for the current packet. 246 247References 248---------- 249 250* Donald E. Knuth, The Art of Computer Programming, Volume 3: Sorting and Searching (2nd Edition), 1998, Addison-Wesley Professional 251