Performing predecode-time optimized instructions in conjunction with predecode time optimized instruction sequence caching

• US 9,378,022 B2
• Filed: 12/09/2013
• Issued: 06/28/2016
• Est. Priority Date: 03/28/2012
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Gschwind, Michael K., Salapura, Valentina
• Primary Examiner(s): Geib, Benjamin
• Application Number:
  US14/100,044
• Publication Number:
  US 20140095835A1
• Time in Patent Office: 932 Days
• US Class Current: 1/1
• CPC Class Codes: G06F 9/3017 Runtime instruction translation, e.g. macros
• CPC Class Codes: G06F 9/3808 for instruction reuse, e.g. trace cache, branch target cache
• CPC Class Codes: G06F 9/382 Pipelined decoding, e.g. using predecoding
• CPC Class Codes: G06F 9/384 Register renaming

Abstract

A method for performing predecode-time optimized instructions in conjunction with predecode time optimized instruction sequence caching. The method includes receiving a first instruction of an instruction sequence and a second instruction of the instruction sequence and determining if the first instruction and the second instruction can be optimized. In response to the determining that the first instruction and second instruction can be optimized, the method includes, preforming a pre-decode optimization on the instruction sequence and generating a new second instruction, wherein the new second instruction is not dependent on a target operand of the first instruction and storing a pre-decoded first instruction and a pre-decoded new second instruction in an instruction cache. In response to determining that the first instruction and second instruction can not be optimized, the method includes, storing the pre-decoded first instruction and a pre-decoded second instruction in the instruction cache.

Claims

1. A computer program product for performing predecode-time optimized instructions in conjunction with predecode time optimized instruction sequence caching, the computer program product comprising:

a non-transitory storage medium readable by a processing circuit and storing instructions for execution by the processing circuit for performing a method comprising;

receiving a first instruction of an instruction sequence by a first decoder and a second instruction of the instruction sequence by a second decoder, wherein the first instruction is received from an instruction cache that is a level one cache;

decoding the first instruction of the instruction sequence by the first decoder to create a decoded first instruction;

decoding the second instruction of the instruction sequence by the second decoder to create a decoded second instruction;

determining, by an optimization analysis engine in communication with the first decoder and the second decoder, if the first instruction and the second instruction can be optimized based on the first decoded instruction and the second decoded instruction;

responsive to the determining that the first instruction and second instruction can be optimized;

performing a pre-decode optimization on the instruction sequence and generating, by the second decoder;

a new decoded second instruction, wherein the new decoded second instruction is not dependent on a target operand of the first instruction; and

storing the pre-decoded first instruction and the new decoded second instruction directly in the instruction cache;

responsive to the determining that the first instruction and second instruction can not be optimized, storing the pre-decoded first instruction and a pre-decoded second instruction directly in the instruction cache and storing an entry point in an entry point vector stored in the instruction cache, the entry point corresponding to the first decoded instruction.

2. The computer program product of claim 1, further comprising storing an entry point corresponding to the first instruction to an entry point vector.

3. The computer program product of claim 2, wherein the entry point vector is stored in the instruction cache.

4. The computer program product of claim 1, wherein the new second instruction specifies the source operand locations of the first instruction and at least one source operand location of the second instruction.

5. The computer program product of claim 1, wherein the target operand location is a first target register of the first instruction and the source operand location is a source register of the second instruction.

6. A method for performing predecode-time optimized instructions in conjunction with predecode time optimized instruction sequence caching, the computer program product comprising:

a storage medium readable by a processing circuit and storing instructions for execution by the processing circuit for performing a method comprising;

receiving a first instruction of an instruction sequence by a first decoder and a second instruction of the instruction sequence by a second decoder, wherein the first instruction is received from an instruction cache that is a level one cache;

decoding the first instruction of the instruction sequence by the first decoder to create a decoded first instruction;

decoding the second instruction of the instruction sequence by the second decoder to create a decoded second instruction;

determining, by an optimization analysis engine in communication with the first decoder and the second decoder, if the first instruction and the second instruction can be optimized based on the first decoded instruction and the second decoded instruction;

responsive to the determining that the first instruction and second instruction can be optimized;

performing a pre-decode optimization on the instruction sequence and generating, by the second decoder;

a new decoded second instruction, wherein the new decoded second instruction is not dependent on a target operand of the first instruction; and

storing the first decoded instruction and the new decoded second instruction directly in an instruction cache;

responsive to the determining that the first instruction and second instruction can not be optimized, storing the pre-decoded first instruction and a pre-decoded second instruction directly in the instruction cache and storing an entry point in an entry point vector stored in the instruction cache, the entry point corresponding to the first decoded instruction.

7. The method of claim 6, further comprising storing an entry point corresponding to the first instruction to an entry point vector.

8. The method of claim 7, wherein the entry point vector is stored in the instruction cache.

9. The method of claim 6, wherein the new second instruction specifies the source operand locations of the first instruction and at least one source operand location of the second instruction.

10. The method of claim 6, wherein the target operand location is a first target register of the first instruction and the source operand location is a source register of the second instruction.