Read channel buffer management for higher throughput decoding

1. A method comprising:

identifying a set of failed data blocks read from a storage medium during execution of a read command;

populating sample buffers in a read channel with data of a first subset of the set of failed data blocks;

initiating an error recovery process on the data in the sample buffers, the error recovery process being a read sample combining process that entails multiple rounds of read averaging to update previously-stored samples based on newly-sampled read data;

responsive to successful recovery of one or more data blocks in the first subset, releasing the recovered data from the sample buffers and repopulating sample buffer locations previously-storing the recovered data with data of a second subset of the set of failed data blocks prior to completion of the multiple rounds of read averaging on the first subset of the failed data blocks; and

initiating the error recovery process on the data of the second subset of the failed data blocks while the error recovery process is ongoing with respect to data of the first subset of failed data blocks remaining in the sample buffers.

2. The method of claim 1 , wherein the multiple rounds of read averaging are performed without firmware intervention.

3. The method of claim 2 , further comprising:

during each of the multiple rounds of the read sample combining recovery process, allocating different numbers of decoding iterations to different data blocks in the set of failed data blocks.

4. The method of claim 1 , wherein releasing the recovered data from the sample buffers and repopulating the sample buffers with the data of the second subset of failed data blocks further comprises:

dynamically freeing up one or more of the sample buffers prior to a last round of the multiple rounds of the read sample combining recovery process; and

repopulating the freed-up sample buffers with the data of the second subset of the set of failed data blocks without freeing up another subset of the sample buffers storing data of the first subset of the failed data blocks not yet successfully recovered.

5. The method of claim 1 , wherein the multiple rounds of the read sample combining process are effective to perform decoding operations on a greater number of data blocks than there exist sample buffers in the read channel.

6. The method of claim 1 , further comprising:

dynamically tracking, in hardware, a subset of the failed data blocks that remain to be recovered as the subset is reduced in size with each subsequent round of an ongoing read sample combining recovery process.

7. The method of claim 1 , further comprising:

dynamically tracking, in hardware, a count value for data currently held in each one of a subset of the sample buffers allocated to a read sample combining process, the count value indicating a number of update rounds that a data sample stored within the buffer has been subjected to.

8. The method of claim 1 , further comprising:

decoding the set of failed data blocks according to an order that is different than the order for which failed data blocks are arranged on the storage medium.

9. A system comprising:

read channel hardware adapted to:

receive a set of failed data blocks read from a storage medium during execution of a read command;

populate sample buffers in a read channel with data of a first subset of the set of failed data blocks;

initiate an error recovery process on the data in the sample buffers, the error recovery process being a read sample combining process that entails multiple rounds of read averaging to update previously-stored samples based on newly-sampled read data;

responsive to successful recovery of one or more data blocks in the first subset, release the recovered data from the sample buffers and repopulate sample buffer locations previously-storing the recovered data with data of a second subset of the set of failed data blocks prior to completion of the multiple rounds of read averaging on the first subset of the failed data blocks; and

initiate the error recovery process on the data of the second subset of the failed data blocks while the error recovery process is ongoing with respect to data of the first subset of failed data blocks remaining in the sample buffers.

10. The system of claim 9 , wherein the multiple rounds of read averaging are performed without firmware intervention.

11. The system of claim 9 , wherein the read channel hardware is further adapted to:

release the recovered data from the sample buffers without releasing other data of the first subset of failed data blocks that has not yet been recovered.

12. The system of claim 9 , wherein the multiple rounds of the read sample combining process are effective to perform decoding operations on a greater number of data blocks than there exist sample buffers in the read channel.

13. The system of claim 9 , wherein the read channel hardware is further adapted to:

during each of the multiple rounds of the read sample combining process, allocate different numbers of decoding iterations to different data blocks in the set of failed data blocks.

14. The system of claim 9 , wherein the read channel hardware is further adapted to:

dynamically track a subset of the failed data blocks that remain to be recovered as the subset is reduced in size with each subsequent round of an ongoing read sample combining recovery process.

15. The system of claim 9 , wherein the read channel hardware is further adapted to:

dynamically track a count value for data currently held in each one of a subset of the sample buffers allocated to a read sample combining process, the count value indicating a number of update rounds that data stored within the buffer has been subjected to.

16. The system of claim 9 , wherein the read channel hardware is further adapted to:

manage a block-to-buffer map to dynamically track, in hardware, a buffer location for data stored in association with each one of the failed data blocks.

17. One or more tangible computer-readable storage media encoding instructions for executing a computer process, the computer process comprising:

identifying a set of failed data blocks read from a storage medium during execution of a read command;

populating sample buffers in a read channel with data of a first subset of the set of failed data blocks;

initiating an error recovery process on the data in the sample buffers, the error recovery process being a read sample combining process that entails multiple rounds of read averaging to update previously-stored samples based on newly-sampled read data;

responsive to successful recovery of one or more data blocks in the first subset, releasing the recovered data from the sample buffers and repopulating sample buffers locations previously-storing the recovered data with data of a second subset of the set of failed data blocks prior to completion of the multiple rounds of read averaging on the first subset of the failed data blocks; and

initiating the error recovery process on the data of the second subset of the failed data blocks while the error recovery process is ongoing with respect to data of the first subset of failed data blocks remaining in the sample buffers.