Verifying data integrity of a non-volatile memory system during data caching process

• US 8,037,380 B2
• Filed: 07/08/2008
• Issued: 10/11/2011
• Est. Priority Date: 07/08/2008
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Elliott, John C., Kubo, Robert A., Lucas, Gregg S., Cagno, Brian J.
• Primary Examiner(s): BRITT, CYNTHIA H
• Application Number:
  US12/169,273
• Publication Number:
  US 20100011261A1
• Time in Patent Office: 1,190 Days
• Field of Search: 714/24, 714/25, 714/718, 714/723, 711/118, 365/201, 365/228
• US Class Current: 714/718
• CPC Class Codes: G06F 11/1064 in cache or content addressable memories
• CPC Class Codes: G11C 2029/0407 on power on
• CPC Class Codes: G11C 29/56 External testing equipment for static stores, e.g. automatic test equipment [ATE]; Interfaces therefor

Abstract

To ensure integrity of non-volatile flash, the controller programs the non-volatile memories with background test patterns and verifies the non-volatile memories during power on self test (POST) operation. In conjunction with verifying the non-volatile memories, the controller may routinely run diagnostics and report status to the storage controller. As part of the storage controller power up routines, the storage controller issues a POST command to the controller via an I²C register that is monitored by the storage controller. The storage controller may determine that the non-volatile flash is functional without any defects, and the controller may remove power from the non-volatile flash to extend its reliability. Periodically, in the background, the controller may run diagnostic routines to detect any failures associated with the volatile memory and the controller itself.

Claims

1. A computer program product comprising a computer recordable medium having a computer readable program recorded thereon, wherein the computer readable program, when executed on a memory system controller, causes the memory system controller to:

responsive to powering on a storage controller, wherein the storage controller comprises a volatile memory and a non-volatile memory, the volatile memory stores cache data for input/output operations performed by the storage controller, and the non-volatile memory hardens cache data from the volatile memory responsive to a power event, and concurrent with the storage controller performing input/output operations using the volatile memory, verify the non-volatile memory, wherein verifying the non-volatile memory comprises;

programming the non-volatile memory with a set of test patterns;

reading the set of test patterns from the non-volatile memory;

verifying whether the non-volatile memory stored the set of test patterns correctly; and

posting a failure status to the storage controller responsive to the non-volatile memory storing the set of test patterns incorrectly.

2. The computer program product of claim 1, wherein the computer readable program further causes the computing device to:

run background diagnostics concurrent with the storage controller performing input/output operations.

3. The computer program product of claim 2, wherein running background diagnostics comprises:

snooping an access to the volatile memory system;

responsive to the access being a cache read, directing read data to internal buffer locations of a non-volatile memory controller; and

validating error correction code information in the non-volatile memory controller to validate the internal data pathing within the non-volatile memory controller.

4. The computer program product of claim 2, wherein running background diagnostics comprises:

responsive to the storage controller writing to the volatile memory, reading the written data from the volatile memory; and

determining whether a fault exists in the written data.

5. The computer program product of claim 2, wherein running background diagnostics comprises:

incrementing a programming cycles counter responsive to verifying whether the non-volatile memory stored the set of test patterns correctly; and

switching to an alternate background diagnostic routine responsive to the programming cycles counter reaching a predetermined value.

6. The computer program product of claim 5, wherein the alternate background diagnostic routine comprises:

sending write commands to the non-volatile memory;

verifying that the write commands are acknowledged; and

reading vendor identification information from the non-volatile memory.

7. The computer program product of claim 1, wherein verifying the non-volatile memory further comprises:

performing a bulk erase to the non-volatile memory responsive to the non-volatile memory storing the set of test patterns correctly; and

sending a command to the storage controller to confirm that the non-volatile memory is pre-programmed.

8. The computer program product of claim 7, wherein verifying the non-volatile memory further comprises:

responsive to sending the command to the storage controller, power down the non-volatile memory.

9. The computer program product of claim 1, wherein posting a failure status to the storage controller comprises:

posting a status code in an inter-integrated circuit interface register that is monitored by the storage controller.

10. A method, in a memory system, for verifying data integrity of a non-volatile memory system during data caching process, the method comprising:

responsive to powering on a storage controller, wherein the storage controller comprises a volatile memory and a non-volatile memory, the volatile memory stores cache data for input/output operations performed by the storage controller, and the non-volatile memory hardens cache data from the volatile memory responsive to a power event, and concurrent with the storage controller performing input/output operations using the volatile memory, verify, by a memory system controller in the memory system, the non-volatile memory, wherein verifying the non-volatile memory comprises;

programming a non-volatile memory with a set of test patterns;

reading the set of test patterns from the non-volatile memory;

verifying whether the non-volatile memory stored the set of test patterns correctly; and

posting a failure status to the storage controller responsive to the non-volatile memory storing the set of test patterns incorrectly.

11. The method of claim 10, further comprising:

running, by the memory system controller, background diagnostics concurrent with the storage controller performing input/output operations.

12. The method of claim 11, wherein running background diagnostics comprises:

snooping an access to the volatile memory system;

responsive to the access being a cache read, directing read data to internal buffer locations of a non-volatile memory controller; and

validating error correction code information in the non-volatile memory controller, to validate the internal data pathing within the non-volatile memory controller.

13. The method of claim 11, wherein running background diagnostics comprises:

responsive to the storage controller writing to the volatile memory, reading the written data from the volatile memory; and

determining whether a fault exists in the written data.

14. The method of claim 11, wherein running background diagnostics comprises:

incrementing a programming cycles counter responsive to verifying whether the non-volatile memory stored the set of test patterns correctly; and

switching to an alternate background diagnostic routine responsive to the programming cycles counter reaching a predetermined value.

15. The method of claim 14, wherein the alternate background diagnostic routine comprises:

sending write commands to the non-volatile memory;

verifying that the write commands are acknowledged; and

reading vendor identification information from the non-volatile memory.

16. The method of claim 10, wherein verifying the non-volatile memory further comprises:

performing a bulk erase to the non-volatile memory responsive to the non-volatile memory storing the set of test patterns correctly;

sending a command to the storage controller to confirm that the non-volatile memory is pre-programmed; and

powering down the non-volatile memory.

17. The method of claim 10, wherein posting a failure status to the storage controller comprises:

posting, by the memory system controller, a status code in an inter-integrated circuit interface register that is monitored by the storage controller.

18. A memory system comprising:

a controller, wherein the controller detects a power event from a power supply that provides power to a storage controller and the memory system;

a volatile memory, wherein the volatile memory stores cache data for input/output operations performed by a storage controller; and

a non-volatile memory, wherein the controller hardens the cache data from the volatile memory for the storage controller to the non-volatile memory responsive to the power event;

wherein responsive to powering on the storage controller and concurrent with the storage controller performing input/output operations using the volatile memory, the controller verifies the non-volatile memory, wherein verifying the non-volatile memory comprises;

programming the non-volatile memory with a set of test patterns;

reading the set of test patterns from the non-volatile memory;

verifying whether the non-volatile memory stored the set of test patterns correctly; and

posting a failure status to the storage controller responsive to the non-volatile memory storing the set of test patterns incorrectly.

19. The memory system of claim 18, wherein the controller runs background diagnostics concurrent with the storage controller performing input/output operations.

20. The memory system of claim 18, wherein the controller posts the failure status to the storage controller in an inter-integrated circuit interface register that is monitored by the storage controller.