Reduced-overhead error detection and correction

1. A method of operation within a memory control component, the method comprising:

reading from one or more memory components, as part of a data block, a plurality of first-format words each consisting of M+N bits organized within an N-bit payload field and an M-bit error detection/correction field;

reading from the one or more memory components, as part of the data block, a second-format word consisting of M+N bits organized in an (N−1)-bit payload field, an M-bit error detection/correction field, and a one-bit parity field; and

determining, for each of the first-format words, whether an error exists at any of N bit positions within the N-bit payload field based at least a part on the M-bit error detection/correction field.

2. The method of claim 1 wherein the N-bit payload field within each of the first-format words comprises a plurality of non-contiguous subfields.

3. The method of claim of claim 1 wherein the N-bit payload field within each of the plurality of first-format words comprises an (N−1)-bit data field and a one-bit encryption field, the method further comprising decrypting contents of the (N−1)-bit data fields of the plurality of first-format words using an encryption value that includes bits from respective encryption fields of the plurality of first-format words.

4. The method of claim 1 further comprising determining whether an error within the N-bit payload field of one of the first-format words is correctable based at least in part on the parity field within the second-format word.

5. The method of claim 4 wherein determining whether the error within the N-bit payload field of one of the first-format words is correctable comprises determining whether the error is correctable based at least in part on (i) the parity field within the second-format word, (ii) the error detection/correction fields within the others of the first-format words and (iii) the error detection/correction field within the second-format word.

6. The method of claim 1 further comprising identifying, for at least one of the first-format words for which an error is determined to exist within the N-bit payload field, a bit position within the N-bit payload field at which the bit value is in error.

7. The method of claim 6 wherein identifying the bit position within the N-bit payload field of the at least one of the first-format words at which the bit value is in error comprises identifying the bit position based at least in part on the M-bit error detection/correction field within the at least one of the first-format words and the one-bit parity field within the second-format word.

8. The method of claim 1 further comprising identifying, for at least two of the first-format words for which respective errors are determined to exist within respective N-bit payload fields thereof, a respective bit position within the N-bit payload field of each of the at least two of the first-format words at which the bit value is in error.

9. The method of claim 1 wherein determining, for each of the first-format words, whether an error exists at any of N bit positions within the N-bit payload field comprises detecting occurrence of two single-bit errors exist at respective positions of the N-bit payload field of one of the first-format words based on the M-bit error detection/correction field of the one of the first-format words and on the one-bit parity field of the second-format word.

10. The method of claim 1 wherein the N-bit payload field within each of the plurality of first-format words spans an odd number of bits such that reading the plurality of first-format words comprises reading, for each of the plurality of first-format words, an odd number of payload bits.

11. A memory control component comprising:

receiver circuitry to receive a block of data from one or more memory components, the block of data including (i) a plurality of first-format words each consisting of M+N bits organized within an N-bit payload field and an M-bit error detection/correction field and (ii) a second-format word consisting of M+N bits organized in an (N−1)-bit payload field, an M-bit error detection/correction field, and a one-bit parity field; and

error detection circuitry to determine, for each of the first-format words, whether an error exists at any of N bit positions within the N-bit payload field based at least a part on the M-bit error detection/correction field.

12. The memory control component of claim 11 wherein the N-bit payload field within each of the first-format words comprises a plurality of non-contiguous subfields.

13. The memory control component of claim 11 wherein the N-bit payload field, within each of the plurality of first-format words, comprises an (N−1)-bit data field and a one-bit encryption field, the memory control component further comprising decryption circuitry to decrypt contents of the (N−1)-bit data fields of the plurality of first-format words using an encryption value that includes bits from respective encryption fields of the plurality of first-format words.

14. The memory control component of claim 11 further comprising error correction circuitry to determine whether an error within the N-bit payload field of one of the first-format words is correctable based at least in part on the parity field within the second-format word.

15. The memory control component of claim 14 wherein the error correction circuitry to determine whether the error within the N-bit payload field of one of the first-format words is correctable comprises circuitry to determine whether the error is correctable based at least in part on (i) the parity field within the second-format word, (ii) the M-bit error detection/correction fields within the others of the first-format words and (iii) the M-bit error detection/correction field within the second-format word.

16. The memory control component of claim 11 wherein the error detection circuitry comprises logic circuitry to identify, for at least one of the first-format words for which an error is determined to exist within the N-bit payload field, a bit position within the N-bit payload field at which the bit value is in error.

17. The memory control component of claim 16 wherein the logic circuitry to identify the bit position within the N-bit payload field of the at least one of the first-format words at which the bit value is in error comprises circuitry that identifies the bit position based at least in part on the M-bit error detection/correction field within the at least one of the first-format words and the one-bit parity field within the second-format word.

18. The memory control component of claim 11 wherein the error detection circuitry comprises logic circuitry to identify, for at least two of the first-format words for which respective errors are determined to exist within respective N-bit payload fields thereof, a respective bit position within the N-bit payload field of each of the at least two of the first-format words at which the bit value is in error.

19. The memory control component of claim 11 wherein the error circuitry to determine, for each of the first-format words, whether an error exists at any of N bit positions within the N-bit payload field comprises circuitry to detect occurrence of two single-bit errors exist at respective positions of the N-bit payload field of one of the first-format words based on the M-bit error detection/correction field of the one of the first-format words and on the one-bit parity field of the second-format word.

20. A memory control component comprising:

means for reading from one or more memory components, as constituents of a data block, (i) a plurality of first-format words each consisting of M+N bits organized within an N-bit payload field and an M-bit error detection/correction field, and (ii) a second-format word consisting of M+N bits organized in an (N−1)-bit payload field, an M-bit error detection/correction field, and a one-bit parity field; and

means for determining, for each of the first-format words, whether an error exists at any of N bit positions within the N-bit payload field based at least a part on the M-bit error detection/correction field.