Method of handling page fault in nonvolatile main memory system

• US 10,719,263 B2
• Filed: 11/30/2016
• Issued: 07/21/2020
• Est. Priority Date: 12/03/2015
• Status: Active Grant

• Current Assignee: Samsung Electronics Co. Ltd., SungKyunKwan University's Research & Business Foundation (SungKyunKwan University (SKKU))
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: Choi, Jung Sik, Han, Hwan Soo
• Primary Examiner(s): Thammavong, Prasith
• Application Number:
  US15/364,628
• Publication Number:
  US 20170160991A1
• Time in Patent Office: 1,329 Days
• CPC Class Codes: G06F 11/00 Error detection; Error correction; Monitoring error detection, correction or monitoring in information storage based on relative movement between record carrier and transducer G11B20/18; monitoring, i.e. supervising the progress of recording or reproducing G11B27/36; in static stores G11C29/00
• CPC Class Codes: G06F 11/073 in a memory management context, e.g. virtual memory or cache management memory management G06F12/00; testing of static memory units G11C29/00
• CPC Class Codes: G06F 3/0619 in relation to data integrity, e.g. data losses, bit errors
• CPC Class Codes: G06F 3/0653 Monitoring storage devices or systems
• CPC Class Codes: G06F 3/0659 Command handling arrangements, e.g. command buffers, queues, command scheduling
• CPC Class Codes: G06F 3/0679 Non-volatile semiconductor memory device, e.g. flash memory, one time programmable memory [OTP]

Abstract

A method of handling a page fault occurring in a non-volatile main memory system including analyzing a pattern of occurrence of the page fault based on the page fault when the page fault occurs, setting the number of pages to be consecutively processed based on analysis result of the analyzing, and consecutively processing as many pages as the number may be provided.

Claims

1. A method of handling a page fault occurring in a non-volatile main memory system, the method comprising:

analyzing a page fault occurrence pattern referring to previous page information having a page fault, in response to an occurrence of a page fault;

setting a first number of pages to be consecutively processed based on a result of the analyzing; and

consecutively processing as many pages as the first number,

wherein the consecutively processing includes

synchronously processing a second number of pages,

asynchronously processing a third number of pages, and

a sum of the second number of pages and the third number of pages is equal to the first number of pages to be consecutively processed.

2. The method of claim 1, wherein the analyzing includes determining whether the page fault occurrence pattern is sequential.

3. The method of claim 1, wherein the setting comprises:

increasing the first number in response to the page fault occurrence pattern being sequential; and

decreasing the first number in response to the page fault occurrence pattern being random.

4. The method of claim 1, wherein the synchronously processing is performed in a kernel mode.

5. The method of claim 4, wherein the asynchronously processing is performed in a user mode.

6. The method of claim 4, wherein the asynchronously processing comprises determining whether the page fault occurs each time one of the third number of pages is processed.

7. The method of claim 6, further comprising:

decreasing or maintaining a ratio of the third number of pages to the second number of pages in response to the page fault occurring while performing the asynchronously processing; and

increasing the ratio of the third number of pages to the second number of pages in response to the page fault not occurring while performing the asynchronously processing.

8. The method of claim 1, further comprising:

determining whether the page fault has occurred before the analyzing.

9. The method of claim 8, further comprising:

predicting page fault occurrence using hint information, the hint information obtained from a system call in response to a non-occurrence of the page fault;

setting a fourth number of pages to be consecutively processed based on the hint information; and

consecutively processing as many pages as the fourth number.

10. A method of handling a page fault occurring in a non-volatile main memory system, the method comprising:

setting a first number of pages to be consecutively processed based on a result of analyzing a page fault occurrence pattern referring to previous page information having a page fault, in response to an occurrence of a page fault;

synchronously processing a second number of pages; and

asynchronously processing a third number of the pages,

wherein a sum of the second number of pages and the third number of pages is equal to the first number of pages to be consecutively processed.

11. The method of claim 10, wherein the analyzing the page fault occurrence pattern includes determining whether the page fault occurrence pattern is sequential.

12. The method of claim 10, wherein the setting comprises:

increasing the first number of pages to be consecutively processed in response to the page fault occurrence pattern being sequential; and

decreasing the first number of pages to be consecutively processed in response to the page fault occurrence pattern being random.

13. The method of claim 10, wherein the synchronously processing is performed in a kernel mode, and the asynchronously processing is performed in a user mode.

14. The method of claim 10, wherein the non-volatile main memory system comprises a memory cell array, and the memory cell array comprises three-dimensional memory cells.

15. A method of handling a page fault occurring in a non-volatile main memory system, the method comprising:

determining whether a page fault has occurred;

analyzing a page fault occurrence pattern referring to previous page information having a page fault to determine whether the page fault occurrence pattern is sequential, in response to a result of the determining indicating that the page fault has occurred;

setting a first number of pages to be consecutively processed based on a result of the analyzing; and

consecutively processing as many pages as the first number,

wherein the consecutively processing includes

synchronously processing a second number of pages,

asynchronously processing a third number of pages, and

a sum of the second number of pages and the third number of pages is equal to the first number of pages to be consecutively processed.

16. The method of claim 15, further comprising:

decreasing or maintaining a ratio of the third number of pages to the second number of pages in response to the page fault occurring while performing the asynchronously processing; and

increasing the ratio of the third number of pages to the second number of pages in response to the page fault not occurring while performing the asynchronously processing.

17. The method of claim 15, wherein the first number is set based on a plurality of predefined data states.

18. The method of claim 15 further comprising:

predicting page fault occurrence using hint information, the hint information obtained from a system call in response to a non-occurrence of the page fault;

setting a second number of pages to be consecutively processed ages based on the hint information; and

consecutively processing as many pages as the second number.