Method of managing defects in flash disk memories

• US 5,473,753 A
• Filed: 10/30/1992
• Issued: 12/05/1995
• Est. Priority Date: 10/30/1992
• Status: Expired due to Term

• Current Assignee: Intel Corporation
• Original Assignee: Intel Corporation
• Inventors: Wells, Steven E., Hasbun, Robert N., Magnusson, Eric J.
• Primary Examiner(s): Beausoliel, Jr., Robert W.
• Assistant Examiner(s): Snyder, Glenn
• Application Number:
  US07/969,749
• Time in Patent Office: 1,131 Days
• Field of Search: 371/13, 371/10.1, 371/10.2, 371/12, 371/7, 371/8.1, 365/218, 395/182.03
• US Class Current: 714/5.1
• CPC Class Codes: G06F 11/1068 in sector programmable memories, e.g. flash disk G06F11/1072 takes precedence
• CPC Class Codes: G11C 2029/0409 Online test
• CPC Class Codes: G11C 29/4401 for self repair
• CPC Class Codes: G11C 29/765 in solid state disks
• CPC Class Codes: G11C 29/88 with partially good memories

Abstract

A method for monitoring the operations of a flash memory array divided into individually erasable blocks of memory in order to assure the integrity of data stored in the array in which each read or write operation is verified to detect an error which may have occurred in the operation including the steps of attempting at least one retry operation whenever an error occurs to determine whether the error is repeatable, marking the block to indicate valid data should be removed from the block if the error is found to be repeatable, removing the valid information from the block if the error is found to be repeatable, and removing a block with a repeatable error from operation.

Claims

1. A method for write defect management in a memory array divided into individually erasable blocks of memory, said method comprising the steps of:

generating a block status table for each block in each memory array;

executing a program operation with data to a block of memory at a first offset address;

determining whether an error occurred in said program operation to said block;

checking said block status table to determine whether a previous error occurred in said block if an error occurred in said program operation;

generating a first questionable indication for said block in said block status table including identifying said first offset address in said block if an error occurred in said program operation and if said block has no previous questionable indications at said first offset address;

allocating a new offset address in said block for said data if an error occurred in said program operation and if said status table indicates said block was previously marked questionable at said first offset address;

generating a second questionable indication for said block in said block status table if an error occurred in said program operation, and said block has a previous questionable indication but said block has no previous questionable indications at said block offset address; and

marking said block bad if an error occurred in said program operation and if said block has two previous questionable indications; and

allocating a new block for said data if an error occurred in said program operation.

2. The method as set forth in claim 1, further comprising the steps of:

determining whether a high programming voltage is low if an error occurred in said program operation;

disabling write operations to said memory array if said high program voltage is low; and

executing, upon request, read operations to said block.

3. The method as set forth in claim 1, wherein the step of determining whether an error occurred in said program operation to said block comprises the steps of:

determining whether an error occurred in said program operation when programming a header field;

reading a previous header field in said block if said program operation error occurred when programming a header field;

writing said previous header field data to a subsequent header position in said block; and

marking said previous header field dirty.

4. The method as set forth in claim 1, further comprising the steps of:

executing a second program operation for said data to a second block if an error occurred in said program operation;

determining whether an error occurred in said second program operation;

disabling write operations to said array if an error occurred in said second program operation; and

executing, upon request, read operations to said array.

5. The method as set forth in claim 1, wherein the step of marking said block bad if an error occurred in said program operation and if said block has two previous questionable indications further comprises the steps of:

flagging said block for a priority cleanup; and

marking said block bad.

6. The method as set forth in claim 1, wherein the step of executing a program operation further comprises the steps of:

generating a plurality of header fields in a first area of said block for data entries;

generating a block storage structure for said block; and

programming data corresponding to said data entries between said header fields and said block storage structure.

7. The method as set forth in claim 6, wherein the step of generating a header field in said block for said data comprises the steps of:

identifying a sector number for said data entry, said sector number being utilized by an operating system to address said data;

identifying whether said data entry is valid;

identifying a revision number for said data entry;

identifying whether said data entry includes valid data;

identifying whether said data entry is compressed; and

identifying a pointer for said data entry including an offset address to specify a location in said block for said data entry.

8. A method for read defect management in a memory comprising a plurality of memory arrays divided into individually erasable blocks of memory, said method comprising the steps

generating a bad block status area on said array;

executing a read operation to a block at an offset address to retrieve data stored in said block;

determining whether an error occurred in said read operation to said block;

signaling a priority clean-up for said block;

marking said block bad in said bad block status area after priority clean-up for said block;

polling said bad block status area on said array to determine a status for any block in said array such that determination of bad blocks on said array is readily ascertainable;

generating a questionable status table on said array;

marking said block questionable in said questionable status table including identifying said offset address if an error occurred in said read operation to said block;

executing an erase operation on said block if marked questionable twice including saving said questionable status table; and

marking said block bad in said bad block status area if said erase operation fails; and

removing said questionable status of said block if said erase operation is successful.

9. The method as set forth in claim 8, further comprising the step of executing said read operation again at a slower speed if an error occurred in said read operation.

10. The method as set forth in claim 8, further comprising the step of performing error correction on said data retrieved in said read operation.