摘要:
Novel data structures, methods and apparatus for finding the longest prefix match search when searching tables with variable length patterns or prefixes. To find the exact match or the best matching prefix, patterns have to be compared a bit at a time until the exact or first: match is found. This requires “n” number of comparisons or memory accesses to identify the closest matching pattern. The trees are built in such a way that the matching result is guaranteed to be a best match, whether it is an exact match or a longest prefix match. Using the trail of all the birds and associated prefix lengths enables determination of the correct prefix result from the trail. By construction, the search tree provides the best matching prefix at or after the first compare during walking of the trail or tree.
摘要:
Novel data structures, methods and apparatus for finding a full match between a search pattern and a pattern stored in a leaf of the search tree. A key is input, a hash function is performed on the key, a direct table (DT) is accessed, and a tree is walked through pattern search control blocks (PSCBS) until reaching a leaf. The search mechanism uses a set of data structures that can be located in a few registers and regular memory, and then used to build a Patricia tree structure that can be manipulated by a relatively simple hardware macro. Both keys and corresponding information needed for retrieval are stored in the Patricia tree structure. The hash function provides an n->n mapping of the bits of the key to the bits of the hash key. The data structure that is used to store the hash key and the related information in the tree is called a leaf. Each leaf corresponds to a single key that matches exactly with the input key. The leaf contains the key as well as additional information. The length of the leaf is programmable, as is the length of the key. The leaf is stored in random access memory and is implemented as a single memory entry. If the key is located in the direct table then it is called a direct leaf.
摘要:
The invention includes a bridge having n ports (n>1), each port being connected to a Token-Ring physical segment, each physical segment having one native Token-Ring workstation attached. The bridge to the workstations a single Token-Ring logical segment with a single Active Monitor and a single Ring Number. The invention includes a centralized medium access control (MAC) function inside a centralized processor instead of a MAC function implemented at each port of the bridge; the frame handling function, due to the fixed and limited configuration (same bridge Active Monitor seen by all connected stations), does not require a multi-port bridge function, but a simpler switch function between ports. Bridge clocking is also simplified, and a cost effective unshield twisted pair (UTP) retiming solution is presented.
摘要:
A method of operating a packet parser in a computing system includes providing a configurable packet pointer by the packet parser, the packet pointer configured to index a configurable number of atomic parsing elements, the atomic parsing elements having a configurable size, in a data stream received by the computing system for extraction, wherein the indexed atomic parsing elements are non-contiguous in the data stream; and receiving the extracted indexed atomic parsing elements from the data stream by the packet parser.
摘要:
The CRC for the CPS Header of an ATM AAL2 cell is generated by a CRC generator which uses the 8 bits of the CID field to generate partial 5 bits CRCs which are loaded in a first table. The 6 bits LI field and 5 bits UUI field are added to the partial 5 bits CRC to form 16 bits. The CRC generator uses the 2.sup.16 bits to generate a second CRC table. The CRC for a particular CPS header is generated by correlating bits in the CID field, LI field and UUI field with the two tables.
摘要:
A network switch apparatus, components for such an apparatus, and methods of operating such an apparatus in which data flow handling and flexibility is enhanced by the cooperation of a control point and a plurality of interface processors formed on a semiconductor substrate. The control point and interface processors together form a network processor capable of cooperating with other elements including an optional switching fabric device in executing instructions directing the flow of data in a network.
摘要:
An improved arbiter is described for arbitrating requests by a plurality of first data processing units for access to a plurality of second data processing units interconnected by a switching system of a type in which at any time each first unit can only access one second unit and each second unit can only be accessed by one first unit. The arbiter comprises a scheduler mechanism for repeatedly selecting access requests with a defined minimum probability of selecting a request for each first unit-second unit combination. Rearrangement storage means records requests selected by the scheduler mechanism. A rearranger is provided for repeatedly selecting a set of requests recorded in the rearrangement storage means, so that only one request per first unit and per second unit is selected, using a priority mechanism which increases the probability of selection with the length of time a request is stored in the rearrangement storage means. Finally, means are provided for communicating the grant of the selected set of requests to the switching system and for deleting the selected set of requests from the rearrangement storage means. In one embodiment, the arbiter is used for controlling switching paths in a packet data switch.
摘要:
A communication system comprises a plurality of nodes interconnected by links comprising a plurality of connections. The traffic between the nodes is set up by a reserved bandwidth service and/or a non reserved bandwidth service. The non reserved bandwidth service is controlled by a hop by hop backpressure mechanism. When the traffic entering a node exceeds a high threshold, the backpressure mechanism generates stop backpressure primitives in order to throttle the entering traffic. In case of congestion the mechanism is either able to selectively interuppt the connection contributing to the congestion without affecting the rest of the link traffic, or to globally stop all link traffic. Traffic can be resumed if traffic rates fall below the low threshold values.
摘要:
A hub featuring ports for attachment of stations to a LAN comprises concentration logic (14) for the handling of multiplexed incoming and outgoing Token-Ring and isochronous signal streams. The concentration logic comprises clock recovery logic (42) from incoming Token-Ring packet data stream (40), for regeneration of Differential Manchester encoded data on output (400), and recovering of Token-Ring clock (401). A cycle framing generator (43) receives a 125 us synchronization clock from the hub backplane (402), and the Token-Ring clock (401), and generates control signals (403) to each of the 10 ports. Each port is comprised of a port transmit interface (44), and a port receive interface (45). Data from a connected station is input (404) to port receive interface (45). Token-Ring packet Differential Manchester encoded data are output (406) to the next active port, specifically to its port transmit interface, along with a recovered strobe clock (405), while ISO data are output (407) to switch (46). The switch and other concentration logic receive a hub local clock (412). Isochronous traffic interchanges with the hub backplane through leads 410 and 411; between ports or between ports and the hub through leads 407 and 409. Data to a connected station is output (408) from port transmit interface (44). Differential Manchester encoded data are received (400) along with Token-Ring clock (401). Control signals are input (403). Isochronous data are received (409). Token-Ring packet Differential Manchester encoded data are finally output from the concentration logic (41).
摘要:
A data switching device, such as an ATM or Asynchronous Transfer Mode switch, includes a switching fabric with multiple input and output leads. The device also includes at least one input adapter for receiving data cells on each of a number of input ports and at least one output adapter for delivering data cells switched through the switching fabric to a target port in a set of output ports. Error and format checks are performed on incoming cells and counts are kept of the number of good cells and invalid cells received on a particular input port. To reduce hardware costs, the counts are kept in a random access memory which is shared among the input ports. Several storage locations are allocated to each input port to maintain the necessary counts.