Method and apparatus for store dependence prediction

US9619750B2 · US · B2

Patent metadata
FieldValue
Publication numberUS-9619750-B2
Application numberUS-201313931872-A
CountryUS
Kind codeB2
Filing dateJun 29, 2013
Priority dateJun 29, 2013
Publication dateApr 11, 2017
Grant dateApr 11, 2017

How to read this patent

A practical reading order for non-experts. Skip the full description unless you need deep technical detail.

  1. Title

    What the patent document calls the invention.

  2. Abstract

    A short plain-language summary of the technical disclosure.

  3. Assignees and inventors

    Who owns or filed the patent and who is credited as inventor.

  4. Key dates

    Filing, priority, publication, and grant dates set the timeline.

  5. First independent claim

    The legal scope of protection — read this for what is actually claimed.

  6. CPC / IPC classifications

    Technology tags used to group this patent with similar filings.

  7. Citations and related patents

    Prior art links and similar publications in this corpus.

Abstract

Official abstract text for this publication.

An apparatus and method for store dependence prediction is described. For example, one embodiment of the invention includes a processor comprising a store buffer for buffering store operations prior to completion, the store operations to store data to a memory hierarchy; and a store dependence predictor to predict whether load operations should be permitted to speculatively skip over each store operation and responsively setting an indication within an entry associated with each store operation in the store buffer; wherein a load operation checks the indication in the store buffer to determine whether to speculatively execute ahead of each store operation.

First claim

Opening claim text (preview).

What is claimed is: 1. A processor comprising: a store buffer for buffering store operations prior to completion, the store operations to store data to a memory hierarchy; and a store dependence predictor to predict whether load operations should be permitted to speculatively skip over each store operation based on a lookup of an array and responsively setting an indication within an entry associated with each store operation in the store buffer, the entry including bits of an instruction address of the store operation, wherein the store dependence predictor comprises the array containing a counter value associated with each store operation, wherein the store dependence predictor decrements a counter for a store operation each time a load operation hits an address of the store operation and increments the counter each time a load operation does not hit the address of the store operation; wherein a load operation checks the indication in the store buffer to determine whether to speculatively execute ahead of each store operation. 2. The processor as in claim 1 wherein the indication comprises a bit for each entry which the store dependence predictor sets for store operations which may be skipped by load operations. 3. The processor as in claim 1 wherein the array comprises an array indexed by a current instruction address or a portion of the current instruction address of each of the store operations. 4. The processor as in claim 1 further comprising threshold comparison logic to determine whether a counter threshold has been reached for a new store operation, the threshold comparison logic responsively setting the indication within the entry associated with the store operation in the store buffer. 5. The processor as in claim 1 further comprising: a load-based dependence predictor to predict whether specified load operations may skip over store operations using one or more load-based dependence prediction techniques. 6. A method comprising: buffering store operations prior to completion, the store operations to store data to a memory hierarchy; performing a lookup in an array containing a counter value associated with each store operation, wherein the counter is decremented for a store operation each time a load operation hits an address of the store operation and incremented each time a load operation does not hit the address of the store operation; and predicting whether load operations should be permitted to speculatively skip over each store operation based on the lookup and responsively setting an indication within an entry associated with each store operation in the store buffer, the entry including bits of an instruction address of the store operation; wherein a load operation checks the indication in the store buffer to determine whether to speculatively execute ahead of each store operation. 7. The method as in claim 6 wherein the indication comprises a bit for each entry which the store dependence predictor sets for store operations which may be skipped by load operations. 8. The method as in claim 6 wherein the array comprises an array indexed by a current instruction address or a portion of a current instruction address of each of the store operations. 9. The method as in claim 6 further comprising determining whether a counter threshold has been reached for a new store operation and responsively setting the indication within the entry associated with the store operation in the store buffer. 10. The method as in claim 6 further comprising: predicting whether specified load operations may skip over store operations using one or more load-based dependence prediction techniques.

Assignees

Inventors

Classifications

  • Speculative instruction execution · CPC title

  • G06F9/3834Primary

    Maintaining memory consistency · CPC title

  • G06N5/04Primary

    Inference or reasoning models · CPC title

  • Physics · mapped topic

  • Reordering of instructions, e.g. using queues or age tags · CPC title

Patent family

Related publications grouped by family.

External sources

Frequently asked questions

Answers are generated from the same data shown on this page.

What does patent US9619750B2 cover?
An apparatus and method for store dependence prediction is described. For example, one embodiment of the invention includes a processor comprising a store buffer for buffering store operations prior to completion, the store operations to store data to a memory hierarchy; and a store dependence predictor to predict whether load operations should be permitted to speculatively skip over each store…
Who is the assignee on this patent?
Intel Corp
What technology area does this patent fall under?
Primary CPC classification G06F9/3834. Mapped technology areas include Physics.
When was this patent published?
Publication date Tue Apr 11 2017 00:00:00 GMT+0000 (Coordinated Universal Time) (B2). Legal status and post-grant events are not shown on this page.
What related patents are in patentsdb?
We list 3 related publications on this page (citations in our corpus or others sharing the same primary CPC).