PREFETCH FILTER CACHE FOR A PROCESSOR

• US 20170344483A1
• Filed: 05/31/2016
• Published: 11/30/2017
• Est. Priority Date: 05/31/2016
• Status: Active Grant

• Current Assignee: Intel Corporation
• Original Assignee: Intel Corporation
• Inventors: SHWARTSMAN, STANISLAV, RAIS, RON
• Granted Patent: US 10,579,530 B2
• Time in Patent Office: Days
• CPC Class Codes: G06F 12/0862 with prefetch
• CPC Class Codes: G06F 12/0875 with dedicated cache, e.g. instruction or stack
• CPC Class Codes: G06F 2212/1016 Performance improvement
• CPC Class Codes: G06F 2212/602 Details relating to cache prefetching
• CPC Class Codes: G06F 2212/6022 Using a prefetch buffer or dedicated prefetch cache
• CPC Class Codes: Y02D 10/00 Energy efficient computing, e.g. low power processors, power management or thermal management

Abstract

In an embodiment, a processor includes a plurality of cores, with at least one core including prefetch logic. The prefetch logic comprises circuitry to: receive a prefetch request; compare the received prefetch request to a plurality of entries of a prefetch filter cache; and in response to a determination that the received prefetch request matches one of the plurality of entries of the prefetch filter cache, drop the received prefetch request. Other embodiments are described and claimed.

Claims

1. A processor comprising:

a plurality of cores, at least one core including prefetch logic, the prefetch logic comprising circuitry to;

receive a prefetch request;

compare the received prefetch request to a plurality of entries of a prefetch filter cache; and

in response to a determination that the received prefetch request matches one of the plurality of entries of the prefetch filter cache, drop the received prefetch request.

2. The processor of claim 1, the circuitry further to:

in response to a determination that the received prefetch request does not match any of the plurality of entries of the prefetch filter cache, attempt to execute the received prefetch request.

3. The processor of claim 2, the circuitry further to:

in response to a determination that an attempted execution of the received prefetch request resulted in a prefetch fault, add an entry for the received prefetch request to the prefetch filter cache.

4. The processor of claim 2, the circuitry further to:

upon a successful execution of the received prefetch request, store prefetch data in a cache of the core.

5. The processor of claim 1, wherein each entry of the prefetch filter cache comprises memory address information associated with a previous prefetch request that resulted in a prefetch fault.

6. The processor of claim 5, wherein the circuitry is to compare a memory address of the received prefetch request to the memory address information stored in the plurality of entries of the prefetch filter cache.

7. The processor of claim 1, the circuitry further to:

automatically generate the prefetch request based on an analysis of a program execution.

8. The processor of claim 1, wherein the prefetch filter cache is implemented as a Bloom filter.

9. A machine-readable medium having stored thereon data, which if used by at least one machine, causes the at least one machine to fabricate at least one integrated circuit to perform a method comprising:

receiving a prefetch request for data to be stored in a cache of a processor;

determining whether the prefetch request matches at least one entry of a prefetch filter cache of the processor; and

in response to a determination that the prefetch request matches at least one entry one of the prefetch filter cache, dropping the prefetch request without initiating execution of the prefetch request.

10. The machine-readable medium of claim 9, wherein the method further comprises:

in response to a determination that the prefetch request does not match at least one entry of the prefetch filter cache, initiating an execution of the prefetch request.

11. The machine-readable medium of claim 10, wherein the method further comprises:

in response to a prefetch fault in the execution of the prefetch request, adding an entry associated with the prefetch request to the prefetch filter cache.

12. The machine-readable medium of claim 11, wherein the entry associated with the prefetch request comprises memory address information associated with the prefetch request.

13. The machine-readable medium of claim 9, wherein the method further comprises:

upon completing the execution of the prefetch request, storing prefetch data in the cache of the processor.

14. The machine-readable medium of claim 9, wherein the prefetch filter cache is implemented using a Bloom filter.

15. A system comprising:

a processor comprising prefetch logic, the prefetch logic comprising circuitry to;

receive a prefetch request;

compare the received prefetch request to entries of a prefetch filter cache; and

in response to a determination that the received prefetch request matches an entry of the prefetch filter cache, drop the received prefetch request without attempting to execute the prefetch request; and

a system memory coupled to the processor.

16. The system of claim 15, wherein each entry of the prefetch filter cache includes memory address information associated with a previous prefetch request that resulted in a prefetch fault, and wherein the circuitry is to compare a memory address of the received prefetch request to the memory address information included in the plurality of entries of the prefetch filter cache.

17. The system of claim 15, the circuitry further to:

in response to a determination that the received prefetch request does not match the entries of the prefetch filter cache, attempt to execute the received prefetch request.

18. The system of claim 17, the circuitry further to:

in response to a determination that an attempted execution of the received prefetch request resulted in a prefetch fault, add an entry for the received prefetch request to the prefetch filter cache.

19. The system of claim 15, the circuitry further to:

execute the received prefetch request to read prefetch data; and

store the prefetch data in a cache.

20. The system of claim 15, wherein the prefetch filter cache is implemented as a Bloom filter.