Method for copying data in reprogrammable non-volatile memory
View Patent ↗Methods are presented for improving data relocation operations. Rather than check data quality based on its associated error correction code (ECC) in every relocation operation, it can determine whether to check ECC based on predetermined selection criteria: if ECC checking is not selected, the memory can perform an on-chip copy of the data from a first location to a second location. If ECC checking is selected, the data is transferred to the controller and checked; when an error is found, a correction operation is performed and when no error is found, an on-chip copy is performed. Predetermined selection criteria may comprise a sampling mechanism, which may be random based or deterministic. Additionally, data transfer flags can be introduced to indicate data has been corrected and should be transferred back to the memory. Header and user data can be considered separately if each has a distinct associated ECC.
1. A method for copying data having an associated error correction code (ECC) from a first location to a second location of a reprogrammable non-volatile memory array of a system that includes a memory controller and a memory portion having said memory array, comprising:
reading the data from the first location into a cache on the memory portion;
determining whether to check the ECC based on predetermined selection criteria, and if ECC checking is not selected, causing the system to copy the data from the cache to the second location without transferring the data to the controller; and
if ECC checking is selected, checking the ECC on the read data in the controller, including:
if an ECC error is found, correcting the ECC error on the read data in the controller and causing the controller to write the corrected data to the second location of the memory array; and
if an ECC error is not found, causing the memory to copy the data from the cache to the second location without transferring the data to the controller.
2. The method of claim 1 , wherein the predetermined selection criteria comprises a sampling mechanism.
3. The method of claim 2 , wherein the sampling mechanism is random based.
4. The method of claim 2 , wherein the sampling mechanism is deterministic.
5. The method of claim 2 , wherein the sampling mechanism comprises:
comparing a plurality of bits associated with a the number of a logical sector in the reprogrammable non-volatile memory with a timestamp to produce a comparison result; and
selecting to check an ECC in accordance with the comparison result.
6. The method according to claim 2 , wherein said sampling mechanism is based on overhead data associated with said data.
7. The method of claim 6 , wherein said sampling mechanism comprises comparing a time stamp associated with the data to logical address information of the data.
8. The method of claim 2 , wherein said determining whether to check ECC is randomly or pseudo-randomly based.
9. The method of claim 1 , further comprising setting an error flag associated with the data if the error correction is not successfully completed.
10. The method of claim 9 , wherein the data comprises a header portion and a user data portion and said error flag is written into the header portion.
11. The method of claim 1 , wherein said determining whether to check ECC is a deterministic mechanism.
12. The method of claim 1 , further comprising:
setting a flag value based on whether the data has been corrected as part of said correcting the ECC error;
if the flag value is set to indicate that said data has been corrected, transferring the corrected data back to the cache and subsequently writing the contents of the cache to the second memory location.
13. The method of claim 12 , wherein the correcting the ECC error comprises correcting the copy of the read data using the associated ECC, wherein said flag value is set in response to a successfully completing said correcting.
14. The method of claim 12 , wherein the data comprises a user data portion and an overhead portion and wherein the user data has an associated ECC and the overhead portion has a distinct associated ECC wherein said transferring a copy of the data, said ECC checking and correcting comprises:
transferring a copy of the overhead portion of the data from the cache to the controller while maintaining said data in the cache; and
performing in the controller an error detection and correction operation on the overhead portion of the data based on the associated overhead ECC and, in response to completing a successful correction operation on the overhead portion of the data, transferring the corrected overhead portion back to the cache;
and in response to a successful error detection and correction operation on the overhead portion of the data;
transferring a copy of the user portion of the data from the cache to the controller while maintaining said data in the cache; and
performing in the controller an error detection and correction operation on the user portion of the data based on the associated user data ECC,
where said flag value is set based on whether the user data portion has been corrected.