摘要:
The disclosure describes queue management based on queue sets. A queue set comprises a group of packets or packet references that are processed as a single entity or unit. For example, when a queue set reaches the head of a queue in which it is stored, the entire queue set including its packets or packet references is passed for scheduling as a single unit. A queue set provides the benefit of a single operation associated with enqueuing and a single operation associated with dequeuing. Since only one operation on a queue is required for the typical case of several packets in a queue set rather than for every packet, the rate of queue operations may be significantly reduced. A queue set has a target data unit size, for example, a roughly equal number of packet bytes represented by each queue set, regardless of the number of packets referenced by a queue set. This means that a scheduler of a queue manager, which is tasked with metering the number of packet bytes transmitted from each queue per time unit, is provided with a list of packets which represents a predictable quantity of packet bytes, and this predictability streamlines the scheduling task and significantly reduces the number of operations.
摘要:
Enhanced switching functionality is achieved by a switching system having of traffic management devices connected with a generic Ethernet switch fabric and an optional flow control device. The traffic management devices use an encapsulated Ethernet frame format to provide an additional of addressing.
摘要:
An architecture for a multi-port switching device is described having a very regular structure that lends itself to scaling for performance speed and a high level of integration. The distribution of packet data internal to the chip is described as using a cell-based TDM packet transport configuration such as a ring. Similarly, a method of memory allocation in a transmit buffer of each port allows for reassembly of the cells of a packet for storage in a contiguous manner in a queue. Each port includes multiple queues. The destination queue and port for a packet is identified in a multi-bit destination map that is prepended to the start cell of the packet and used by a port to identify packets destined for it. The architecture is useful for a single-chip multi-port Ethernet switch where each of the ports is capable of 10 Gbps data rates.
摘要:
A multi-slice network processor processes a packet in packet slices for transfer over a multi-port network interface such as a switch fabric. The network processor segments a packet into cells having a target size. A group of cells of a common packet form a packet slice which is independently processed by one of a number of parallel processing and storage slices. Load balancing may be used in the selection of processing slices. Furthermore, the network processor may load balance slices across the multi-port network interface to one or more destination slices of another network processor. The multi-slice processor uses post header storage delivery on ingress processing to the multi-port interface thereby reducing temporary storage requirements. The multi-slice network processor may also utilize sequence numbers associated with each packet to ensure that prior to transmission onto a destination network, the packet is in the correct order for a communication flow.
摘要:
A system, method and tangible medium for unified exception handling with distributed exception identification includes a packet processing pipeline with at least two processing stages for processing data packets, each of the data packets being processed having an associated exception map in a memory of the apparatus. An exception detector at each processing stage detects whether any exception conditions apply to the data packet at the processing stage, and if so, a bit setter sets, modifies, or resets one or more bits in the exception map associated with exception conditions detected at the processing stage. An exception handler processes the exception map in response to the state of in the exception map after all the processing stages are complete.
摘要:
Methods and apparatus for enabling communication between a source network device and one or more destination network devices are disclosed. A system enabling communication between a source network device and one or more destination network devices includes a switch and a ring interconnect. The switch is adapted for connecting to the source network device and the one or more destination network devices. More particularly, the switch is capable of storing data provided by the source network device and retrieving the data for the one or more destination network devices. The ring interconnect is adapted for connecting the source network device and the one or more destination network devices to one another. In addition, the ring interconnect is capable of passing one or more free slot symbols along the ring interconnect. Thus, the ring interconnect is capable of expanding when one or more of the free slot symbols are each replaced by a frame notify message indicating that the data has been stored by the switch for retrieval by the one or more destination network devices.
摘要:
The disclosure describes implementations for accessing in parallel a plurality of banks across a plurality of DRAM devices. These implementations are suited for operation within a parallel packet processor. A data word in partitioned into data segments which are stored in the plurality of banks in accordance with an access scheme that hides pre-charging of rows behind data transfers. A storage distribution control module is communicatively coupled to a memory comprising a plurality of storage request queues, and a retrieval control module is communicatively coupled to a memory comprising a plurality of retrieval request queues. In one example, each request queue may be implemented as a first-in-first-out (FIFO) memory buffer. The plurality of storage request queues are subdivided into sets as are the plurality of retrieval queues. Each is set is associated with a respective DRAM device. A scheduler for each respective DRAM device schedules data transfer between its respective storage queue set and the DRAM device and between its retrieval queue set and the DRAM device independently of the scheduling of the other devices, but based on a shared criteria for queue service.
摘要:
Unified exception handling may be provided by processing a data packet through at least two pipelined processing stages in a data packet processor such as a switch, router, bridge, or similar network device, each of the data packets having associated with it (while it is being processed) an exception map disposed in a memory of the network device. The bits in the exception map are set, modified, or reset in response to exception conditions detected at the various processing stages. After the packet has been fully processed, an exception handler takes as an input the exception map and further processes the packet in response to the state of the exception map.
摘要:
Methods and apparatus for providing a source interface device and destination interface device are disclosed. A method of enabling communication between the source device and one or more destination devices includes sending data from the source device to the switch for storage. A frame notify message addressed to the one or more destination devices and indicating that the data has been stored by the switch for retrieval is then sent on the ring interconnect. One of the specified destination devices obtains the frame notify message from the source device via the ring interconnect. A frame retrieval message identifying the data is then sent from the destination device to the switch in response to the frame notify message. In addition, the destination device modifies the frame notify message to indicate whether the destination device was capable of accepting the frame notify message. The modified frame notify message is then sent on the ring interconnect for retrieval by another destination device or the source device.