Intel Corporation

John Wishneusky

Sanjeev Jain

Niall D McDonnell

David Romano

Gavin J Stark

Gilbert M Wolrich

A method and apparatus for scheduling packets using a pre-sort deficit round-robin method. Scheduling decisions for packets are made when packets are received, and entries for the received packets are stored in a pre-sorted scheduling array. A packet is transmitted by dequeuing the packet from the pre-sorted scheduling array.

Method and apparatus for scheduling packets

A method and apparatus for scheduling packets using a pre-sort scheduling array having one or more smoothing registers. The scheduling array includes a number of round buffers, each round buffer having an associated smoothing register. To schedule a packet for transmission, the packet's transmission round and relative position within that round are determined, and an identifier for the packet is placed at the appropriate position within the scheduling array. A bit of the associated smoothing register is set, the set bit corresponding to the entry receiving the packet identifier. During transmission, the set bits of the smoothing register associated with a current round buffer are read to identify packets that are to be dequeued.

In general, in one aspect, the disclosure describes a method of processing bits of a frame. The method includes accessing a subset of bits of a frame and based, at least in part, on the subset of bits, determining an address of an instruction within a set of instructions that perform at least one framing operation on the bits. The method executes instructions at the determined instruction address.

Processing frame bits

According to some embodiments, a method determining a number of stages associated with an instruction to be executed via a processor pipeline, determining a number of stages associated with a subsequent instruction, and stalling the pipeline based on the number of stages associated with the instruction to be executed and the number of stages associated with the subsequent instruction is provided.

Stall optimization for an in-order, multi-stage processor pipeline which analyzes current and next instructions to determine if a stall is necessary

A programmed state processing machine architecture and method that provides improved efficiency for processing data manipulation tasks. In one embodiment, the processing machine comprises a control engine and a plurality coprocessors, a data memory, and an instruction memory. A sequence of instructions having a plurality of portions are issued by the instruction memory, wherein the control engine and each of the processors is caused to perform a specific task based on the portion of the instructions designated for that component. Accordingly, a data manipulation task can be divided into a plurality of subtasks that are processed in parallel by respective processing components in the architecture.

Multiple coprocessor architecture to process a plurality of subtasks in parallel

According to some embodiments, storage registers are provided for a processor pipeline.

Storage registers for a processor pipeline

A Media Access Control (MAC) Bus interface definition and multiplexor scheme that may be implemented to provide chip layout-insensitive connections between a number of communication physical layer port entities and a single buffer manager or communications controller entity, utilizing a set of independent pipelined buses. The interface comprising three buses: A MAC In Data bus, a MAC Out Data bus, and a MAC Out Message bus. Each bus can operated with an independent set of timing signals to enable data transfers between a system side block and one or more network side blocks. The multiplexor scheme provides a multiplexor for each of the MAC buses, and enables a single system side block to connect to multiple network side blocks. The multiplexors may be also be cascaded.

MAC bus interface

A programmable processing system that executes multiple instruction contexts includes an instruction memory for storing instructions that are executed by the system, fetch logic for determining an address of an instruction, with the fetch logic including scheduling logic that schedules execution of the instruction contexts based on condition signals indicating an availability of a hardware resource, with the condition signals being divided into groups of condition signals, which are sampled in turn by the scheduling logic to provide a plurality of scan sets of sampled conditions.

Context scheduling

A processing system that executes multiple instruction contexts includes an instruction memory for storing instructions that are executed by the system, a processor unit executing the instructions in a pipelined fashion, a plurality of context registers for storing instructions and instruction addresses for contexts to be executed and fetch logic for selecting an address from one of the context registers and for selecting an instruction from a second of the context registers for execution substantially simultaneously for each cycle of execution of processor unit.

Context processing by substantially simultaneously selecting address and instruction of different contexts

A programmable processing system includes a first processor for executing a first portion of an instruction, a second processor for executing a second portion of the instruction, where the second portion of the instruction is interpreted by the first processor as an extension to an immediate operand field included in the first portion of the instruction.

Growing community of inventors

Fitzwilliam, NH, United States of America

John Wishneusky