One embodiment of an invention especially relates to computer and communications systems, especially network routers and switches; and more particularly, one embodiment of an invention relates to generating and merging lookup results to apply multiple features.
The communications industry is rapidly changing to adjust to emerging technologies and ever increasing customer demand. This customer demand for new applications and increased performance of existing applications is driving communications network and system providers to employ networks and systems having greater speed and capacity (e.g., greater bandwidth). In trying to achieve these goals, a common approach taken by many communications providers is to use packet switching technology. Increasingly, public and private communications networks are being built and expanded using various packet technologies, such as Internet Protocol (IP).
A network device, such as a switch or router, typically receives, processes, and forwards or discards a packet based on one or more criteria, including the type of protocol used by the packet, addresses of the packet (e.g., source, destination, group), and type or quality of service requested. Additionally, one or more security operations are typically performed on each packet. But before these operations can be performed, a packet classification operation must typically be performed on the packet.
Packet classification as required for, inter alia, access control lists (ACLs) and forwarding decisions, is a demanding part of switch and router design. The packet classification of a received packet is increasingly becoming more difficult due to ever increasing packet rates and number of packet classifications. For example, ACLs require matching packets on a subset of fields of the packet flow label, with the semantics of a sequential search through the ACL rules. IP forwarding requires a longest prefix match.
Known approaches of packet classification include using custom application-specific integrated circuits (ASICs), custom circuitry, software or firmware controlled processors, and associative memories, including, but not limited to binary content-addressable memories (binary CAMs) and ternary content-addressable memories (ternary CAMs or TCAMs). Each entry of a binary CAM typically includes a value for matching against, while each TCAM entry typically includes a value and a mask. The associative memory compares a lookup word against all of the entries in parallel, and typically generates an indication of the highest priority entry that matches the lookup word. An entry matches the lookup word in a binary CAM if the lookup word and the entry value are identical, while an entry matches the lookup word in a TCAM if the lookup word and the entry value are identical in the bits that are not indicated by the mask as being irrelevant to the comparison operations.
Associative memories are very useful in performing packet classification operations. In performing a packet classification, it is not uncommon for multiple lookup operations to be performed in parallel or in series using multiple associative memories basically based on a same search key or variant thereof, as one lookup operation might be related to packet forwarding while another related to quality of service determination. Desired are new functionality, features, and mechanisms in associative memories to support packet classification and other applications.
Additionally, as with most any system, errors can occur. For example, array parity errors can occur in certain content-addressable memories as a result of failure-in-time errors which are typical of semiconductor devices. Additionally, communications and other errors can occur. Prior systems are known to detect certain errors and to signal that some error condition has occurred, but are typically lacking in providing enough information to identify and isolate the error. Desired is new functionality for performing error detection and identification.
One problem with performing packet classification is the rate at which it must be performed, especially when multiple features of a certain type are to be evaluated. A prior approach uses a series of lookups to evaluate an action to be taken for each of these features. This approach is too slow, so techniques, such as Binary Decision Diagram (BDD) and Order Dependent Merge (ODM), were used for combining these features so they can be evaluated in a single lookup operation. For example, if there are two ACLs A (having entries A1 and A2) and B (having entries B1 and B2, then ODM combines these original lists to produce one of two cross-product equivalent ordered lists, each with four entries: A1B1, A1B2, A2B1, and A2B2; or A1B1, A2B1, A1B2, and A2B2. These four entries can then be programmed into an associative memory and an indication of a corresponding action to be taken placed in an adjunct memory. Lookup operations can then be performed on the associative and adjunct memories to identify a corresponding action to use for a particular packet being processed. There are also variants of ODM and BDD which may filter out the entries which are unnecessary as their values will never allow them to be matched. However, one problem with these approaches is that there can be an explosion of entries generated by these algorithms. A typical worst case would be to multiply the number of items in each feature by each other. Thus, two features of one hundred items each can generate one thousand entries, and if a third feature is considered which also has one hundred items, one million entries could be generated. Desired is a new mechanism for efficiently performing lookup operations which may reduce the number of entries required.
A known approach of identifying traffic flows for the purpose of prioritizing packets uses CAMs to identify and “remember” traffic flows allowing a network switch or router to identify packets belonging to that flow, at wire speed, without processor intervention. In one approach, learning new flows is automatic. Once a flow is identified, the system software assigns the proper priority to the newly identified flow. In each of the cases where learning is necessary (i.e., adding a new connection), the next free address of the device is read out so the system software can keep track of where the new additions are being placed. This way, the system software can efficiently remove these entries when they are no longer active. If aging is not used, the system software would need to keep track of the locations of every entry, and when a session ends, remove the corresponding entries. This is not a real-time issue, so software can provide adequate performance. Additionally, it is possible, even desirable to store timestamp information in the device to facilitate aging and purging of inactive flow identifiers.
For a purpose and context different from prioritizing packets, it is desirable to collect statistics about traffic flows (also referred to as “netflows”). These statistics can provide the metering base for real-time and post-processing applications including network traffic accounting, usage-based network billing, network planning, network monitoring, outbound marketing, and data mining capabilities for both service provider and enterprise customers. While this approach may work well for systems dealing with a relatively small amount of traffic with thousands of flows, this approach is not very scalable to systems handling larger amounts of data and flows as the collection of data on the raw flows generally produces too much unneeded data and requires a heavy burden on systems to collect all the information, if possible. Desired is a new mechanism for collecting accounting and other data.
Disclosed are, inter alia, methods, apparatus, data structures, computer-readable media, mechanisms, and means for merging lookup results, such as from one or more associative memory banks and/or memory devices. One embodiment includes multiple associative memories or associative memory banks configured to substantially simultaneously generate a plurality of lookup results based on a lookup value. Multiple memories are each configured to generate a corresponding result based on the lookup result generated by its corresponding associative memory or associative memory bank. A combiner is configured to receive and merge these corresponding results generated substantially simultaneously in order to identify the merged lookup result.
The appended claims set forth the features of the invention with particularity. The invention, together with its advantages, may be best understood from the following detailed description taken in conjunction with the accompanying drawings of which:
Methods and apparatus are disclosed for generating and merging lookup results to apply multiple features, for generating accounting or other data based on that indicated in an access control list or other specification, for defining and using associative memory entries with force no-hit and priority indications of particular use in implementing policy maps in communication devices, and for performing lookup operations using associative memories, including, but not limited to modifying search keys within an associative memory based on modification mappings, forcing a no-hit condition in response to a highest-priority matching entry including a force no-hit indication, selecting among various associative memory blocks or sets or banks of associative memory entries in determining a lookup result, and detecting and propagating error conditions.
Embodiments described herein include various elements and limitations, with no one element or limitation contemplated as being a critical element or limitation. Each of the claims individually recites an aspect of the invention in its entirety. Moreover, some embodiments described may include, but are not limited to, inter alia, systems, networks, integrated circuit chips, embedded processors, ASICs, methods, and computer-readable medium containing instructions. One or multiple systems, devices, components, etc. may comprise one or more embodiments, which may include some elements or limitations of a claim being performed by the same or different systems, devices, components, etc. The embodiments described hereinafter embody various aspects and configurations within the scope and spirit of the invention, with the figures illustrating exemplary and non-limiting configurations.
As used herein, the term “packet” refers to packets of all types or any other units of information or data, including, but not limited to, fixed length cells and variable length packets, each of which may or may not be divisible into smaller packets or cells. The term “packet” as used herein also refers to both the packet itself or a packet indication, such as, but not limited to all or part of a packet or packet header, a data structure value, pointer or index, or any other part or identification of a packet. Moreover, these packets may contain one or more types of information, including, but not limited to, voice, data, video, and audio information. The term “item” is used generically herein to refer to a packet or any other unit or piece of information or data, a device, component, element, or any other entity. The phrases “processing a packet” and “packet processing” typically refer to performing some steps or actions based on the packet contents (e.g., packet header or other fields), and such steps or action may or may not include modifying, storing, dropping, and/or forwarding the packet and/or associated data.
The term “system” is used generically herein to describe any number of components, elements, sub-systems, devices, packet switch elements, packet switches, routers, networks, computer and/or communication devices or mechanisms, or combinations of components thereof. The term “computer” is used generically herein to describe any number of computers, including, but not limited to personal computers, embedded processing elements and systems, control logic, ASICs, chips, workstations, mainframes, etc. The term “processing element” is used generically herein to describe any type of processing mechanism or device, such as a processor, ASIC, field programmable gate array, computer, etc. The term “device” is used generically herein to describe any type of mechanism, including a computer or system or component thereof. The terms “task” and “process” are used generically herein to describe any type of running program, including, but not limited to a computer process, task, thread, executing application, operating system, user process, device driver, native code, machine or other language, etc., and can be interactive and/or non-interactive, executing locally and/or remotely, executing in foreground and/or background, executing in the user and/or operating system address spaces, a routine of a library and/or standalone application, and is not limited to any particular memory partitioning technique. The steps, connections, and processing of signals and information illustrated in the figures, including, but not limited to any block and flow diagrams and message sequence charts, may be performed in the same or in a different serial or parallel ordering and/or by different components and/or processes, threads, etc., and/or over different connections and be combined with other functions in other embodiments in keeping within the scope and spirit of the invention. Furthermore, the term “identify” is used generically to describe any manner or mechanism for directly or indirectly ascertaining something, which may include, but is not limited to receiving, retrieving from memory, determining, defining, calculating, generating, etc.
Moreover, the terms “network” and “communications mechanism” are used generically herein to describe one or more networks, communications mediums or communications systems, including, but not limited to the Internet, private or public telephone, cellular, wireless, satellite, cable, local area, metropolitan area and/or wide area networks, a cable, electrical connection, bus, etc., and internal communications mechanisms such as message passing, interprocess communications, shared memory, etc. The term “message” is used generically herein to describe a piece of information which may or may not be, but is typically communicated via one or more communication mechanisms of any type.
The term “storage mechanism” includes any type of memory, storage device or other mechanism for maintaining instructions or data in any format. “Computer-readable medium” is an extensible term including any memory, storage device, storage mechanism, etc., tangibly embodying computer-readable instructions and/or data. The term “memory” includes any random access memory (RAM), read only memory (ROM), flash memory, integrated circuits, and/or other memory components or elements. The term “storage device” includes any solid state storage media, disk drives, diskettes, networked services, tape drives, and other storage devices. Memories and storage devices may store computer-executable instructions to be executed by a processing element and/or control logic, and data which is manipulated by a processing element and/or control logic. The term “data structure” is an extensible term referring to any data element, variable, data structure, database, and/or one or more organizational schemes that can be applied to data to facilitate interpreting the data or performing operations on it, such as, but not limited to memory locations or devices, sets, queues, trees, heaps, lists, linked lists, arrays, tables, pointers, etc. A data structure is typically maintained in a storage mechanism. The terms “pointer” and “link” are used generically herein to identify some mechanism for referencing or identifying another element, component, or other entity, and these may include, but are not limited to a reference to a memory or other storage mechanism or location therein, an index in a data structure, a value, etc. The term “associative memory” is an extensible term, and refers to all types of known or future developed associative memories, including, but not limited to binary and ternary content addressable memories, hash tables, TRIE and other data structures, etc. Additionally, the term “associative memory unit” may include, but is not limited to one or more associative memory devices or parts thereof, including, but not limited to regions, segments, banks, pages, blocks, sets of entries, etc.
The term “one embodiment” is used herein to reference a particular embodiment, wherein each reference to “one embodiment” may refer to a different embodiment, and the use of the term repeatedly herein in describing associated features, elements and/or limitations does not establish a cumulative set of associated features, elements and/or limitations that each and every embodiment must include, although an embodiment typically may include all these features, elements and/or limitations. In addition, the phrase “means for xxx” typically includes computer-readable medium containing computer-executable instructions for performing xxx.
In addition, the terms “first,” “second,” etc. are typically used herein to denote different units (e.g., a first element, a second element). The use of these terms herein does not necessarily connote an ordering such as one unit or event occurring or coming before another, but rather provides a mechanism to distinguish between particular units. Additionally, the use of a singular tense of a noun is non-limiting, with its use typically including one or more of the particular thing rather than just one (e.g., the use of the word “memory” typically refers to one or more memories without having to specify “memory or memories,” or “one or more memories” or “at least one memory”, etc.). Moreover, the phrases “based on x” and “in response to x” are used to indicate a minimum set of items x from which something is derived or caused, wherein “x” is extensible and does not necessarily describe a complete list of items on which the operation is performed, etc. Additionally, the phrase “coupled to” is used to indicate some level of direct or indirect connection between two elements or devices, with the coupling device or devices modifying or not modifying the coupled signal or communicated information. The term “subset” is used to indicate a group of all or less than all of the elements of a set. The term “subtree” is used to indicate all or less than all of a tree. Moreover, the term “or” is used herein to identify a selection of one or more, including all, of the conjunctive items.
Methods and apparatus are disclosed for defining and using associative memory entries with force no-hit and priority indications of particular use in implementing policy maps in communication devices. In one embodiment, a set of entries is determined based on a policy map with a force no-hit indication being associated with one or more of the entries. Additionally, programmable priority indications may be associated with one or more of the entries, or with the associative memory devices, associative memory banks, etc. The force no-hit indications are often used in response to identified deny instructions in an access control list or other policy map. A lookup operation is then performed on these associative memory entries, with highest matching result or results identified based on the programmed and/or implicit priority level associated with the entries, or with the associative memory devices, associative memory banks, etc.
Methods and apparatus are disclosed for performing lookup operations using associative memories, including, but not limited to modifying search keys within an associative memory based on modification mappings, forcing a no-hit condition in response to a highest-priority matching entry including a force no-hit indication, selecting among various associative memory blocks or sets or banks of associative memory entries in determining a lookup result, and detecting and propagating error conditions. In one embodiment, each block retrieves a modification mapping from a local memory and modifies a received search key based on the mapping and received modification data. In one embodiment, each of the associative memory entries includes a field for indicating that a successful match on the entry should or should not force a no-hit result. In one embodiment, an indication of which associative memory sets or banks or entries to use in a particular lookup operation is retrieved from a memory.
One embodiment performs error detection and handling by identifying, handling and communication errors, which may include, but is not limited to array parity errors in associative memory entries and communications errors such as protocol errors and interface errors on input ports. Array parity errors can occur as a result of failure-in-time errors which are typical of semiconductor devices. One embodiment includes a mechanism to scan associative memory entries in background, and to identify any detected errors back to a control processor for re-writing or updating the flawed entry. In one embodiment, certain identified errors or received error conditions are of a fatal nature in which no processing should be performed. For example, in one embodiment, a fatal error causes an abort condition. In response, the device stops an in-progress lookup operation and just forwards error and possibly no-hit signals. Typically, these signals are generated at the time the in-progress lookup operation would have generated its result had it not been aborted so as to maintain timing among devices in a system including the associative memory.
In one embodiment, including cascaded or connected associative memory devices, error status messages indicating any error type and its corresponding source are propagated to indicate the error status to the next device and/or a control processor. In addition, the communicated signal may indicate and generate an abort condition in the receiving device. In one embodiment, the receiving device does not perform its next operation or the received instruction, or it may abort its current operation or instruction. Moreover, the receiving device may or may not delay a time amount corresponding to that which its processing would have required in performing or completing the operation or instruction so as to possibly maintain the timing of a transactional sequence of operations.
One embodiment generates accounting or other data based on that indicated in an access control list or other specification, and typically using associative memory entries in one or more associative memory banks and/or memory devices. One embodiment identifies an access control list including multiple access control list entries, with a subset of these access control list entries identifying accounting requests. Accounting mechanisms, such as, but not limited to counters or data structures, are associated with each of said access control list entries in the subset of access control list entries identifying accounting requests. An item is identified. A particular one of the accounting mechanisms corresponding to the item is identified and updated. In one embodiment, the item corresponds to one or more fields of a received packet. In one embodiment, the item includes at least one autonomous system number, said at least one autonomous system number identify a set of communication devices under a single administrative authority. In one embodiment, at least one of the accounting mechanisms is associated with at least two different access control list entries in the subset of access control list entries identifying accounting requests.
One embodiment merges lookup results, such as from one or more associative memory banks and/or memory devices. One embodiment identifies an access control list including multiple access control list entries. A first set of access control list entries corresponding to a first feature of the access control list entries and a second set of access control list entries corresponding to a second feature of the access control list entries are identified. A first associative memory bank is programmed with the first associative memory entries and a second associative memory bank is programmed with the second associative memory entries, with the first associative memory entries having a higher lookup precedence than the second associative memory entries. A lookup value is then identified, such as that based on a packet or other item. Lookup operations are then typically performed substantially simultaneously on the first and second sets of associative memory entries to generate multiple lookup results, with these results typically being identified directly, or via a lookup operation in an adjunct memory or other storage mechanism. These lookup results are then combined to generate a merged lookup result.
In one embodiment, packets 101 are received by packet processor 105. In addition to other operations (e.g., packet routing, security, etc.), packet processor 105 typically generates one or more items, including, but not limited to one or more packet flow identifiers based on one or more fields of one or more of the received packets 101 and possibly from information stored in data structures or acquired from other sources. Packet processor 105 typically generates a lookup value 103 which is provided to control logic 110 for providing control and data information (e.g., lookup words, modification data, profile IDs, etc.) to associative memory or memories 115, which perform lookup operations and generate one or more results 117. In one embodiment, a result 117 is used is by memory 120 to produce a result 125. Control logic 110 then relays result 107, based on result 117 and/or result 125, to packet processor 105. In response, one or more of the received packets are manipulated and forwarded by packet processor 105 as indicated by packets 109. Note, results 117, 125 and 107 may include indications of error conditions.
Moreover, in one embodiment, a synchronization bit field is included in messages 141-145 sent between devices 140 and 146-148, with the value being set or changed at predetermined periodic intervals such that each device 140, 146-148 expects the change. One embodiment uses a single synchronization bit, and if this bit is set in the request or input data 141-145 to a device 146-148, then the device 146-148 will set this bit in the corresponding reply or output data 143-145. For example, in one embodiment, control processor or logic 140 sets the sync bit in its request data 141 periodically, say once in every eight requests. Control processor or logic 140 also monitors the sync bit in the reply data 145. If any kind of error altered the request-reply association (or transaction timing) between the control processor or logic 140 and the associative memories 146-148, then control processor or logic 140 can detect it and recover from that error (by flushing the pipeline, etc.)
In this manner, devices, especially those as part of a transactional sequence, can synchronize themselves with each other. Resynchronization of devices may become important, for example, should an error condition occur, such as an undetected parity error in a communicated instruction signal (e.g., the number of parity errors exceed the error detection mechanism). There is a possibility that a parity error in an instruction goes undetected and that completely changes the transaction timing. Also, there could be other types of “unknown” errors that can put the control processor or logic and the associative memory chain out of synchronization.
Various embodiments of system 180 may include more or less elements. The operation of system 180 is typically controlled by processing element 181 using memory 182 and storage devices 183 to perform one or more tasks or processes, such as programming and performing lookup operations using associative memory or memories 184. Memory 182 is one type of computer-readable medium, and typically comprises random access memory (RAM), read only memory (ROM), flash memory, integrated circuits, and/or other memory components. Memory 182 typically stores computer-executable instructions to be executed by processing element 181 and/or data which is manipulated by processing element 181 for implementing functionality in accordance with one embodiment of the invention. Storage devices 183 are another type of computer-readable medium, and typically comprise solid state storage media, disk drives, diskettes, networked services, tape drives, and other storage devices. Storage devices 183 typically store computer-executable instructions to be executed by processing element 181 and/or data which is manipulated by processing element 181 for implementing functionality in accordance with one embodiment of the invention.
In one embodiment, processing element 181 provides control and data information (e.g., lookup words, modification data, profile IDs, etc.) to associative memory or memories 184, which perform lookup operations to generate lookup results and possibly error indications, which are received and used by processing element 181 and/or communicated to other devices via interface 185.
In one embodiment, input data 201, which may include, but is not limited to search keys and modification data, is received by associative memory 200 and distributed to associative memory blocks 218-219, and possibly forwarded to other downstream associative memories in a cascaded configuration. In addition, input control information 202, which may include, but is not limited to profile IDs (e.g., a value), instructions, programming information, is received by control logic 210, and possibly forwarded to other downstream associative memories in a cascaded configuration. In addition, in one embodiment, previous stage lookup results and/or error indications are received from previous stage associative memories in a cascaded configuration or from other devices by control logic 210. Note, in one embodiment, input data 201, input control 202, previous stage results and errors 203, and/or portions thereof are communicated directly to associative memory blocks 218-219 and/or output selectors 231-232.
Control logic 210 possibly processes and/or forwards the received information via block control signals 211-212 to associative memory blocks 218-219 and via selector control signals and previous stage results 215 (which typically includes the received profile ID) to output selectors 231-232. In addition, control logic 210 may generate error signals 216 based on a detected error in the received information or in response to received error condition indications. Note, in one embodiment, control logic 210 merely splits or regenerates a portion of or the entire received input control 202 and optional previous stage results and errors 203 signals as selector control signals and previous stage results signals 215 and/or error signals 216. In addition, control logic 210 could initiate an abort operation wherein a lookup operation will not occur because of a detected or received notification of an error condition.
In one embodiment, control logic 210 identifies data representing which associative memory blocks 218-219 to enable, which associative memory blocks 218-219 each output selector 231-232 should consider in determining its lookup result, and/or modification mappings each associative memory block 218-219 should use in modifying an input search key. In one embodiment, this data is retrieved, based on received input control information 202 (e.g., a profile ID or other indication), from one or more memories, data structures, and/or other storage mechanisms. This information is then communicated as appropriate to associative memory blocks 218-219 via block control signals 211-212, and/or output selectors 231-232 via selector control signals and previous stage results signals 215.
In one embodiment, associative memory blocks 218-219 each receive a search key and possibly modification data via signal 201, and possibly control information via block control signals 211-212. Each enabled associative memory block 218-219 then performs a lookup operation based on the received search key, which may include generating a lookup word by modifying certain portions of the search key based on received modification data and/or modification mappings. Each associative memory block 218-219 typically generates a result 228-229 which are each communicated to each of the output selectors 231-232. In one embodiment, each associative memory block 218-219 that is not enabled generates a no-hit signal as its corresponding result 228-229. In one embodiment, output selectors 231-232 receive an indication of the associative memory blocks 218-219 that is not enabled.
Output selectors 231 evaluate associative memory results 228-229 to produce results 240. In one embodiment, each output selector has a corresponding identified static or dynamic subset of the associate memory results 228-229 to evaluate in determining results 240. In one embodiment, an identification of this corresponding subset is provided to each output selector 231-232 via selector control signals 215. In one embodiment, each of the output selectors 231-232 receives a profile ID via selector control signals 215 and performs a memory lookup operation based on the received profile ID to retrieve an indication of the particular associate memory results 228-229 to evaluate in determining results 240.
Moreover, in one embodiment, results 240 are exported over one or more output buses 240, each typically connected to a different set of one or more pins of a chip of the associative memory. In one embodiment, the number of output buses used and their connectivity to outputs selectors 231-232 are static, while in one embodiment the number of output buses used and their connectivity to outputs selectors 231-232 are configurable, for example, at initialization or on a per or multiple lookup basis. In one embodiment, an output bus indication is received by an output selector 231-232, which uses the output bus indication to determine which output bus or buses to use. For example, this determination could include, but is not limited to a direct interpretation of the received output bus indication, performing a memory read operation based on the received output bus indication, etc. In one embodiment, an output selector 231-232 performs a memory access operation based on a profile ID to determine which output bus or buses to use for a particular lookup operation. Thus, depending on the configuration, a single or multiple output buses/pins can selectively be used to communicate results 240, with this decision possibly being made based on the tradeoff of receiving multiple results simultaneously versus the number of pins required.
Associative memory 200 provides many powerful capabilities for simultaneously producing one or more results 240. For example, in one embodiment, based on a received profile ID, control logic 210 identifies which of the one or more associative memory blocks 218-219 to enable and then enables them, and provides the profile ID to output selectors 231 for selecting a lookup result among the multiple associative memory blocks 218-219. Each of the associative memory blocks 218-219 may receive/identify a modification mapping based on the profile ID, with this modification mapping possibly being unique to itself. This modification mapping can then be used in connection with received modification data to change a portion of a received search key to produce the actual lookup word to be used in the lookup operation. Also, certain entries may be programmed with force no-hit indications to generate a no-hit result for the corresponding associative memory block 218-219 should a corresponding entry be identified as the highest priority entry matching the lookup word. Each of these enabled associative memory block 218-219 typically generate a result (e.g., no-hit, hit with highest priority matching entry or location thereof identified) which is typically communicated to each of the output selectors 231-232. Note, in one embodiment, the results are only communicated to the particular output selectors 231-232 which are to consider the particular result in selecting their respective highest priority result received from associative memory blocks 218-219 and possibly other lookup results from previous stage associative memories. Additionally, in certain configurations, multiple associative memories 200 are cascaded or coupled in other methods so that results from one or more stages may depend on previous stage results, such that a lookup can be programmed to be performed across multiple associative memories 200. These and other constructs provided by associative memory 200 and configurations thereof provide powerful programmable lookup search capabilities and result selection mechanisms using one or more stages of associative memories 200, each including N associative memories blocks 218-219 and M output selectors 231-232. In one embodiment, the actual values of N and M may vary among associative memories 200.
Returning to
Output selector array 330 is programmed with an output selector ID 335 identifying which output selector, such as, but not limited to output selectors 231-232 (
Modification mapping array 340 is programmed with a modification mapping 345 for each tuple (profile ID 341, output selector ID 342). Thus, based on a profile ID 341 received over via input control 302 (
Moreover, one embodiment of associative memory block 400 includes a memory 413 for storing configuration information, which may allow an associative memory block 400 to retrieve the information from memory 413 rather than receive it from another source. For example, in one embodiment, modification mapping data (e.g., modification mapping 345 of
Additionally, in one embodiment, a search key 402, modification data 403, modification mapping 404, an enable signal 405, a global mask enable signal 406, and a global mask select signal 407 are received. In response to performing a lookup operation and/or detecting an error condition, such as a parity fault in one of the associative memory entries, result and error indications 411 are generated. In one embodiment, associative memory entries are checked for parity errors in background. The use of these signals and information in one embodiment are further described in relation to
Turning to
As determined in process block 478, if at least one match has been identified, then processing proceeds to process block 480, otherwise to process block 490, wherein a result with a no hit indication is generated and processing proceeds to process block 484. Otherwise, as determined in process block 480, if the highest priority matching entry includes a force no hit indication, then processing proceeds to process block 490, wherein a result with a no hit indication is generated and processing proceeds to process block 484. Otherwise, in process block 482, a result indicating a hit (i.e., successful match) with the highest priority matching entry identified is generated.
In process block 484, the result is communicated to at least the identified output selector or selectors. In one embodiment, the output selector to which to communicate the result is identified by output selector ID 335 (
Returning to
Moreover, in one embodiment, selector control signal 501 including an enable indication, the enable indication including an enabled or not enabled value, such that in when a not enable value is received, output selector 500 is not enabled and does not select among results from blocks 1−N 503 or optional previous stage results 502. In one embodiment, when not enabled, output selector 500 generates a result signal 515 indicting a no hit, not enabled, or some other predetermined or floating value.
Additionally, in one embodiment, result 515 is communicated over a fixed output bus, which may or may not be multiplexed with other results 515 generated by other output selectors 500. In one embodiment, the associative memory may include one or more output buses, each typically connected to a single pin of a chip of the associative memory, with the selection of a particular output bus possibly being hardwired or configurable, with the configuration possibly being on a per lookup basis, such as that determined from a received value or configuration information retrieved from a memory (e.g., based on the current profile ID.) In such a configuration, control logic 510 (or other mechanism) typically selects which output bus (and the timing of sending result 515) to use for a particular or all results 515.
A process used in one embodiment for receiving and selecting a highest-priority associative memory entry, if any, is illustrated in
By using the optional priority indications, entries 621 and 622 can be stored in different associative memories and/or associative memory banks, etc., to possibly consider in determining where to store the entries in order to efficiently use the space available for the entries. By associating a priority level with each entry, entries within a same associative memory and/or associate memory block, etc. can have different priority levels, which gives great flexibility in programming and managing the entries and space available for storing the entries.
Processing of the flow diagram illustrated in
Based on an identified packet, autonomous system numbers, and/or other information, a lookup value 833 is identified.
Based on lookup value 833, a lookup operation is performed in associative memory entries 834 in one or more associative memory banks and/or one or more associative memories to generate a counter indication 835. The corresponding counting mechanism within counters and decoder/control logic 836 is updated. Counter values 837 are typically communicated via any communication mechanism and/or technique to packet processor 832 or another device to be forwarded or processed.
Based on an identified packet, autonomous system numbers, and/or other information, a lookup value 842 is identified.
Based on lookup value 842, a lookup operation is performed in associative memory entries 843 in one or more associative memory banks and/or one or more associative memories to produce a lookup result 844, which is then used to perform a lookup operation in adjunct memory 845 generate a counter indication 846, and the corresponding counting mechanism within counters and decoder/control logic 847 is updated. In one embodiment, adjunct memory 845 stores counter indications for corresponding locations of access control list entries programmed in associative memory 843, and some of these counter indications may be the same value such that a same counting mechanism is updated for different matching access control list entries. Counter values 848 are typically communicated via any communication mechanism and/or technique to packet processor 841 or another device to be forwarded or processed.
Based on an identified packet, autonomous system numbers, and/or other information, a lookup value 903 is identified.
Based on lookup value 903, a lookup operation is performed in associative memory entries 904 (e.g., access control list, security, quality of service, accounting entries) in multiple associative memory banks and/or one or more associative memories to generate a results 905, based on which, memories 906 generate results 907. Combiner mechanism 910 merges results 907 to produce one or more merged results 911, which are typically used by packet processor 902 in the processing of packets. In one embodiment, combiner mechanism 910 includes a processing element responsive to computer-readable medium (e.g., software, firmware, etc.), custom hardware (e.g., circuits, ASICs, etc.) and/or via any other means or mechanism. In one embodiment, a merged result 911 includes a counter indication which is used by counters and decoder/control logic 912 to update a value. The accumulated accounting values 913 are typically communicated to packet processor 902 or another device.
For example, one embodiment includes four associative memory banks for supporting one to four features. An associative memory lookup operation is performed in parallel on the four banks and then in the adjunct memories (SRAMs), which indicate the action, type of entry (e.g., ACL, QoS, Accounting), and precedence for combiner mechanism. The combiner mechanism merges the results to get the final merged result. A miss in an ACL lookup in a bank is treated as a permit with lowest precedence. If in more than one bank there is a hit with same specified precedence in the retrieved adjunct memory entry, the precedence used by the combiner mechanism is determined based on the implied or specified precedence of the associative memory bank. If there is a miss in all the banks, default result is used from global registers. A similar merge operation is performed for the QoS and accounting lookup results.
As shown, result value 965 includes a result type 965A, an action or counter indication 965B, and a precedence indication 965C. In one embodiment, result value 965 is programmed in the adjunct memories. One embodiment uses all, less than all, or none of fields 965A-965C.
As shown, merged result value 967 includes a result type 967A and an action or counter indication 967B. One embodiment uses all, less than all, or none of fields 967A-967B.
In one embodiment, the merge result of a Quality of Service (QoS) lookup operation is illustrated in security combiner logic 972, and is based on the results of a previously merged security lookup operation and up to four substantially simultaneous (or not) lookup operations with differing precedence indicated in columns 972A-970E, with the corresponding merged result shown in column 972F.
In one embodiment, the merge result of an accounting lookup operation is illustrated in accounting combiner logic 972, and is based on the results of a previously merged security lookup operation and up to four substantially simultaneous (or not) lookup operations with differing precedence indicated in columns 974A-974E, with the corresponding merged result shown possibly identifying a counter to be updated in column 972F.
In process block 986, the QoS lookup value is identified. In process block 987, lookup operations are performed based on the QoS lookup value in multiple associative memory banks and one or more adjunct memories to identify multiple QoS results, which, in process block 988, are merged along with the previously determined merged security result to identify the merged QoS result.
In process block 989, the accounting lookup value is identified. In process block 990, lookup operations are performed based on the accounting lookup value in multiple associative memory banks and one or more adjunct memories to identify multiple accounting results, which, in process block 991, are merged along with the previously determined merged security result to identify the merged accounting result. Also, an identified counter or other accounting mechanism is updated. Processing is complete as indicated by process block 992.
In view of the many possible embodiments to which the principles of our invention may be applied, it will be appreciated that the embodiments and aspects thereof described herein with respect to the drawings/figures are only illustrative and should not be taken as limiting the scope of the invention. For example and as would be apparent to one skilled in the art, many of the process block operations can be re-ordered to be performed before, after, or substantially concurrent with other operations. Also, many different forms of data structures could be used in various embodiments. The invention as described herein contemplates all such embodiments as may come within the scope of the following claims and equivalents thereof.
This is a continuation of nonprovisional application Ser. No. 10/630,174, filed Jul. 29, 2003 now U.S. Pat. No. 7,177,978, which is a continuation-in-part of application Ser. No. 10/215,719, filed Aug. 10, 2002 now U.S. Pat. No. 7,103,708, with both of these applications being hereby incorporated by reference in their entirety.
Number | Name | Date | Kind |
---|---|---|---|
3648254 | Beausoleil | Mar 1972 | A |
4296475 | Nederlof et al. | Oct 1981 | A |
4791606 | Threewitt et al. | Dec 1988 | A |
4996666 | Duluk, Jr. | Feb 1991 | A |
5088032 | Bosack | Feb 1992 | A |
5319763 | Ho et al. | Jun 1994 | A |
5339076 | Jiang | Aug 1994 | A |
5383146 | Threewitt | Jan 1995 | A |
5404482 | Stamm et al. | Apr 1995 | A |
5428565 | Shaw | Jun 1995 | A |
5440715 | Wyland | Aug 1995 | A |
5450351 | Heddes | Sep 1995 | A |
5481540 | Huang | Jan 1996 | A |
5515370 | Rau | May 1996 | A |
5684954 | Kaiserswerth et al. | Nov 1997 | A |
5740171 | Mazzola et al. | Apr 1998 | A |
5748905 | Hauser et al. | May 1998 | A |
5802567 | Liu et al. | Sep 1998 | A |
5841874 | Kempke et al. | Nov 1998 | A |
5842040 | Hughes et al. | Nov 1998 | A |
5852569 | Srinivasan et al. | Dec 1998 | A |
5898689 | Kumar et al. | Apr 1999 | A |
5920886 | Feldmeier | Jul 1999 | A |
5930359 | Kempke et al. | Jul 1999 | A |
5956336 | Loschke et al. | Sep 1999 | A |
5978885 | Clark, II | Nov 1999 | A |
6000008 | Simcoe | Dec 1999 | A |
6041389 | Rao | Mar 2000 | A |
6047369 | Colwell et al. | Apr 2000 | A |
6061368 | Hitzelberger | May 2000 | A |
6069573 | Clark, II et al. | May 2000 | A |
6071440 | Wang et al. | Jun 2000 | A |
6091725 | Cheriton et al. | Jul 2000 | A |
6097724 | Kartalopoulos | Aug 2000 | A |
6134135 | Andersson | Oct 2000 | A |
6137707 | Srinivasan et al. | Oct 2000 | A |
6141738 | Munter et al. | Oct 2000 | A |
6148364 | Srinivasan et al. | Nov 2000 | A |
6154384 | Nataraj et al. | Nov 2000 | A |
6175513 | Khanna | Jan 2001 | B1 |
6181698 | Hariguchi | Jan 2001 | B1 |
6199140 | Srinivasan et al. | Mar 2001 | B1 |
6219748 | Srinivasan et al. | Apr 2001 | B1 |
6236658 | Essbaum et al. | May 2001 | B1 |
6237061 | Srinivasan et al. | May 2001 | B1 |
6240003 | McElroy | May 2001 | B1 |
6240485 | Srinivasan et al. | May 2001 | B1 |
6243667 | Kerr et al. | Jun 2001 | B1 |
6246601 | Pereira | Jun 2001 | B1 |
6285378 | Duluk, Jr. | Sep 2001 | B1 |
6289414 | Feldmeier et al. | Sep 2001 | B1 |
6295576 | Ogura et al. | Sep 2001 | B1 |
6307855 | Hariguchi | Oct 2001 | B1 |
6308219 | Hughes | Oct 2001 | B1 |
6317350 | Pereira et al. | Nov 2001 | B1 |
6374326 | Kansal et al. | Apr 2002 | B1 |
6377577 | Bechtolsheim et al. | Apr 2002 | B1 |
6389506 | Ross et al. | May 2002 | B1 |
6430190 | Essbaum et al. | Aug 2002 | B1 |
6434662 | Greene et al. | Aug 2002 | B1 |
6467019 | Washburn | Oct 2002 | B1 |
6526474 | Ross | Feb 2003 | B1 |
6535951 | Ross | Mar 2003 | B1 |
6546391 | Tsuruoka | Apr 2003 | B1 |
6597595 | Ichiriu et al. | Jul 2003 | B1 |
6606681 | Uzun | Aug 2003 | B1 |
6651096 | Gai et al. | Nov 2003 | B1 |
6658002 | Ross et al. | Dec 2003 | B1 |
6707692 | Hata | Mar 2004 | B2 |
6715029 | Trainin et al. | Mar 2004 | B1 |
6717946 | Hariguchi et al. | Apr 2004 | B1 |
6718326 | Uga et al. | Apr 2004 | B2 |
6738862 | Ross et al. | May 2004 | B1 |
6775737 | Warkhede et al. | Aug 2004 | B1 |
6928430 | Chien et al. | Aug 2005 | B1 |
7028136 | Priyadarshan et al. | Apr 2006 | B1 |
7065609 | Eatherton et al. | Jun 2006 | B2 |
7082492 | Pullela et al. | Jul 2006 | B2 |
7103708 | Eatherton et al. | Sep 2006 | B2 |
7177978 | Kanekar et al. | Feb 2007 | B2 |
20010038554 | Takata et al. | Nov 2001 | A1 |
20020073073 | Cheng | Jun 2002 | A1 |
20030005146 | Miller et al. | Jan 2003 | A1 |
20030223259 | Roth | Dec 2003 | A1 |
20030231631 | Pullela | Dec 2003 | A1 |
20040030802 | Eatherton et al. | Feb 2004 | A1 |
20040030803 | Eatherton et al. | Feb 2004 | A1 |
20040111556 | Torkelsson et al. | Jun 2004 | A1 |
Number | Date | Country | |
---|---|---|---|
20070002862 A1 | Jan 2007 | US |
Number | Date | Country | |
---|---|---|---|
Parent | 10630174 | Jul 2003 | US |
Child | 11497171 | US |
Number | Date | Country | |
---|---|---|---|
Parent | 10215719 | Aug 2002 | US |
Child | 10630174 | US |