# US 2002/0057708 A1 · Enhancing performance by pre-fetching and caching data directly in a communication processor's register set

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2002/0057708 A1
- Document ID: US20020057708A1
- Status: Published application
- Kind: A1
- Application number: 09/919,216
- Filed: 2001-07-31
- Published: 2002-05-16
- Earliest claimed: 2000-07-31
- Assignees: Mindspeed Technologies, Inc.
- Inventors: Duane E. Galbi; Wilson P. Snyder II; Daniel J. Lussier
- Web page: https://patentgrep.com/patent/US20020057708A1

## CPC

- G06F 9/30141: {Implementation provisions of register files, e.g. ports}
- G06F 9/30123: {according to context, e.g. thread buffers}
- G06F 9/383: {Operand prefetching (cache prefetching G06F12/0862)}
- G06F 9/3851: {from multiple instruction streams, e.g. multistreaming}
- G06F 15/7825: {Globally asynchronous, locally synchronous, e.g. network on chip}
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

Circuitry to free the core processor from performing the explicit read operation required to read data into the internal register set. The processor's register set is expanded and a “shadow register” set is provided. While the core processor is processing one event the “context” and “data” and other associated information for the next event is loaded into the shadow register set. When the core processor finishes processing an event, the core processor switches to the shadow register set and it can begin processing the next event immediately. With short service routines, there might not be time to fully pre-fetch the “context” and “data” associated with the next event before the current event ends. In this case, the core processor still starts processing the next event and the pre-fetch continues during the event processing. If the core processor accesses a register which is associated with part of the context for which the pre-fetch is still in progress the core processor will automatically stall or delay until the pre-fetch has completed reading the appropriate data.

## Claims

### Claim 1

An integrated circuit for processing communication packets, said integrated circuit comprising:
a core processor configured to execute software to process a series of communication packets, the processing of each packet being an event and having associated data and context information, said core processor having two sets of data registers, each set of data registers being capable of storing the context and data information required to process one event, said core processor using said sets of registers alternatively; and
a co-processor having a plurality of data buffers configured to store data and context information associated with a plurality of packets, data and context from one event being transferred to one of said sets of data registers in said core processor while said core processor is utilizing data and context information stored in a different set of data registers in said processor, whereby said processor need not wait between event processing in order to load data in said registers.

### Claim 2 (depends on 1)

The integrated circuit of claim 1 wherein said data registers comprise a low latency memory.

### Claim 3 (depends on 1)

The integrated circuit of claim 1 wherein said data registers comprise a low latency memory and said data buffers in said co-processor comprises a medium latency memory.

### Claim 4 (depends on 1)

The integrated circuit of claim 1 wherein said data registers comprise a low latency memory, said data buffers in said co-processor comprises a medium latency memory, and said system included off chip high latency memory.

### Claim 5 (depends on 1)

The integrated circuit of claim 1 wherein said core processor is configured to pre-fetch data into one set of registers while said core processor processes an event using data in another set of registers.

### Claim 6 (depends on 1)

The integrated circuit of claim 1 further comprising means for preventing the occurrence of two back to back events which use the same context data.

### Claim 7 (depends on 1)

The integrated circuit of claim 1 comprising a queue of events.

### Claim 8 (depends on 7)

The integrated circuit of claim 7 comprising logic configured to detect when an event being processes uses the same context as that used by an event immediately preceding the preceding event (the ABA case), and which in the ABA case, delay the selection of the data for the second event until data which affects the selection mechanism has been emptied from the queue.

### Claim 9 (depends on 1)

The integrated circuit of claim 1 further comprising a work queue configured to prioritize communication packets for processing, said work queue comprising a detector configured to determine if sequential communications packets queued for transmission to said core processor utilize the same context data, and delay transmission of the second such communication packet until processing of the first such communication packet is complete.

### Claim 10

A method of processing communication packets in a system which comprises a core processor comprising a first set of registers and a shadow set of registers, the method comprising:
in said first set of registers and said shadow set of registers, storing the context and data necessary to process one communication packet and a co-processor with a plurality of buffers configured to store data and context information necessary to process a plurality of packets; and
transferring data and context information associated with a second communication packet from said co-processor to said shadow set of registers while said core processor is using the data and context information said first set of registers to process a first communication packet.

### Claim 11 (depends on 10)

The method of claim 10 wherein said registers are a low latency memory.

### Claim 12 (depends on 10)

The method of claim 10 wherein said data registers are a low latency memory and said data buffers in said co-processor are a medium latency memory.

### Claim 13 (depends on 10)

The method of claim 10 further comprising:
in said core processor, pre-fetching data into one set of registers while said core processor is processor is processing an event using data in another set of registers.

### Claim 14 (depends on 10)

The method of claim 10 further comprising:
preventing the occurrence of two back to back events which use the same context data.

### Claim 15 (depends on 10)

The method of claim 10 wherein said co-processor includes a queue of events.

### Claim 16 (depends on 15)

The method of claim 15 wherein when an event being processes uses the same context as that used by an event immediately preceding the preceding event (the ABA case), the selection of the data for the second A event is delayed until data which affects the selection mechanism has been emptied from the queue.

### Claim 17

A system for processing communication packets comprising a core processor and a co-processor, said core processor including two sets of registers, means in said core processor for initiating pre-fetch of data into one set of said registers from said co-processor and for at the same time initiating processing of an event using data in the other set of registers, whereby transfer of data from the co-processor to the core processor can take place at the same time as said core processor is processing data for a different communication packet.

### Claim 18 (depends on 17)

The system of claim 17 wherein said registers are a low latency memory.

### Claim 19 (depends on 17)

The system of claim 17 wherein said data registers are a low latency memory and said co-processor includes medium latency memory.

### Claim 20 (depends on 17)

The integrated circuit of claim 17 including means for preventing the occurrence of two back to back events which use the same context data.

### Claim 21 (depends on 17)

The integrated circuit of claim 17 including a queue of events.

### Claim 22 (depends on 21)

The integrated circuit of claim 21 including logic configured to detect when an event being processes uses the same context as that used by an event immediately preceding the preceding event (the ABA case), and which in the ABA case, delay the selection of the data for the second A event until data which affects the selection mechanism has been emptied from the queue.

### Claim 23

An integrated circuit for processing communication packets, said integrated circuit including:
a core processor configured to execute software to process a series of communication packets, the processing of each packet requiring data and context information, said core processor comprising two sets of low latency data registers, each set of data registers configured to store the context and data information required to process one communication packet, said core processor using said sets of registers alternatively;
a co-processor comprising a plurality of medium latency data buffers configured to store data and context information associated with a plurality of communication packets; and
pre-fetch circuitry configured to transfer data and context information required for processing one communication packet from said co-processor to one of said sets of data registers in said core processor while said core processor is utilizing data and context information stored in a different set of data registers to processes a different communication packet.
