Computer processor with generation renaming

• US 9,710,272 B2
• Filed: 10/31/2014
• Issued: 07/18/2017
• Est. Priority Date: 04/25/2014
• Status: Active Grant

• Current Assignee: Avago Technologies International Sales Pte Limited (Broadcom, Inc.)
• Original Assignee: Avago Technologies General IP PTE Limited (Broadcom, Inc.)
• Inventors: Wilson, Sophie, Redford, John, Barrett, Geoffrey, Kurd, Tariq
• Primary Examiner(s): Partridge, William B
• Application Number:
  US14/530,370
• Publication Number:
  US 20150309797A1
• Time in Patent Office: 991 Days
• Field of Search: None
• CPC Class Codes: G06F 9/30043 LOAD or STORE instructions; Clear instruction
• CPC Class Codes: G06F 9/30058 Conditional branch instructions
• CPC Class Codes: G06F 9/30065 Loop control instructions; iterative instructions, e.g. LOOP, REPEAT
• CPC Class Codes: G06F 9/30098 Register arrangements
• CPC Class Codes: G06F 9/3013 according to data content, e.g. floating-point registers, address registers
• CPC Class Codes: G06F 9/30145 Instruction analysis, e.g. decoding, instruction word fields
• 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 9/3832 Value prediction for operands; operand history buffers
• CPC Class Codes: G06F 9/3834 Maintaining memory consistency
• CPC Class Codes: G06F 9/3836 Instruction issuing, e.g. dynamic instruction scheduling or out of order instruction execution
• CPC Class Codes: G06F 9/384 Register renaming
• CPC Class Codes: G06F 9/3842 Speculative instruction execution
• CPC Class Codes: G06F 9/3848 using hybrid branch prediction, e.g. selection between prediction techniques
• CPC Class Codes: G06F 9/3858 Result writeback, i.e. updating the architectural state or memory

Abstract

A processor including a register file having a plurality of registers, and configured for out-of-order instruction execution, further includes a renamer unit that produces generation numbers that are associated with register file addresses to provide a renamed version of a register that is temporally offset from an existing version of that register rather than assigning a non-programmer-visible physical register as the renamed register. The processor includes a small reset DHL Gshare branch prediction unit coupled to an instruction cache and configured to provide speculative addresses to the instruction cache.

Claims

1. A processor, comprising:

a reservation queue;

a renamer coupled to the reservation queue and configured to produce a generation number and to communicate the generation number to the reservation queue;

a register coupled to the reservation queue and configured to store a value;

an execution pipe coupled to the reservation queue; and

a stunt box coupled to the execution pipe, the stunt box comprising;

a first storage, including a plurality of registers, configured to temporarily store an execution pipe result as an intermediate result;

a second storage, communicatively coupled to the first storage, configured to receive and store a plurality of storage results which includes one or more of the intermediate results;

a first arbiter, communicatively coupled to the first storage and the second storage, configured to receive a plurality of intermediate results, and the plurality of storage results and to select an output of the first arbiter to retire from the plurality of intermediate results, and the plurality of storage results;

a second arbiter, communicatively coupled to receive execution pipe results and the output from the first arbiter; and

a third storage, communicatively coupled to receive a plurality of speculative execution pipe results from the second arbiter.

2. The processor of claim 1, further comprising:

a small reset DHL Gshare branch prediction unit.

3. The processor of claim 1, further comprising:

a load/store unit configured to schedule access to memory for each of a plurality of queued load and store instructions;

wherein the load/store unit includes a resource locking circuit configured to prevent clashing memory operations from being concurrently scheduled.

4. A processor, comprising:

an instruction cache;

an instruction decoder coupled to the instruction cache;

a reservation queue coupled to the instruction decoder;

a register coupled to the reservation queue and configured to store a value;

a renamer coupled to the reservation queue and configured to produce a generation number and to communicate the generation number to the reservation queue;

a small reset DHL Gshare branch prediction unit coupled to the instruction cache and configured to generate an address and communicate the address to the instruction cache;

a plurality of execution pipes coupled to the reservation queue; and

a stunt box coupled to each execution pipe of the plurality of execution pipes, the stunt box comprising;

a first storage, including a plurality of registers, configured to temporarily store a plurality of execution pipe results as a plurality of intermediate results;

a second storage, communicatively coupled to the first storage, configured to receive and store a plurality of storage results which includes one or more of the plurality of intermediate results;

a first arbiter, communicatively coupled to the first storage and the second storage, configured to receive the plurality of intermediate results, and the plurality of storage results and to select an output of the first arbiter to retire from the plurality of intermediate results, and the plurality of storage results;

a second arbiter, communicatively coupled to receive execution pipe results and the output from the first arbiter; and

a third storage, communicatively coupled to receive a plurality of speculative execution pipe results from the second arbiter.

5. A processor, comprising:

an instruction cache;

an instruction decoder coupled to the instruction cache;

a branch prediction unit coupled to the instruction cache and configured to generate an instruction address and provide the instruction address to the instruction cache;

a register file coupled to receive a register address;

a renamer unit coupled to the instruction cache and to the instruction decoder;

a plurality of reservation queues, each reservation queue of the plurality of reservation queues coupled to the register file, the instruction decoder, and the renamer unit;

a plurality of execution pipes, each execution pipe of the plurality of execution pipes coupled to a corresponding one of the plurality of reservation queues, and configured to receive an instruction and data from the corresponding one of the plurality of reservation queues, and to execute the instruction;

a stunt box coupled to each execution pipe of the plurality of execution pipes, and configured to receive an output of each execution pipe of the plurality of execution pipes the stunt box comprising;

a first storage, including a plurality of registers, configured to temporarily store a plurality of execution pipe results as a plurality of intermediate results;

a second storage, communicatively coupled to the first storage, configured to receive and store a plurality of storage results which includes one or more of the plurality of intermediate results;

a first arbiter, communicatively coupled to the first storage and the second storage, configured to receive the plurality of intermediate results and the plurality of storage results, and to select an output of the first arbiter to retire from the plurality of intermediate results, and the plurality of storage results;

a second arbiter, communicatively coupled to receive execution pipe results and the output from the first arbiter; and

a third storage, communicatively coupled to receive a plurality of speculative execution pipe results from the second arbiter;

wherein, responsive to an output of the instruction decoder, the register file is configured to provide contents of a register specified as a source in an instruction;

and responsive to the output of the instruction decoder, the renamer is configured to concurrently provide a generation number.

6. The processor of claim 5, wherein the branch prediction unit includes a small reset DHL Gshare branch predictor.

7. The processor of claim 5, wherein the instruction cache is configured to receive and store VLIW instructions.

8. The processor of claim 5, wherein an output of the stunt box is coupled to an input of each of the execution pipes of the plurality of execution pipes, and further coupled to an input of each of the reservation queues of the plurality of reservation queues.

9. The processor of claim 8, further comprising:

a load/store unit coupled the stunt box and to the register file; and

a memory coupled to the load/store unit;

wherein the load/store unit is configured to receive input from the stunt box and to receive input from the register file.

10. The processor of claim 5, wherein each reservation queue of the plurality of reservation queues is configured to store an instruction opcode, a destination register address, and a destination register generation number.

11. The processor of claim 10, wherein each reservation queue of the plurality of reservation queues is further configured to store a first source register address, a first register generation number, a first register value, a first valid bit, a second source register address, a second register generation number, a second register value, and a second valid bit.

12. The processor of claim 11, wherein the instruction decoder identifies an instruction that specifies a paired register operation.

13. The processor of claim 5, wherein the register file comprises a plurality of registers;

and the renamer unit comprises a register renamer block for each of the plurality of registers.

14. The processor of claim 5, wherein the stunt box is configured to concurrently place a pair-wide output and a single-wide output onto an operand copy network.

15. The processor of claim 5, further comprising an instruction fetch unit.

16. The processor of claim 15, wherein the instruction fetch unit comprises:

a loopcount register configured to store a either a loop count value or a speculative loop count value;

a loopstart register configured to store a loop start address; and

a loopmatch register configured to store a loop match address.

17. The processor of claim 15, wherein the instruction fetch unit is coupled to the instruction decoder and a SIN allocator, and configured to provide a signal to the SIN allocator indicative of a loop count value representing a speculative loop execution.

18. The processor of claim 16, wherein the instruction decoder is coupled to the loopcount register, the loopstart register and the loopmatch register.

19. The processor of claim 15, wherein the instruction fetch unit is configured to generate a loop match signal.

20. The processor of claim 19, wherein the instruction decoder is configured to receive the loop match signal from the instruction fetch unit.