Processor having a bus interconnect which is dynamically reconfigurable in response to an instruction field

• US 5,771,362 A
• Filed: 05/17/1996
• Issued: 06/23/1998
• Est. Priority Date: 05/17/1996
• Status: Expired due to Term

• Current Assignee: Samsung Electronics Co. Ltd. (Samsung Group)
• Original Assignee: Advanced Micro Devices, Inc.
• Inventors: Lynch, Thomas W., Bartkowiak, John G.
• Primary Examiner(s): AUVE, GLENN ALLEN
• Application Number:
  US08/649,810
• Time in Patent Office: 767 Days
• Field of Search: 395/312, 395/311, 395/309, 395/310, 395/306, 395/562, 395/384, 395/566
• US Class Current: 710/317
• CPC Class Codes: G06F 9/30145 Instruction analysis, e.g. decoding, instruction word fields
• CPC Class Codes: G06F 9/3824 Operand accessing
• CPC Class Codes: G06F 9/3826 Bypassing or forwarding of data results, e.g. locally between pipeline stages or within a pipeline stage
• CPC Class Codes: G06F 9/3885 using a plurality of independent parallel functional units
• CPC Class Codes: G06F 9/3893 controlled in tandem, e.g. multiplier-accumulator
• CPC Class Codes: G06F 9/3897 with adaptable data path

Abstract

A processor employing a dynamically configurable bus interconnect is provided. The interconnect routes data between functional units and memories included within the processor in response to an instruction field. As opposed to a particular hard-wired interconnect, the dynamically variable interconnect may be modified to form an optimum interconnect for the particular algorithm being executed. Still further, the interconnect may be modified between several configurations during the execution of the algorithm, as often as each clock cycle. Because an instruction field is used to directly specify the configuration of the interconnect during execution of that instruction, control over the interconnect is afforded to the programmer writing the code which implements a particular algorithm.

Claims

1. A processor, comprising:

an instruction sequencer configured to execute instructions having an opcode field and a bus configuration field, wherein said instruction sequencer is configured to produce at least one control signal in response to a first encoding of said bus configuration field; and

a bus interconnect coupled to receive said control signal from said instruction sequencer, said bus interconnect including a first plurality of buses and a second plurality of buses, wherein said bus interconnect is configured to route data from one of said first plurality of buses to one of said second plurality of buses if said control signal is asserted,

wherein data from said one of said first plurality of buses is prevented from being routed to said one of said second plurality of buses if said control signal is deasserted.

2. The processor as recited in claim 1 wherein said bus interconnect comprises a crossbar interconnect, and wherein said crossbar interconnect is configured to receive a different control signal for each one of said first plurality of buses and each one of said second plurality of buses, and wherein said different control signal, if asserted, causes routing of data from said one of said first plurality of buses to said one of said second plurality of buses.

3. The processor as recited in claim 1 further comprising a plurality of functional units configured to perform instruction operations according to said instructions executed by said instruction sequencer.

4. The processor as recited in claim 3 wherein one of said plurality of functional units comprises a multiply and accumulate unit configured to multiply a pair of operands and accumulate a resulting product with other products calculated therein.

5. The processor as recited in claim 3 wherein each one of said plurality of functional units is coupled to at least one of said second plurality of buses to receive operands upon which to perform said instruction operations.

6. The processor as recited in claim 5 wherein said one of said plurality of functional units is configured to store said operands in at least one input register coupled to said at least one of said second plurality of buses.

7. The processor as recited in claim 5 wherein said each one of said plurality of functional units is coupled to one of said first plurality of buses to convey a result of performing said instruction operations.

8. The processor as recited in claim 7 wherein said each one of said plurality of functional units is configured to store said result in an output register prior to conveying said result upon said one of said first plurality of buses.

9. The processor as recited in claim 3 further comprising a register file coupled to said bus interconnect, wherein said register file is configured to store a plurality of data values for use by said plurality of functional units.

10. The processor as recited in claim 9 wherein said register file is coupled to a third bus, and wherein said register file is configured to provide and receive said plurality of data values upon said third bus.

11. The processor as recited in claim 10 wherein said bus interconnect is configured to convey data from one of said first plurality of buses to said third bus upon assertion of a second control signal from said instruction sequencer, said instruction sequencer being configured to assert said second control signal in response to a second encoding of said bus configuration field.

12. The processor as recited in claim 9 further comprising a first data memory and a second data memory, wherein said first data memory and said second data memory are coupled to said bus interconnect, and wherein said first and second data memories are configured to store a second plurality of data values and a third plurality of data values, respectively.

13. The processor as recited in claim 12 wherein said first data memory is configured to provide and receive said second plurality of data values upon a fourth bus, and wherein said second data memory is configured to provide and receive said third plurality of data values upon a fifth bus.

14. The processor as recited in claim 13 wherein said bus interconnect is configured to route data from one of said first plurality of buses to said fourth bus upon assertion of a third control signal from said instruction sequencer, and wherein said bus interconnect is configured to route data from one of said first plurality of buses to said fifth bus upon assertion of a fourth control signal from said instruction sequencer, and wherein said instruction sequencer is configured to assert said third control signal in response to a third encoding of said bus configuration field, and wherein said instruction sequencer is configured to assert said fourth control signal in response to a fourth encoding of said bus configuration field.

15. A method for efficient data routing in a processor, comprising:

encoding a connection between a first output of a first functional unit and a first input of a second functional unit within a first field of an instruction;

detecting said first field upon execution of said instruction;

asserting at least one control signal upon detection of said first field;

routing data from said first output to said first input in a bus interconnect upon assertion of said at least one control signal, and

preventing routing of data from said first output to said first input upon deassertion of said at least one control signal.

16. The method as recited in claim 15 wherein said bus interconnect comprises a crossbar interconnect connecting a plurality of outputs of a plurality of functional units to a plurality of inputs from said plurality of functional units.

17. The method as recited in claim 15 further comprising routing data from said first output to a register file upon assertion of a second control signal.

18. The method as recited in claim 15 further comprising routing data from said first output to a first and second data memories upon assertion of a third and fourth control signal, respectively.