Search Images Maps Play YouTube News Gmail Drive More »
Sign in
Screen reader users: click this link for accessible mode. Accessible mode has the same essential features but works better with your reader.

Patents

  1. Advanced Patent Search
Publication numberUS20080256455 A1
Publication typeApplication
Application numberUS 12/103,683
Publication date16 Oct 2008
Filing date15 Apr 2008
Priority date14 Dec 2001
Also published asUS7298754, US7298756, US7304987, US7352694, US7965624, US20080253294
Publication number103683, 12103683, US 2008/0256455 A1, US 2008/256455 A1, US 20080256455 A1, US 20080256455A1, US 2008256455 A1, US 2008256455A1, US-A1-20080256455, US-A1-2008256455, US2008/0256455A1, US2008/256455A1, US20080256455 A1, US20080256455A1, US2008256455 A1, US2008256455A1
InventorsAlberto Alessandro Della Ripa, Peter Benschop, Philip Michael Clovis, Peter Mark Bouvier, Steven Dean Michel, David Dvorman, Diego Escobar
Original AssigneeAlberto Alessandro Della Ripa, Peter Benschop, Philip Michael Clovis, Peter Mark Bouvier, Steven Dean Michel, David Dvorman, Diego Escobar
Export CitationBiBTeX, EndNote, RefMan
External Links: USPTO, USPTO Assignment, Espacenet
Method for Defining the Physical Configuration of a Communication System
US 20080256455 A1
Abstract
A method is provided for automatically generating code to define and control a system of connected hardware elements. The method comprises: accepting a system configuration macro with sub-macros for system elements, subsystem elements, and connections there between; accepting a plurality of tables with a plurality of system element behaviors, a plurality of subsystem element behaviors, and a plurality of connection options; defining the system of connected elements in response to selecting sub-macros; defining the physical links between the system elements and the behavior of the system and subsystem elements in response to populating the selected sub-macro parameters; expanding the selected sub-macros; generating executable code; and, accessing the tables in response to parameters in the executable code. Advantageously, the form and function of the system can be defined with programming, or writing application specific code.
Images(12)
Previous page
Next page
Claims(12)
1-25. (canceled)
26. A machine-readable medium having stored thereon instructions for automatically generating code to define and control a system of connected elements, the instructions comprising:
supplying a system configuration macro with sub-macros for system elements, subsystem elements embedded in the system elements, and connections there between;
supplying a plurality of tables with a plurality of system element behaviors, a plurality of subsystem element behaviors, and a plurality of connection options;
defining a system of connected elements in response to receiving instructions selecting sub-macros;
defining the physical links between the system elements and the behavior of the system and subsystem elements in response to populating the selected sub-macro parameters;
expanding the selected sub-macros;
automatically generating executable code;
accessing the tables in response to parameters in the executable code; and,
storing the tables in a tangible memory medium.
27. The instructions of claim 26 further comprising:
populating parameters in the selected sub-macros; and,
elaborating the executable code.
28. The instructions of claim 27 wherein elaborating the executable code includes differentiating the executable code into code sections corresponding to system and subsystem elements; and,
wherein storing the tables includes loading accessed tables into corresponding subsystem memories.
29. The instructions of claim 28 further comprising:
following the elaboration of the executable code, monitoring the system performance;
comparing the system performance to a standard;
in response to sub-standard system performance:
automatically repopulating selected parameters in the selected sub-macros;
instantiating new executable code; and,
storing accessed indexing tables in response to the repopulated parameters.
30. The instructions of claim 26 wherein supplying a system configuration macro with sub-macros includes supplying a system configuration macro with sub-macros for switch card and port card system elements, integrated circuit (chip) subsystem elements on the cards, and connections there between; and,
wherein automatically generating executable code includes automatically generating code to define and control a system of cards connected in a switch fabric network.
31. A graphical user interface (GUI) for defining the physical configuration of a network, the GUI comprising:
a first user interface for accessing a system configuration macro text file with selectable sub-macros for system elements, subsystem elements on the cards, and connections there between;
a second user interface for accessing tables defining the behavior of system elements, subsystem elements, and connection options;
a third user interface for defining a network of connected elements supplied in response to selecting sub-macros;
a fourth user interface for defining the physical links between the system elements and the behavior of the system and subsystem elements supplied in response to populating the selected sub-macro parameters; and,
a machine interface to supply code with parameters for loading tables into system element memories.
32. The GUI of claim 31 wherein the first user interface accesses a hierarchical set of fields, with subsystem elements being indentured to system elements; and,
wherein the machine interface supplies code for:
accessing a sub-table in response to indexing a table; and,
loading values from the sub-table in a subsystem memory.
33. The GUI of claim 31 wherein the first user interface accesses a switch fabric configuration macro for defining switch fabric cards, firmware processing integrated circuits (chips) on the cards, and the connections between the cards.
34. The GUI of claim 33 wherein the first user interface accesses sub-macros to hierarchically define the system level (slot number), the slot level (card type), card level (chip type), and chip level (link types and endpoint).
35. A method for dynamically maintaining optimum performance levels in a system, the method comprising:
following the elaboration of a network including connected system and subsystem elements, monitoring the network performance;
comparing the network performance to a standard;
in response to sub-standard network performance:
automatically repopulating selected parameters in selected sub-macros defining system elements, subsystem elements, and connections there between;
instantiating executable code; and,
loading accessed tables into system elements in response to the repopulated parameters, the tables defining system element behaviors, subsystem element behaviors, and connection options.
36. The method of claim 35 wherein monitoring the network performance includes monitoring a network of switch fabric cards, firmware processing integrated circuits (chips) on the cards, and connections between the cards.
Description
    RELATED APPLICATIONS
  • [0001]
    This application is a Continuation of a pending application entitled; METHOD FOR AUTOMATICALLY GENERATING CODE TO DEFINE A SYSTEM OF HARDWARE ELEMENTS, invented by Ripa et al., Ser. No. 10/400,380, filed Mar. 26, 2003; which is a Continuation of the following applications:
  • [0002]
    SYSTEM AND METHOD FOR SWITCHING VARIABLY SIZED INFORMATION GROUPS, invented by Yun et al., Ser. No. 10/023,266, filed Dec. 14, 2001, now abandoned;
  • [0003]
    SYSTEM AND METHOD FOR SIMULTANEOUS DEFICIT ROUND ROBIN PRIORITIZATION, invented by Yun et al., Ser. No. 10/022,673, filed Dec. 17, 2001, now U.S. Pat. No. 7,079,545;
  • [0004]
    SYSTEM AND METHOD FOR GRANTING ARBITRATED BIDS IN THE SWITCHING OF INFORMATION, invented by Yun et al., Ser. No. 10/029,581, filed Dec. 20, 2001, still pending;
  • [0005]
    SYSTEM AND METHOD FOR HIERARCHICAL SWITCHING, invented by Yun et al., Ser. No. 10/035,835, filed Dec. 24, 2001, now U.S. Pat. No. 7,020,131; and,
  • [0006]
    SYSTEM AND METHOD FOR TOLERATING DATA LINK FAULTS IN A PACKET COMMUNICATIONS SWITCH FABRIC, invented by Clovis et al., Ser. No. 10/373,139, filed Feb. 24, 2003, now U.S. Pat. No. 7,352,694. All the above-referenced applications are incorporated herein by reference.
  • BACKGROUND OF THE INVENTION
  • [0007]
    1. Field of the Invention
  • [0008]
    This invention generally relates to software programming and, more particularly, to a method for automating generating code that defines and controls a system of network-connected hardware elements.
  • [0009]
    2. Description of the Related Art
  • [0010]
    Conventionally, software can be generating by using macros to construct calls to a particular interface in the effort of promoting hardware independent software. However, the interfaces so constructed are not configuration independent. If the system hardware connections are modified, the macros must also be modified. This can be very difficult to for a chip vendor to manage as new customers decide to use a chipset of complex programmable devices, especially if the original design team has been redirected to other projects.
  • [0011]
    There is industry demand for integrated circuits (ICs) switching systems that can be easily scaled for Network Access, the Network Edge, or a Core Switch Router, at the high end. SCSA (Signal Computing System Architecture) defines a switch fabric to be the facility for connecting any two (or more) transmitting or receiving Service Providers.
  • [0012]
    FIG. 10 is a schematic block diagram of a general switch fabric system (prior art). Traffic enters the system on the ingress side (the left side of the figure). Traffic exits the system on the egress side (the right side of the figure). The port card ingress and egress structures are sometimes subdivided into channels and subchannels. The packet headers specify a required destination port card, channel, and subchannel. The system must synchronously route each ingress packet to the correct egress destination.
  • [0013]
    Packets are converted into frames by ingress traffic managers (iTMs). A frame is a logical unit of data, which is often a small piece of a much larger data set such as a file or image. The iTMs feed the ingress side of the fabric. The switch fabric might convert the frame format to a “native” format, and then on egress, convert the data back into the TM frame format before sending the data to the egress traffic managers (eTMs). If the frames are fixed size (for example: 53 bytes, 64 bytes, or 80 bytes), the frames are often called cells.
  • [0000]
    Protocol Specific vs. Protocol Agnostic
  • [0014]
    A switch fabric can be protocol specific or protocol agnostic. An example of a protocol specific switch fabric would be a system designed and optimized specifically for asynchronous transfer mode (ATM) traffic. Another example would be a switch that handles only TCP/IP traffic. The obvious disadvantage of a switch fabric hardwired to handle a specific protocol is lack of flexibility. Service providers want to sell services to a variety of customers in various industries. Protocols vary from industry to industry.
  • [0015]
    Even within one protocol type, there can be protocol upgrades. For example, TCP/IP, the machine language of Internet routers, now runs primarily “IPv4.” This protocol, successful as it has been, has a lack of available addresses, poor security features, and no “quality of service” (QoS) provisions. The next generation Internet protocol is “IPv6.” It provides solutions to these limitations.
  • [0016]
    A protocol agnostic switch fabric works equally well with all protocol types, however, the traffic manager must be responsible for recognizing specific protocol types. The disadvantage of a protocol agnostic switch fabric is that it may be more complicated, and perhaps slower than a switch fabric dedicated to a particular protocol.
  • [0000]
    Packet Striping vs. Single Link per Packet
  • [0017]
    The simplest way for a traffic manager to transmit a packet into a switch fabric is to transmit the packet serially along one line. Striping is a way of achieving higher bandwidth by transmitting a single packet across multiple ingress/egress lines. For example, a TM can transmit a packet into a switch fabric eight times as fast if the packet is sliced into eight pieces (stripes) by the TM, and conveyed into the fabric along 8 parallel lines simultaneously. The fabric captures the packet in memory, routes it to the required egress destination, and slices the packet into 8 parallel lines before transmitting the packet to the egress TM.
  • [0018]
    The upside to packet striping is the potential for lower latency. There are several negative aspects of packet striping:
  • [0019]
    if one of the links is damaged (1 of 8 in the example above), the entire channel is out of service, degrading fault tolerance; and,
  • [0020]
    the interface between the TM and switch fabric is more complicated. Circuitry must be used to slice the packet into stripes and reassemble it into packets.
  • [0000]
    Single-Cell Packet vs. Multi-Cell Packets
  • [0021]
    Many switch fabrics now deployed and passing revenue traffic, especially in wide area networks (WANs), use asynchronous transfer mode (ATM) packets. ATM packets are single-cell packets, with a fixed cell size of 53 bytes, 48 bytes of which is the information payload. The ATM specification evolved in the 1980s and early 1990s. It was then believed that variable length (multi-cell) packets would be too difficult to implement at rates needed for wire-speed traffic. The single-cell solution was a compromise that would work for voice data, video data, multimedia data, email data, file data, etc. With a fixed frame size, switch designs are simplified and ultimately faster.
  • [0022]
    However, files are better sent in large frames. Voice switching performs better with small bursts of data, corresponding to analog-to-digital converter time slots. Large frames and concomitant switching latencies can render a switch useless for most 2-way voice applications. Voice communications require low latency (time delay). Since ATM had to work for all network data, the small payload (voice) requirements prevailed at the expense of the large frame applications.
  • [0023]
    For large frames or packets it is much more efficient to establish a path from an ingress port card to the required egress destination, and then leave this path undisturbed until the packet is completed. With single cell packets, the ingress port card must bid for, accept grants, and then schedule each cell of a multi-cell packet.
  • [0024]
    There are also Frame Relay switch fabrics. Frame Relay is a single-cell protocol, albeit with frame size ranging from 6 to 4096 bytes. Such switch fabrics have been deployed and passing revenue traffic since the early 1990s. These switch fabrics have generally not been used for voice data because of the large latency. Only highly compressed voice traffic works well over frame relay. Voice-over-frame relay was added as an afterthought by network engineers. Frame relay excels at data communications such as local area network internetworking (LAN-to-LAN). Such communications are very high speed and bursty, with non-critical latency constraints.
  • [0000]
    Cut-Through vs. Store-and-Forward
  • [0025]
    The conventional method of switch fabric packet routing is called Store-and-Forward. In this method, the switch fabric accepts an input packet and buffers the packet on the ingress side of the fabric, making sure the packet was received intact, knowing the exact number of cells in the packet. The problem with the store-and-forward method is the added latency of buffering the packet. In Cut-through Packet Routing (CPR), a switch fabric is able to send the incoming packet cells to the correct egress port as soon as the destination address is known.
  • [0000]
    Memory-Based vs. Arbitrated Crossbar
  • [0026]
    A switch fabric can use memory-based crossbars or arbitrated crossbars. A memory-based crossbar is sometimes called a “shared memory switch.” Ingress packets flow from the port cards into a huge memory bank, which serve as the switch. From the memory bank, the destination address is determined from egress port destination information in the cell headers. The problem with these switch fabrics is that they become prohibitively expensive and large from all the required high-speed memory. Such fabrics cannot be used to reach terabit total switching speeds with technology available today.
  • [0027]
    It would be advantageous if new firmware executable code could be easily created for new customers by field application engineers (FAEs). Preferably, such code generation could occur without the need for delving into the finer details of the firmware, knowing the chipset register definitions, or modifying the source code. If the firmware generation process could be highly automated, customers could even learn how to deploy a wide variety of system configurations themselves, with little FAE support, after receiving training for the chipset and its firmware environment.
  • [0028]
    It would be advantageous to deploy a new system by running a graphical user interface (GUI) on a workstation, with the GUI prompting the user(s) through a process that automatically defines the form and function of the system, without the usual arduous (and expensive) programming. Using “drag and drop” mouse commands, pull down menus, and icons, the customer or FAE could create a completely new firmware build customized for a completely new hardware configuration.
  • [0029]
    It would be advantageous if the maintenance of the system firmware could be focused on the code generation system as well as the algorithmic code, without the drudgery of “starting from scratch” for each new system or customer.
  • [0030]
    A good example of a candidate chipset for employing the method described above is the Cyclone Switch Fabric. This set of four chips has thousands of registers and 100's of millions of logic gates. A huge and steep learning curve must be overcome for even the most seasoned firmware engineer to master the multiplicity of register settings, system configuration options, and complex algorithms used in the firmware. It would be advantageous of the Cyclone chipset firmware can developed for new systems with the help of automated tools.
  • [0031]
    It would be advantageous if the automatic system definition process could also be applied to a broader range communications chipsets. Other examples chipsets might include framers, media access controllers (MACs), PHY devices (used in the OSI Physical layer), Physical Medium Dependent (PMD) devices, framers, network processors, and traffic managers. All such devices have registers, which must be loaded to configure the devices and registers, which must be read to monitor the health and status of the devices. All such devices also must be connected together in a specific manner in a newly proposed system (the form of the system). And all such devices have a multiplicity of parameter selections (the function of the system). The system engineer, solutions architect, or FAE could use the automated firmware tools on customer visits to propose new solutions to customer needs.
  • SUMMARY OF THE INVENTION
  • [0032]
    A method is provided for automatically generating code to define and control a system of connected hardware elements. The method comprises: accepting a system configuration macro with sub-macros for system elements, subsystem elements, and connections there between; accepting a plurality of tables with a plurality of system element behaviors, a plurality of subsystem element behaviors, and a plurality of connection options; defining the system of connected elements in response to selecting sub-macros; defining the physical links between the system elements and the behavior of the system and subsystem elements in response to populating the selected sub-macro parameters; expanding the selected sub-macros; generating executable code; and, accessing the tables in response to parameters in the executable code. Alternately stated, the expansion of the sub-macros defines the form and the function of the network hardware, with programming.
  • [0033]
    Additional details of the present inventions methods are provided below.
  • BRIEF DESCRIPTION OF THE DRAWINGS
  • [0034]
    FIG. 1 a is a flowchart illustrating the present invention method for automatically generating code to define and control a system of connected hardware elements.
  • [0035]
    FIG. 1 b illustrates specifics relating to Steps 102 and 104 of FIG. 1 a.
  • [0036]
    FIG. 2 is a flowchart illustrating the present invention method for automatically generating code to define and control a switch fabric of connected cards.
  • [0037]
    FIG. 3 is a flowchart illustrating another aspect of the present invention method for automatically generating code to define and control a system of connected hardware elements.
  • [0038]
    FIG. 4 is a diagram illustrating the hierarchy of the system configuration file.
  • [0039]
    FIG. 5 is an illustration of code that is generated for each macro in the Systemconfiguration.cfg file.
  • [0040]
    FIG. 6 illustrates the relationship between configuration files and configuration tables.
  • [0041]
    FIG. 7 depicts an excerpt of a PQ chip configuration file.
  • [0042]
    FIG. 8 illustrates how the configuration file is indexed, to locate the tables that contain the values to program into the chip registers for a given chip configuration mode.
  • [0043]
    FIG. 9 is a flowchart illustrating the present invention method for autonomously discovering the configuration of system-connected elements.
  • [0044]
    FIG. 10 is a schematic block diagram of a general switch fabric system (prior art).
  • DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
  • [0045]
    FIG. 1 a is a flowchart illustrating the present invention method for automatically generating code to define and control a system of connected hardware elements. Although the method (and methods associated with FIGS. 2 and 3 below) is depicted as a sequence of numbered steps for clarity, no order should be inferred from the numbering unless explicitly stated. It should be understood that some of these steps may be skipped, performed in parallel, or performed without the requirement of maintaining a strict order of sequence. The method starts at Step 100.
  • [0046]
    Step 102 generates a system configuration macro. Step 104 generates tables defining a plurality of system element behaviors, a plurality of subsystem element behaviors, and a plurality of connection options.
  • [0047]
    FIG. 1 b illustrates specifics relating to Steps 102 and 104 of FIG. 1 a. In some aspects, Steps 102 and 104 are accomplished using a graphical user interface (GUI). Alternately, the steps are enabled by the manual entry of parameters.
  • [0048]
    Step 106 accepts the system configuration macro with sub-macros for defining system elements, subsystem elements, and connections there between. Step 108 selects sub-macros. Step 110 populates parameters in the selected sub-macros. Step 112 accepts the plurality of tables defining a plurality of system element behaviors, a plurality of subsystem element behaviors, and a plurality of connection options. Step 114 expands a plurality of sub-macros with populated parameters. In some aspects, Step 114 expands the parameters of the selected sub-macros. Step 116 generates executable table indexing code. For example, the code can be generated by means of a complier specific to the target processor. Step 118 elaborates the executable code. Step 120 accesses the tables in response to parameters in the executable code. The above-mentioned steps describe the steps that a user would perform to define, compile, and run the code.
  • [0049]
    In some aspects of the method, elaborating the executable code in Step 118 includes differentiating the executable code into code sections corresponding to system and subsystem elements. Then, accessing the tables in response to the parameters in the executable code in Step 120 includes loading accessed tables into corresponding subsystem element memories.
  • [0050]
    In other aspects, accepting the system configuration macro with sub-macros for defining system elements, subsystem elements, and the connections there between, in Step 106, includes accepting a hierarchical set of fields, with subsystem elements being indentured to system elements. Then, accessing the tables in response to the parameters in the executable code in Step 120 includes substeps. Step 120 a accesses a sub-table in response to indexing a table. Step 120 b loads values from the sub-table in chip registers.
  • [0051]
    In some aspects, expanding the plurality of sub-macros with populated parameters in Step 114 includes generating C++ language code. Then, generating executable code in Step 116 includes generating machine code.
  • [0052]
    To better illustrate the invention, the above-mentioned steps can be understood as a process of configuring a switch fabric of cards including firmware-enabled chips. Then, accepting the system configuration macro with sub-macros for defining system elements, subsystem elements, and the connections there between, in Step 106 includes accepting a switch fabric configuration macro for defining switch fabric cards, firmware processing integrated circuits (chips) on the cards, and the connections between the cards. Further, Step 106 may include accepting sub-macros to hierarchically define the system level (slot number), the slot level (card type), card level (chip type), and chip level (link types and endpoint).
  • [0053]
    In some aspects, accepting sub-macros to hierarchically define the system level, the slot level, card level, and chip level, (Step 106) includes differentiating at the slot level between port cards to manage traffic flow in and out of the switch fabric, and switch cards to manage backplane traffic flow between the port cards. In other aspects, Step 106 includes differentiating at the port card level between memory subsystems (MSs) and queuing/scheduling (QS) devices. In yet other aspects, Step 106 includes differentiating at the switch card level between arbiter and crossbars. Further, Step 106 may include defining MS chip level data link connections.
  • [0054]
    Defining MS chip level data link connections may further include defining input port card ingress data links, backplane data links, and output port card egress data links. In other aspects, accepting sub-macros to hierarchically define the system level, the slot level, card level, and chip level, in Step 106 includes defining PQ chip level control link connections. That is, defining grant/bid control links between input port card ingress QS devices and switch card arbiters and defining grant/bid control links between output port card ingress QS devices and switch card arbiters.
  • [0055]
    In other aspects, accepting sub-macros in Step 106 includes defining the chip level. Then, accessing the tables in response to parameters in the executable code in Step 120 includes accessing tables to define chip default register settings and chip modes.
  • [0056]
    Some aspects of the method include further steps. Step 122, following the elaboration of the executable code, monitors system performance. Step 124 compares the system performance to a standard. The standard may be that the data and/or control links merely interface. Alternately, the standard may be a bit error rate, for example, or an information bandwidth. Step 126, in response to sub-standard system performance: automatically repopulates selected parameters in the selected sub-macros; accesses tables in response to the repopulated parameters; and, instantiates new executable code. Note, Step 122 through 126 may be preformed iteratively until a minimum level (the standard) of system performance is obtained. In other aspects, the method may shut to system down if a minimum level of performance is not (eventually) achieved.
  • [0057]
    FIG. 2 is a flowchart illustrating the present invention method for automatically generating code to define and control a switch fabric of connected cards. The method starts at Step 200. Step 202 accepts a system configuration macro with sub-macros for defining switch cards, port cards, integrated circuits (chips) on the cards, and the connections there between. Step 204 accepts a plurality of tables defining a plurality of card behaviors, a plurality of chip behaviors, and a plurality of connection options. Step 206 selects sub-macros. Step 208 populates parameters in the selected sub-macros. Step 210 expands the selected sub-macros. Step 212 generates executable code. Step 214 elaborates the executable code. Step 216, following Step 214, indexes the tables in response to the populated parameters.
  • [0058]
    In some aspects, elaborating the executable code in Step 214 includes differentiating the executable code into code sections corresponding to cards and chips. Then, indexing the tables in response to calling the parameters (Step 216) includes loading accessed tables into corresponding card chip memories.
  • [0059]
    In other aspects, Step 218, following the elaboration of the switch fabric with the executable code, monitors switch fabric performance. Step 220 compares the switch fabric performance to a standard. Step 222, in response to sub-standard switch fabric performance: automatically repopulates selected parameters in the selected sub-macros; indexes tables in response to the repopulated parameters; and, instantiates new executable code.
  • [0060]
    FIG. 3 is a flowchart illustrating another aspect of the present invention method for automatically generating code to define and control a system of connected hardware elements. The method starts at Step 300. Step 302 accepts a system configuration macro with sub-macros for system elements, subsystem elements, and connections there between. Step 304 accepts a plurality of tables with a plurality of system element behaviors, a plurality of subsystem element behaviors, and a plurality of connection options. Step 306 defines the system of connected elements in response to selecting sub-macros. Step 308 defines the physical links between the system elements and the behavior of the system and subsystem elements in response to populating the selected sub-macro parameters. Step 310 expands the selected sub-macros. Step 312 generates executable code. Step 314 accesses the tables in response to parameters in the executable code.
  • [0061]
    FIG. 9 is a flowchart illustrating the present invention method for autonomously discovering the configuration of system-connected elements. The method starts at Step 400. Step 402 initially populates parameters in a macro. In some aspects, the macro is populated with default parameters. Step 404 elaborates table indexing executable code. Step 406 compares the system performance to a standard. Step 408, in response to sub-standard system performance: automatically repopulates parameters in the macro; indexes tables in response to the repopulated parameters; and, instantiates new table indexing executable code.
  • Functional Description
  • [0062]
    The present invention methods use C++ macros to define the physical configuration of a multi processing packet switch fabric system in a hierarchical manner that includes automatic code generation and chip register programming.
  • [0063]
    The System Configuration File is a human-readable configuration file and a C++ source file containing C++ macros combine to expand into executable code that forms the run-time implementation of the a packet switch fabric system, for example. The system configuration file is a human-readable text file that defines the hardware layout in its entirety, and performs initial programming of the chips for a discovered hardware configuration.
  • [0064]
    FIG. 4 is a diagram illustrating the hierarchy of the system configuration file. The format of the configuration file is itself hierarchical, indentured to reflect higher and higher specificity of the system. The firmware actively monitors the health of control and data links between switch chips and adjusts data and control paths dynamically to maintain optimum performance levels. This includes automatic re-assignment of control links and addition or removal of data links. Multiple levels of control performance is supported if the number of control links limits credit bandwidth, automatically switching to dual-credit mode to preserve credit bandwidth.
  • [0065]
    The present invention method is different from conventional system configuration macros in that the form and function of an entire physical subsystem is described and defined in a text file consisting of C++ language macros and hierarchical data tables. Typically, a conventional Element Management System (EMS) maintains a provisioning database that describes cross-connect configurations, equipment and facility states. Various layers of system code use this database to program the hardware. The physical configuration of the system is opaque to this software. The physical layout, connectivity is manifest only when viewed by an operator through the use of a graphical user interface (GUI) that creates a diagram, or as a result of a report output.
  • [0066]
    The present invention system configuration file and tables makes the firmware agnostic of the complete subsystem, enabling it to autonomously reconfigure itself based on its own discovery. A complex system is modeled as a hierarchy of autonomous subsystems managed by a high-level communication protocol. The approach is also flexible; practically every physical parameter of the system is captured in the configuration tables. For ease of use, human-readable macro text is pre-processed, not interpreted. Therefore, any ANSII C++ preprocessor can process it. A special program is not required for interpretation, as an interpreter need not be created to process the file. The invention is useful in that the tables that drive the configuration file can be used as input to a GUI that can graphically diagram how the system is configured both physically and functionally. That is, the combination of system configuration file and the tables completely describe both the physical and functional aspects of the switch subsystem.
  • The AMCC Cyclone Firmware System Configuration Mechanism
  • [0067]
    The present invention method is enabled, in one specific embodiment, as the Cyclone Firmware Subsystem (CFS) System Configuration mechanism, that encapsulates the form and function of an entire physical Cyclone Switch Subsystem (CSS) in a text file consisting of C++ language macros and hierarchical data tables.
  • [0068]
    The configuration system is key to ease of use and flexibility, capturing every physical parameter of the CSS in human-readable tables and macros that can be processed by any ANSII C++ preprocessor to generate the code required to run on any variant of a Cyclone switch system. No special interpreters or proprietary pre-processors are required. The CFS configuration mechanism is not hardware independent, but configuration independent, meaning the CFS can re-program the Cyclone chips to accommodate all possible system configurations.
  • [0069]
    The configuration table approach to system programming is more useful than modifying source code because the tables can alternatively serve as input to a Graphical User Interface (GUI) tool that generate a picture of the system, graphically diagramming how the system is configured, both physically and functionally.
  • Glossary
  • [0000]
      • API Cyclone Firmware Subsystem Application Programming Interface—The API is provided as a means for software on the system controller, or “Host”, to communicate with the Cyclone Firmware Subsystem.
      • CFS Cyclone Firmware Subsystem—CFS refers to the entire collection of firmware that represents the a standalone subsystem that controls the operation of the Cyclone switch hardware. A major sub-component of the CRF is the Token Bus Driver, which is a layer of software that handles the low-level, OS and hardware-dependent read and write operations on the token bus.
      • CSS Cyclone Switch Subsystem—CSS refers to all of the hardware and firmware that makes up a complete Cyclone packet switch fabric. The CSS consists of the Cyclone chips on various port cards and switch cards that comprise the hardware and the firmware that operates on those chips. The CSS is the only context that includes the concept of hardware, all other terms that remain refer to firmware entities.
      • Configuration table module: a programming abstraction of the compile-time configuration class and run-time configuration table object that is encapsulated as a manageable entity, dependent only upon its interface and not implementation. The configuration table module is addressed through the configuration table module API.
      • Configuration table class: compile-time definition of the structure, form, properties and methods (functions) of a configuration table object.
      • Configuration table object: run-time instantiation of a configuration table class, or variable of type “ConfigTable”. The configuration table object includes the configuration table and the methods that operate upon the table.
      • Configuration file: a build-time entity, the configuration file contains the macro definitions that, when elaborated (run, or executed) create the methods that execute upon the data in the configuration table. In Release 1.0 of the firmware, the configuration file name is “SystemConfiguration.cfg”.
      • Configuration table: a run-time entity, the configuration table is the RAM-resident aggregation of data tables that define the Cyclone fabric subsystem configuration, and the values contained in every chip register contained in the system.
  • Configuration Table Module
  • [0078]
    This section describes the purpose, structure and function of the configuration table module. The Configuration Table Module consists of the system configuration files and associated methods. This architectural feature of the CFS abstracts the entire Cyclone switch fabric into one central configuration file. The configuration table allows a complete Cyclone fabric subsystem to be defined and modified without the need to re-program, greatly simplifying the system management overhead required to provision, monitor and upgrade the packet fabric subsystem.
  • [0079]
    The SystemConfiguration.cfg file contains macros that describe the hardware configuration to the Cyclone firmware. The macros expand into object code that executes on the switch card CPU. For example the SWITCH_CARD macro will be replaced by code that, upon execution (elaboration), will instantiate a switch_Card object. The System Configuration File is a human-readable configuration file and a C++ source file containing C++ macros combine to expand into executable code that forms the run-time implementation of the Cyclone packet switch fabric system. It defines the hardware layout in its entirety, and performs initial programming of the cyclone chips for a discovered hardware configuration. The format of the configuration file is itself hierarchical, indentured to reflect higher and higher specificity of the system.
  • [0080]
    FIG. 5 is an illustration of code that is generated for each macro in the Systemconfiguration.cfg file. The left side of the page shows the macro lines in the SystemConfiguration.cfg file, and the right side shows the code that gets generated in SystemConfiguration.cpp (actually, SystemConfiguration.cfg is included at the SystemConfiguration.cpp source file, but broken out as a separate file for clarity).
  • [0081]
    Data for the Configuration Table is accumulated from the header files listed below:
  • [0082]
    SystemConfiguration.cfg: as described above, provides the initial overall slot layout that defines which card type occupies what slot and what chips are configured on each card. The macros in this file are expanded into code by the compiler to make up methods in the ConfigTable object. At startup, these methods verify the hardware configuration, and read values from the tables described below, to program the Cyclone chips on the cards that are found.
  • [0083]
    The System Configuration file “SystemConfiguration.cfg” is modified by the system designer at build time to reflect the hardware layout of the Cyclone fabric. The contents are used at compile time to generate methods that make up the Configuration Table object, which is instantiated at run time to verify the hardware configuration and to pro-gram default values into the chips. The configuration information and the class methods that operate upon this information are included in the boot image that is present in the persistent store of the primary switch card (either hard disk, or some type of EPROM). A configuration table can also be downloaded to the primary card from the host during startup.
  • [0084]
    For each chip, in addition to the chip ID, there are three “modes” which are numbers that point to a section in the register tables. For each chip type there is a set of four register tables. The first table is simply a base table that contains an initial value for each register. The other three tables contain the mode specific values, and these tables are used to modify the base tables at runtime (i.e. the relevant entries in the mode tables are merged with the base tables). At runtime, the firmware starts with a copy of the base tables and successively merges in the appropriate entries from the other tables. The final result is programmed in the chips. The following header files contain tables that are hierarchical, with the first configuration file having the broadest scope for a particular chip type, and successive files having more resolution into the settings of individual chips.
      • ScBaseRegs_<chip_type>.h provide default settings for all chips of chip_type. These files are loaded first, and define the list of base or default registers settings for all the registers for a chip-type. All tables loaded after this are intended to be changes to this base configuration.
      • ScSysConfigRegs_<chip_type>.h provide the system level configuration modes per chip type. These files define the register settings based on a predefined system configuration. A chip's system configuration table has a number of modes that can be selected based on the system platform.
      • ScChipConfigRegs_<chip_type>.h provide the chip level configuration modes. These files define the register settings for individual chips of a particular chip-type within the system. The assignment of chip configuration modes within this table is intended to vary among chips of the same chip-type within a system according to system location.
      • ScCustomPrefRegs_<chip_type>.h provide custom chip level configuration modes. These files are used to define specific register settings that need to be modified from previously loaded tables. CustomPrefRegs_<chip-type>.h files define the tables of register settings that differ or that are not set in the base, system configuration, or chip configuration tables. These files provide the proper mechanism of adding or overriding register settings specified by the base, system configuration, or chip configuration tables.
  • [0089]
    FIG. 6 illustrates the relationship between configuration files and configuration tables.
  • Self-Healing and Load Balancing
  • [0090]
    Self healing is a process by which the firmware dynamically changes system configurations to overcome system failures. Once corrected, the system brings itself back to a new “standard” of performance. In performing self healing, the firmware monitors the health of control and data links between switch chips and dynamically re-routes data and control information when links fail, in order to maintain optimum performance levels. Data and control traffic loading is also re-balanced when cards are inserted or removed from the system. Multiple levels of control performance is supported if the number of control links limits credit bandwidth, automatically switching to dual-credit mode to preserve credit bandwidth. The mode changes are the result of selecting different mode tables and re-programming the chip registers with values from those tables.
  • [0091]
    One aspect of Self Healing is Load Balancing. In the example system of the Cyclone Switch Fabric, Load Balancing would equally balance the “load” of port cards being configured and monitored by a processor on a switch card. Each such processor, through its dedicated Arbiter, can supervise up to 32 port cards. But if there are four working switch cards in the system, each processor would supervise 8 port cards in a load-balanced system. If one switch card were then to fail, the self healing firmware would detect this failure, and a load balancing routine might reallocate ports to switch cards such that 11 port cards are each supervised by switch cards 1 and 2, and the remaining 10 port cards are supervised by switch card 3.
  • [0000]
    Overview of the Cyclone Firmware relating to Load Balancing
  • [0092]
    The Cyclone firmware is a collection of task objects that execute on the switch card CPUs. Communicating through a TCP/IP connection, the Cyclone firmware can be controlled by any host processor through a supplied Application Programming Interface (API). Configurable entirely through the use of configuration files, the Cyclone firmware is a distributed system providing fault-tolerant control of up to 16 switch cards and 32 port cards to program, monitor and control the Cyclone chips in a Cyclone switch fabric system. Typically, a switching subsystem contains multiple, intelligent switch cards that are connected through an internal network in a loosely coupled fashion.
  • [0093]
    A Cyclone Switch Subsystem (CSS) is generally distributed among the port cards and the switch cards. The port cards contain input and output queue chips (8505 iPQ and oPQ, 8805 EDFQ) and memory subsystem chips (8905 iMS and eMS). The switch cards contain an Arbiter (8605) and up to five Crossbar chips (Arbiters and Crossbars are both 8605 parts but are configured by a jumper to operate in either Arbiter or Crossbar mode). Included in the CSS is a controlling CPU that is resident on every switch card.
  • Overview of the Cyclone Firmware Relating to Self Healing
  • [0094]
    The firmware provides the capability to resynchronize individual credit counters in the PQ chips on a periodic basis, or on-demand in case errors are detected by the firmware. The firmware will re-synchronize the appropriate flow control counters of those particular ports for which an error was detected. Typically, credit resync is the result of detecting a status flag set by the S8605 Arbiter chip, indicating either a soft or permanent link error, or by the system detecting that a switch card has failed. If a credit-carrying control link goes down, the link will be re-assigned and the credit counters resynchronized.
  • [0095]
    The failure of a control link between an Arbiter and PQ indicates to the firmware that backplane tokens cannot be reliably communicated over the failed link. The port card may need to be managed by a different switch card in order to re-establish connectivity to the chips on the port card. Once the self healing is complete, the BP_CBFC configuration of the switch may need to be updated. To update the BP_CBFC configuration, the register settings of the appropriate Arbiter chips must be updated, followed by those on the MS chips, and ultimately those on the PQ chips. Reconfiguration of the control links is performed by selecting an appropriate map from the PQ chip configuration file.
  • Example of Self-Healing Configuration Table Entry
  • [0096]
    FIG. 7 depicts an excerpt of a PQ chip configuration file. In response to changing hardware conditions, e.g., a control link failure or a swapped-out switch card, the firmware will scan its configuration tables in the file ScChipConfigRegs_PQ_I.h to determine the next best-fit credit resync configuration. The configuration file contains register settings for the different Cyclone configurations. The “mode” field is used to index into the configuration file to select register maps for a given mode. The mode field is made up of subfields that include credit mode and credit mask settings.
  • [0097]
    The “Chip Configuration Mode” is used as an index into the configuration tables to select the appropriate tables containing the chip register values for a given Chip Configuration Mode. The “Family Mask” defines alternate configurations to support load balancing and self healing. Family Mask is combined with Chip Configuration Mode to find compatible modes for alternate credit transmission/reception. The search is done in the order that the various configurations occur in the configuration file, with the most optimal configurations occurring first, followed by less optimal configurations. At some point the number of links becomes small enough to change the credit mode to “dual” mode, so that each credit returned to an iPQ is counted as two credits, to halve the number of credits that are needed to keep the switch running.
  • [0098]
    FIG. 8 illustrates how the configuration file is indexed, to locate the tables that contain the values to program into the chip registers for a given chip configuration mode.
  • Link Failures
  • [0099]
    When the firmware detects a control or data link error during polling, it sends an internal system error message to the system error handler on the primary switch card. At the same time, an error message is sent to the user. The system error handler reconfigures the switch when it receives a control link failure error and, synchronizes the credit counters of all PQ's when it receives a data link failure error.
  • Control Link Failures
  • [0100]
    The failure of a control link between an Arbiter and PQ indicates to the firmware that tokens cannot be reliably communicated over the failed link. The port card may need to be managed by a different switch card in order to re-establish connectivity to the chips on the port card (this is referred to as self healing from the Firmware perspective, as opposed to load balancing from the switch's perspective, which balances cell traffic load among the available links). Once the load balancing is complete, the Credit Based Flow Control (CBFC) configuration of the switch may need to be updated. To update the CBFC configuration, the register settings of the appropriate Arbiter chips updated, followed by those on the MS chips, and ultimately those on the PQ chips.
  • [0101]
    The self healing Process does the following:
      • 1. Disables polling.
      • 2. Disables the configuration table access on all XbarCard tasks.
      • 3. Distributes the available port cards among the running switch cards based on the available (error-free) Arbiter to PQ control links.
      • 4. Re-enables access to the configuration table on all XBarCard tasks.
      • 5. Distributes the register store based on the new port card distribution.
      • 6. Computes and selects a new primary Arbiter chip and switch card.
      • 7. Computes and selects a new primary PQ chip, port id and port card.
      • 8. Computes the next CBFC configuration to use.
      • 9. Disables CBFC.
      • 10. Configures the Arbiter chips with the new CBFC configuration.
      • 11. Configures the MS chips with the new CBFC configuration.
      • 12. For each PQ chip, configures and enables the new CBFC configuration.
      • 13. Distributes the updated register store to the switch cards.
      • 14. Re-enables polling.
        Data Link Failure from Ingress TM to Ingress MS
  • [0116]
    Data link failures are detected on the receive side of the ingress MS during register polling. The Host is notified of these failures. A data link failure detected on the receive side of the ingress MS indicates that there's an error on the data link between the traffic manager and the iMS. The firmware has a map of how the TM links map between the ingress MS and egress MS chips. Using this map, the system error handler will disable the data link in the eMS “Output Link Enable” bit mask that corresponds to the failed data link detected by the ingress MS.
  • [0000]
    Data Link Failures from Ingress MS to Crossbar
  • [0117]
    Data link failures are detected on the receive side of the Crossbar during register polling. The Host is notified of these failures.
  • [0118]
    Each link is monitored for 8B/10B violations, frame misalignment, byte misalignment, loss of CDR lock, and FIFO overflow/underflow. When the link error threshold is exceeded for a link, the Crossbar will initiate a report to the Arbiter over the token bus. The result is that the Arbiter will no longer grant transmissions over that link. An idle cell is automatically generated by the Crossbar for those outputs that have no corresponding valid input. The errored link is detected during register polling and the Host is notified. Disabling a link reduces switch performance and should be cause for removal of the offending card(s).
  • [0000]
    Data Link Failure from Crossbar to Egress MS
    Data link failures are detected on the receive side of the egress MS (from backplane or another MS) during register polling. The Host is notified of these failures. Failed backplane links between the Crossbar and receiving eMS are communicated to the iPQ on the same port card by the MS2PQ_LINK_FC signal, which communicates all receive link status in round robin order. Once the iPQ gets a failed link notification, it passes the information to the Arbiter during bids (byte-7, bit-1) so that the Arbiter will stop using that link to grant connections to the respective egress port and channel (and subchannel).
  • System Error Counter Timeout
  • [0119]
    The iPQ indicates a credit counter timeout “error” in bit 2 of its High Priority Message register (register 254). The error is detected by a polling task, which reports it to its local XbarCard task. This task does not report the error to the user nor to the system error handler task. Instead, the XBarCard task intercepts the error and sends the NOT_SYNCH_TIMED_OUT_COUNTERS message to the local CreditResync task. The local CreditResync task detects which specific counter or counters have timed out by reading the timed out counters bit mask from the iPQ. The CreditResync task resets each timed out counter to the maximum number of credits depending on the current CBFC mode.
  • [0120]
    A method has been provided for automatically generating code to define and control a system of hardware elements. A switch fabric embodiment of the invention has been presented for illustration. However, the invention is not limited to merely this example. Other variations and embodiments of the invention will occur to those skilled in the art.
Patent Citations
Cited PatentFiling datePublication dateApplicantTitle
US5197016 *14 Jul 198923 Mar 1993International Chip CorporationIntegrated silicon-software compiler
US5231633 *11 Jul 199027 Jul 1993Codex CorporationMethod for prioritizing, selectively discarding, and multiplexing differing traffic type fast packets
US5291479 *16 Jul 19911 Mar 1994Digital Technics, Inc.Modular user programmable telecommunications system with distributed processing
US5422881 *20 Sep 19936 Jun 1995Inmos LimitedMessage encoding
US5594657 *7 Jun 199514 Jan 1997Lucent Technologies Inc.System for synthesizing field programmable gate array implementations from high level circuit descriptions
US5610745 *26 Oct 199511 Mar 1997Hewlett-Packard Co.Method and apparatus for tracking buffer availability
US5781745 *20 May 199614 Jul 19983Com CorporationHigh speed communication bus
US5872904 *24 May 199616 Feb 1999Ncr CorporationComputer system using a master processor to automatically reconfigure faulty switch node that is detected and reported by diagnostic processor without causing communications interruption
US5883814 *13 Mar 199716 Mar 1999International Business Machines CorporationSystem-on-chip layout compilation
US5889779 *2 Dec 199630 Mar 1999Rockwell Science CenterScheduler utilizing dynamic schedule table
US6044061 *10 Mar 199828 Mar 2000Cabletron Systems, Inc.Method and apparatus for fair and efficient scheduling of variable-size data packets in an input-buffered multipoint switch
US6055599 *31 Aug 199825 Apr 2000Electronics & Telecommunications Research InstituteHierarchical crossbar interconnection network for a cluster-based parallel processing computer
US6067286 *9 Apr 199623 May 2000General Datacomm, Inc.Data network switch with fault tolerance
US6072772 *12 Jan 19986 Jun 2000Cabletron Systems, Inc.Method for providing bandwidth and delay guarantees in a crossbar switch with speedup
US6073199 *6 Oct 19976 Jun 2000Cisco Technology, Inc.History-based bus arbitration with hidden re-arbitration during wait cycles
US6101193 *10 Sep 19978 Aug 2000Kabushiki Kaisha ToshibaPacket scheduling scheme for improving short time fairness characteristic in weighted fair queueing
US6252853 *24 Jul 199826 Jun 2001Hitachi, Ltd.Switching router and a datagram transfer system
US6275491 *28 May 199814 Aug 2001Texas Instruments IncorporatedProgrammable architecture fast packet switch
US6343081 *24 Jul 199829 Jan 2002International Business Machines CorporationMethod and apparatus for managing contention in a self-routing switching architecture in a port expansion mode
US6345040 *30 Jul 19985 Feb 2002Marconi Communications, Inc.Scalable scheduled cell switch and method for switching
US6351466 *1 May 199826 Feb 2002Hewlett-Packard CompanySwitching systems and methods of operation of switching systems
US6411599 *2 Dec 199825 Jun 2002International Business Machines CorporationFault tolerant switching architecture
US6490690 *22 Jul 19993 Dec 2002International Business Machines CorporationMethod and apparatus for unix system catastrophic recovery aid
US6515991 *9 Nov 20004 Feb 2003Cisco Technology, Inc.Combined unicast and multicast scheduling
US6519540 *4 Oct 199411 Feb 2003Iris Technologies, Inc.Signal router with cross-point view graphical interface
US6549943 *16 Jun 199915 Apr 2003Cisco Technology, Inc.Network management using abstract device descriptions
US6553026 *30 Jun 199922 Apr 2003Nec CorporationCommunication path switching apparatus, communication path switching method, and computer program product for controlling communication path switching system
US6563837 *10 Feb 199813 May 2003Enterasys Networks, Inc.Method and apparatus for providing work-conserving properties in a non-blocking switch with limited speedup independent of switch size
US6570873 *12 Nov 199927 May 2003Nec CorporationSystem and method for scheduling reservation of traffic with priority
US6591286 *18 Jan 20028 Jul 2003Neomagic Corp.Pipelined carry-lookahead generation for a fast incrementer
US6594695 *8 Dec 199915 Jul 2003Covad Communitions Group, Inc.Network profiling system
US6606588 *26 Jan 199912 Aug 2003Interuniversitair Micro-Elecktronica Centrum (Imec Vzw)Design apparatus and a method for generating an implementable description of a digital system
US6717956 *3 Mar 20006 Apr 2004Luminous Networks, Inc.Dual-mode virtual network addressing
US6721273 *3 Feb 200013 Apr 2004Nortel Networks LimitedMethod and apparatus for traffic flow control in data switches
US6781992 *23 Feb 200124 Aug 2004Netrake CorporationQueue engine for reassembling and reordering data packets in a network
US6850490 *6 Oct 20001 Feb 2005Enterasys Networks, Inc.Hierarchical output-queued packet-buffering system and method
US6870831 *1 May 200122 Mar 2005Pmc-Sierra, Inc.Flexible, self-aligning time and space switch fabrics
US6871294 *25 Sep 200122 Mar 2005Sun Microsystems, Inc.Dynamically reconfigurable interconnection
US6876633 *15 May 20025 Apr 2005Intel CorporationApparatus and method for computer telephone integration in packet switched telephone networks
US6882649 *31 Mar 200019 Apr 2005Sun Microsystems, Inc.Least choice first arbiter
US6895528 *3 Aug 200117 May 2005Computer Network Technology CorporationMethod and apparatus for imparting fault tolerance in a switch or the like
US6898728 *25 Sep 200124 May 2005Sun Microsystems, Inc.System domain targeted, configurable interconnection
US6915372 *8 Jun 20015 Jul 2005Tellabs Operations, Inc.Methods and apparatus for managing traffic through a buffered crossbar switch fabric
US6990063 *7 Mar 200024 Jan 2006Cisco Technology, Inc.Distributing fault indications and maintaining and using a data structure indicating faults to route traffic in a packet switching system
US6990072 *14 Aug 200124 Jan 2006Pts CorporationMethod and apparatus for arbitration scheduling with a penalty for a switch fabric
US6999457 *29 Mar 200114 Feb 2006Juniper Networks, Inc.Arbiter circuit and method of carrying out arbitration
US7009985 *21 Jan 20037 Mar 2006Broadcom CorporationFibre channel arbitrated loop bufferless switch circuitry to increase bandwidth without significant increase in cost
US7016350 *7 May 200121 Mar 2006Xyratex Technology LimitedData switch and a method for controlling the data switch
US7020131 *24 Dec 200128 Mar 2006Applied Micro Circuits Corp.System and method for hierarchical switching
US7023857 *12 Sep 20004 Apr 2006Lucent Technologies Inc.Method and apparatus of feedback control in a multi-stage switching system
US7035212 *25 Jan 200125 Apr 2006Optim NetworksMethod and apparatus for end to end forwarding architecture
US7042842 *21 Jun 20049 May 2006Computer Network Technology CorporationFiber channel switch
US7042883 *16 Oct 20019 May 2006Juniper Networks, Inc.Pipeline scheduler with fairness and minimum bandwidth guarantee
US7050448 *5 Apr 200123 May 2006Xyratex Technology LimitedData switching arbitration arrangements
US7058057 *1 May 20016 Jun 2006Integrated Device Technology, Inc.Network switch port traffic manager having configurable packet and cell servicing
US7065046 *6 Apr 200120 Jun 2006Lucent Technologies Inc.Scalable weight-based terabit switch scheduling method
US7072295 *14 Sep 20004 Jul 2006Tellabs Operations, Inc.Allocating network bandwidth
US7075990 *16 Jan 200311 Jul 2006Sbc Properties, L.P.Method and system to improve the transport of compressed video data in real time
US7079545 *17 Dec 200118 Jul 2006Applied Microcircuits Corporation ( Amcc)System and method for simultaneous deficit round robin prioritization
US7158528 *14 Dec 20012 Jan 2007Agere Systems Inc.Scheduler for a packet routing and switching system
US7158536 *28 Jan 20042 Jan 2007Rambus Inc.Adaptive-allocation of I/O bandwidth using a configurable interconnect topology
US7180867 *26 Dec 200020 Feb 2007Lucent Technologies Inc.Apparatus and method for flow path based fault detection and service restoration in a packet based switching system
US7203202 *31 Oct 200210 Apr 2007Polytechnic UniversityArbitration using dual round robin matching with exhaustive service of winning virtual output queue
US7209453 *3 Mar 200324 Apr 2007Applied Micro Circuits CorporationSystem and method for tolerating control link faults in a packet communications switch fabric
US7218635 *31 Aug 200115 May 2007Stmicroelectronics, Inc.Apparatus and method for indexing MPEG video data to perform special mode playback in a digital video recorder and indexed signal associated therewith
US7221652 *3 Mar 200322 May 2007Applied Micro Circuits CorporationSystem and method for tolerating data link faults in communications with a switch fabric
US7227860 *1 Mar 20015 Jun 2007Nec CorporationPacket switch and method of scheduling the same
US7236488 *9 Aug 200226 Jun 2007Gautam KavipurapuIntelligent routing switching system
US7236491 *19 Sep 200126 Jun 2007Industrial Technology Research InstituteMethod and apparatus for scheduling for packet-switched networks
US7327725 *27 Nov 20025 Feb 2008Alcatel Canada Inc.Method and apparatus capable of transferring very high data rates across a midplane or backplane
US7342929 *26 Apr 200211 Mar 2008Cisco Technology, Inc.Weighted fair queuing-based methods and apparatus for protecting against overload conditions on nodes of a distributed network
US7349403 *19 Sep 200225 Mar 2008Bay Microsystems, Inc.Differentiated services for a network processor
US7406041 *31 Jul 200229 Jul 2008Brocade Communications Systems, Inc.System and method for late-dropping packets in a network switch
US7515537 *28 Jun 20067 Apr 2009Mcdata Services CorporationMethod and apparatus for rendering a cell-based switch useful for frame based protocols
US20010010910 *3 Aug 19992 Aug 2001Boston Probes Inc.Pna probes, probe sets, methods and kits pertaining to the universal detection of bacteria and eucarya
US20020012344 *5 Jun 200131 Jan 2002Johnson Ian DavidSwitching system
US20020039351 *15 Mar 20014 Apr 2002Linghsiao WangBandwidth sharing using emulated weighted fair queuing
US20020039364 *1 Oct 20014 Apr 2002Satoshi KamiyaTwo-dimensional pipelined scheduling technique
US20020058258 *5 Mar 200116 May 2002Wittwer Carl T.Monitoring hybridization during PCR using SYBR Green I
US20020061028 *1 Jun 200123 May 2002Polytechnic UniversityScheduling the dispatch of cells in multistage switches using a hierarchical arbitration scheme for matching non-empty virtual output queues of a module with outgoing links of the module
US20020105966 *16 Nov 20018 Aug 2002Ronak PatelBackplane interface adapter with error control and redundant fabric
US20020118692 *4 Jan 200129 Aug 2002Oberman Stuart F.Ensuring proper packet ordering in a cut-through and early-forwarding network switch
US20030026206 *31 Jul 20026 Feb 2003Mullendore Rodney N.System and method for late-dropping packets in a network switch
US20030026267 *5 Jun 20026 Feb 2003Oberman Stuart F.Virtual channels in a network switch
US20030031193 *6 Apr 200113 Feb 2003Andrews Daniel M.Scalable weight-based terabit switch scheduling method
US20030035427 *14 Aug 200120 Feb 2003Mehdi AlastiMethod and apparatus for arbitration scheduling with a penalty for a switch fabric
US20030043847 *31 Aug 20016 Mar 2003Haddad Semir S.Apparatus and method for indexing MPEG video data to perform special mode playback in a digital video recorder and indexed signal associated therewith
US20030053470 *18 Sep 200120 Mar 2003Divivier Robert J.Multicast cell buffer for network switch
US20030058880 *21 Sep 200127 Mar 2003Terago Communications, Inc.Multi-service queuing method and apparatus that provides exhaustive arbitration, load balancing, and support for rapid port failover
US20030072326 *14 Aug 200117 Apr 2003Mehdi AlastiMethod and apparatus for parallel, weighted arbitration scheduling for a switch fabric
US20030076848 *26 Apr 200224 Apr 2003Anat Bremler-BarrWeighted fair queuing-based methods and apparatus for protecting against overload conditions on nodes of a distributed network
US20030112757 *16 Jul 200219 Jun 2003Thibodeau Mark JasonSystem and method for providing gaps between data elements at ingress to a network element
US20040030766 *12 Aug 200212 Feb 2004Michael WitkowskiMethod and apparatus for switch fabric configuration
US20040047408 *10 Sep 200211 Mar 2004Ingo KoenenkampData link analyzer
US20040093601 *7 Nov 200213 May 2004Quicksilver Technology, Inc.Method, system and program for developing and scheduling adaptive integrated circuity and corresponding control or configuration information
US20040153566 *31 Jan 20035 Aug 2004Brocade Communications Systems, Inc.Dynamic link distance configuration for extended fabric
US20040160970 *29 Sep 200319 Aug 2004Avici Systems, Inc.Methods and apparatus for event-driven routing
US20040161767 *30 Jun 200319 Aug 2004Baldwin Brett R.Detection and quantification of aromatic oxygenase genes by real-time PCR
US20060013207 *20 Sep 200519 Jan 2006Mcmillen Robert JReconfigurable, fault tolerant, multistage interconnect network and protocol
US20060127942 *6 Feb 200615 Jun 2006Tore StraumeParticle analysis assay for biomolecular quantification
Referenced by
Citing PatentFiling datePublication dateApplicantTitle
US7779400 *5 Dec 200517 Aug 2010Via Technologies Inc.System and method for modifying firmware of an optical storage medium device without requiring a compiling process
US841812926 Mar 20039 Apr 2013Qualcomm IncorporatedMethod for automatically generating code to define a system of hardware elements
US8669991 *13 Oct 201011 Mar 2014Nvidia CorporationMethod macro expander
US9401004 *12 Oct 201026 Jul 2016Nvidia CorporationState shadowing to support a multi-threaded driver environment
US20070055794 *5 Dec 20058 Mar 2007Willy ChuangSystem and method for modifying firmware of an optical storage medium device without requiring a compiling process
US20110084977 *12 Oct 201014 Apr 2011Duluk Jr Jerome FrancisState shadowing to support a multi-threaded driver environment
US20120072701 *13 Oct 201022 Mar 2012Duluk Jr Jerome FrancisMethod Macro Expander
US20130191569 *25 Jan 201225 Jul 2013Qualcomm IncorporatedMulti-lane high-speed interfaces for high speed synchronous serial interface (hsi), and related systems and methods
Classifications
U.S. Classification715/735, 717/107
International ClassificationG06F15/16, G06F9/44, G06F3/048
Cooperative ClassificationH04L49/252, G06F8/34, H04L49/503, H04L49/3027, H04L49/3018, H04L49/205
European ClassificationG06F8/34, H04L49/25C
Legal Events
DateCodeEventDescription
21 Nov 2008ASAssignment
Owner name: QUALCOMM INCORPORATED, CALIFORNIA
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:APPLIED MICRO CIRCUITS CORPORATION;REEL/FRAME:021876/0013
Effective date: 20080715
Owner name: QUALCOMM INCORPORATED,CALIFORNIA
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:APPLIED MICRO CIRCUITS CORPORATION;REEL/FRAME:021876/0013
Effective date: 20080715
19 May 2009ASAssignment
Owner name: APPLIED MICROCIRCUITS CORPORATION, CALIFORNIA
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:RIPA, ALBERTO;BENSCHOP, PETER;CLOVIS, PHILIP;AND OTHERS;REEL/FRAME:022704/0072;SIGNING DATES FROM 20030324 TO 20030325
Owner name: APPLIED MICROCIRCUITS CORPORATION, CALIFORNIA
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:RIPA, ALBERTO;BENSCHOP, PETER;CLOVIS, PHILIP;AND OTHERS;SIGNING DATES FROM 20030324 TO 20030325;REEL/FRAME:022704/0072