Nvidia Corporation

National Semiconductor Corporation

Advanced Micro Devices Corporation

Freescale Semiconductor,inc.

David William Nuechterlein

Paolo Enrique Sabella

Dennis K Ma

Lincoln G Garlick

John H Edmondson

Samuel Hammond Duncan

Robert Anthony Alfieri

Jonathan B White

David J Husak

Christopher M Herring

Willard S Briggs

Brian Dennis Falardeau

Rajeev Jayavant

Jeffrey M Smith

Michael Asbury Woodmansee

Shail Dave

Matthew Melton

Vikramjeet Singh

David F Barton

Syed I Shah

Jon L Fluker

Sequential fetch requests from a set of fetch requests are combined into longer coalesced requests that match the width of a system memory interface in order to improve memory access efficiency for reading the data specified by the fetch requests. The fetch requests may be of different classes and each data class is coalesced separately, even when intervening fetch requests are of a different class. Data read from memory is ordered according to the order of the set of fetch requests to produce an instruction stream that includes the fetch requests for the different classes.

Request coalescing for instruction streams

A system and method dispatches commands from multiple instruction streams to processing engines, allowing for some of the dispatched commands to be rescinded before they are executed by the processing engines. The dispatching enables several of the processing engines to execute commands concurrently. Dispatched commands may be rescinded to quickly switch processing from one instruction stream to another instruction stream.

Rescindable instruction dispatcher

A processor having multiple independent engines can concurrently support a number of independent processes or operation contexts. The processor can independently schedule instructions for execution by the engines. The processor can independently switch the operation context that an engine supports. The processor can maintain the integrity of the operations performed and data processed by each engine during a context switch by controlling the manner in which the engine transitions from one operation context to the next. The processor can wait for the engine to complete processing of pipelined instructions of a first context before switching to another context, or the processor can halt the operation of the engine in the midst of one or more instructions to allow the engine to execute instructions corresponding to another context. The processor can affirmatively verify completion of tasks for a specific operation context.

Method and apparatus for context switching of multiple engines

Techniques for performing user-configurable traffic management functions on streams of packets. The functions include multicasting, discard, scheduling, including shaping, and segmentation and reassembly. In the techniques, the functions are not performed directly on the packets of the stream, but instead on descriptors that represent stored packets. Output of descriptors from all traffic queues, including discard traffic queues, is scheduled. Scheduling is done using a hierarchy of schedulers. The form of the hierarchy and the scheduling algorithms used by the schedulers in the hierarchy are both user configurable. As disclosed, the techniques are implemented in a traffic management coprocessor integrated circuit. The traffic manager coprocessor is used with a digital communications processor integrated circuit that performs switching functions. The buffers for the packets are in the digital communications processor. Also disclosed are a modified partial packet discard algorithm and a frame based deficit round robin scheduling algorithm.

Manipulating data streams in data stream processors

A data processing system can establish or maintain data coherency by issuing a data flush operation. The data processing system can be configured as a host executing one or more independent processes using one or more lower level devices. The lower level devices can be viewed as peer devices. Any of the host or the plurality of peer devices can be configured to initiate the flush operation. A device can determine whether the initiator of a flush operation is the host or a peer device. The device can perform a flush limited to local memory, or a subset of all available memory, if a peer device initiates the flush operation.

Hierarchical flush barrier mechanism with deadlock avoidance

Software can freeze portions of a pipeline operation in a processor by asserting a predetermined freeze register in the processor. The processor halts operations relating to portions of a common pipeline processing in response to an asserted freeze register. Processor resources that operate downstream from the common pipeline continue to process any scheduled instructions. The processor is prevented from initiating any context switching in which a processor resource is allocated to a different channel. The processor stops supplying any additional data to downstream resources and ensures that the interface to downstream resources is clear of previously sent data. The processor prevents state machines from making additional requests. The processor asserts an acknowledgement indication in response to the freeze assertion when the processing has reached a stable state. Software is allowed to manipulate states and registers within the processor. Clearing the freeze register allows processing to resume.

Indicating acknowledge of stable state of pipeline resource when freeze bit set and context switch inhibited

The present invention introduces circuitry and systems for performing two-dimensional motion compensation using a three-dimensional pipeline, as well as methods of operating the same. According to an exemplary embodiment, image processing circuitry is provided and includes both a two-dimensional image pipeline, which is operable to process two dimensional image data to generate successive two-dimensional image frames, and a three-dimensional image pipeline, which is operable to process three-dimensional image data to render successive three-dimensional image frames. The image processing circuitry further includes dual mode sub-processing circuitry, which is associated with each of the two- and three-dimensional image pipelines. The dual mode sub-processing circuitry is operable to perform motion compensation operations associated with the two-dimensional image pipeline in one mode and to perform rasterization operations associated the three-dimensional image pipeline in another mode.

Circuitry and systems for performing two-dimensional motion compensation using a three-dimensional pipeline and methods of operating the same

There is disclosed an apparatus for providing a virtual PCI device for use in a processing system comprising a data processor having an external peripheral bus coupled thereto in which peripheral devices associated with the external peripheral bus are controlled by accessing configuration circuitry associated with each of the peripheral devices. The apparatus comprises: a) an address trap circuit for detecting a configuration cycle accessing a virtual configuration address space associated with the virtual PCI device and generating an enable signal in response and b) an interrupt generation circuit associated with the address trap circuit that receives the enable signal and, in response, generates an interrupt signal that causes the data processor to execute instructions stored in system memory associated with the virtual device.

System and method for virtualizing the configuration space of PCI devices in a processing system

There is disclosed an apparatus for loading instructions into the instruction execution pipeline of a pipelined processor. The apparatus for loading instructions comprises: 1) an instruction loading circuit that loads instructions from a first instruction thread into the instruction execution pipeline; and 2) a branch instruction detection circuit that detects a branch instruction in the first instruction thread. In response to the branch instruction detection, the instruction loading circuit stops loading instructions from the first instruction thread into the instruction execution pipeline and begins loading instructions from a second instruction thread into the instruction execution pipeline.

Growing community of inventors

Longmont, CO, United States of America

David William Nuechterlein