System and method for optimizing sequential XOR command performance

• US 7,334,081 B2
• Filed: 04/29/2005
• Issued: 02/19/2008
• Est. Priority Date: 04/29/2005
• Status: Expired due to Fees

• Current Assignee: Hitachi Global Storage Technologies Netherlands BV (Western Digital Corporation)
• Original Assignee: Hitachi Global Storage Technologies Netherlands BV (Western Digital Corporation)
• Inventors: Espeseth, Adam Michael, Younk, Edward Henry
• Primary Examiner(s): SONG, HUA JASMINE
• Application Number:
  US11/117,972
• Publication Number:
  US 20060245106A1
• Time in Patent Office: 1,026 Days
• Field of Search: None
• US Class Current: 711/112
• CPC Class Codes: G11B 20/10 Digital recording or reproducing
• CPC Class Codes: G11B 2020/10916 Seeking data on the record carrier for preparing an access to a specific address
• CPC Class Codes: G11B 5/59627 Aligning for runout, eccentricity or offset compensation G11B5/5534, G11B5/59677, G11B5/59688 take precedence

Abstract

An RPO algorithm in a HDD coalesces LBA-sequential XOR commands in pipes, and passes the pipes to a lower level execution engine. The execution engine executes XOR reads and write separately to optimize performance using head and/or cylinder skew information to approach the nominal disk data rate.

Claims

1. A hard disk drive (HDD) comprising:

at least one rotatable disk;

at least one data transfer element; and

at least one controller controlling the data transfer element to execute XOR commands by coalescing LBA-sequential XOR commands into pipes, and separately executing read commands and write commands in the pipes using skew information, each XOR command including at least one read followed by an XOR operation followed by at least one write of the result of the XOR operation, wherein optimal disk access lengths and optimal read/write switching points are determined based on skew information to minimize dead latency, and further wherein a read portion of an XOR command is terminated on a last skew boundary that allows just enough time to seek back to a starting location of the read portion to begin an XOR write operation without incurring an additional disk revolution.

2. The HDD of claim 1, wherein the skew information is cylinder skew information.

3. The HDD of claim 1, wherein the XOR commands are coalesced by a Rotational Position Optimization (RPO) algorithm.

4. The HDD of claim 1, wherein an XOR disk operation is terminated on a head or cylinder skew boundary.

5. A controller executing logic comprising:

grouping multiple disk access commands in pipes, each command requiring at least one read along with at least one associated write;

passing the pipes to an execution engine of a disk drive;

separating multiple disk access command reads from corresponding writes and then executing the reads and writes based on skew, wherein optimal disk access lengths and optimal read/write switching points are determined based on skew to minimize dead latency, and further wherein a read portion of a command is terminated on a last skew boundary that allows just enough time to seek back to a starting location of the read portion to begin a related write operation.

6. The controller of claim 5, wherein the skew is cylinder skew.

7. The controller of claim 5, wherein the commands are coalesced by a Rotational Position Optimization (RPO)) algorithm.

8. The controller of claim 5, wherein a disk operation is terminated on a head or cylinder skew boundary.

9. A HDD comprising:

at least one disk;

a controller writing data to and reading data from the disk in response to commands, each command requiring at least one read along with at least one associated write, the controller invoking;

means for grouping LBA-sequential multiple disk access commands;

means for separating reads from writes in the commands; and

means for executing the reads and writes based on skew information, wherein optimal disk access lenghts and optimal read/write switching points are determined by the means for executing based on skew information to minimize dead latency, and wherein the means for executing terminates a read portion of an XOR command on a last skew boundary that allows just enough time to seek back to a starting location of the read portion to begin an XOR write operation without incurring an additional disk revolution.

10. The HDD of claim 9, wherein the means for grouping is embodied in an RPO algorithm.

11. The HDD of claim 10, wherein the means for separating and means for executing are embodied in an execution engine that is at a lower level in a software architecture of the HDD than the RPO algorithm.

12. The HDD of claim 9, wherein the skew information is cylinder skew information.

13. The HDD of claim 9, wherein the means for executing terminates an XOR disk operation on a head or cylinder skew boundary.