CN104050975A - Sectional reading-correcting-writing operation - Google Patents

Abstract

The application discloses sectional reading-mending-writing operation and discloses a system, a device and a method used for sectional reading-correcting-writing operation. According to the embodiment provided, the system is allowed to perform reading-mending-writing (RMW) operation to data stored by a main storage. The data is capable of storing large data sizes in an available buffering area. An example of the system comprises a magnetic disk provided with an overlapped track strip. In the example, each strip is divided to be a subsection containing one or more tracks. Reading-mending-writing-writing (RMWW) operation can be performed to each subsection. Further, a logic protecting strip is capable of preventing sectional rewriting of a next track in an overlapping strip. Likewise, multiple working-storage sections are achieved.

Description

Read-modification-write operation of sectional type

Background technology

The disclosure relates to a kind of data storage device, and relate more specifically to use read-revise-write the data storage memory of (RMW) operation, such as lapping magnetic recording (SMR).

Summary of the invention

In one embodiment, a kind of device can comprise data storage medium and controller, and described data storage medium comprises the track band with a plurality of data-tracks that configure in lapping mode, wherein the overlapping adjacent orbit of at least one rail portion.This controller can be configured to virtual division track band and become at least the first segmentation and the second segmentation, each fragmented packets is containing at least one track, described the first segmentation is carried out to first read-modification-write operation, and described the second segmentation is carried out to second read-modification-write operation.

In another embodiment, a kind of method can comprise: the lapping band of the track of data storage medium is divided into at least the first segmentation and the second segmentation, each fragmented packets is containing at least one track, described the first segmentation is carried out to first read-modification-write operation, and described the second segmentation is carried out to second read-modification-write operation.

In another embodiment, a kind of equipment can comprise the computer-readable data storage medium of storing instruction, when being carried out by processor, makes described processor carry out a kind of method.The method can comprise: the lapping band of the track of data storage medium is divided into at least the first segmentation and the second segmentation, each fragmented packets is containing at least one track, described the first segmentation is carried out to first read-modification-write operation, and described the second segmentation is carried out to second read-modification-write operation.

Accompanying drawing summary

Fig. 1 reads-revises-schematic diagram of the exemplary embodiment of write operation designed system for sectional type;

Fig. 2 reads-revises-schematic diagram of the exemplary embodiment of the lapping of write operation record for sectional type;

Fig. 3 reads-revises-schematic diagram of another exemplary embodiment of the lapping of write operation record for sectional type;

Fig. 4 reads-revises-schematic diagram of another exemplary embodiment of the system of the lapping of write operation record for sectional type;

Fig. 5 is that sectional type reads-revises-process flow diagram of the exemplary embodiment of the method for write operation;

Fig. 6 is that sectional type reads-revises-another process flow diagram of the exemplary embodiment of the method for write operation;

Fig. 7 reads-revises-schematic diagram of another exemplary embodiment of the lapping of write operation record for sectional type;

Fig. 8 reads-revises-schematic diagram of another exemplary embodiment of the lapping of write operation record for sectional type.

Detailed Description Of The Invention

In the following detailed description of the present embodiment, with reference to forming a part herein and accompanying drawing that wherein diagram by specific embodiment describes.But should be understood that: without departing from the spirit and scope of the present invention, can utilize and change other embodiment.

Fig. 1 describes sectional type and reads-embodiment of revise-write operation designed system, refer generally to be decided to be 100.System 100 can comprise main frame 102 and data storage device (DSD) 104.Main frame 102 also can be called as host computer system or host computer.Main frame 102 can be desk-top computer, laptop computer, server, flat computer, phone, music player, another electronic equipment or their combination in any.Equally, DSD104 can be any equipment listed above, or is used for any miscellaneous equipment of storage or retrieve data, as hard disk drive (HDD) or mixed type disk drive.Main frame 102 can or be connected by Local Area Network or wide area network (WAN) by the mode (such as wired or wireless connection) of interface with DSD104.In certain embodiments, DSD104 can be that not to be directly connected to the autonomous device of main frame 102 or main frame 102 and DSD104 can be a part for single unit.

This DSD104 can comprise one or more nonvolatile memories.In described embodiment, DSD104 can comprise in rotating magnetic disk memory 106.In other embodiments, DSD104 can comprise other storeies or type of memory, comprises volatibility and non-volatile solid state memory.For example, DSD104 can be the mixing HDD with magnetic disk memory 106 and non-volatile solid state memory 108.

In the embodiment of system 100, magnetic disk memory 106 can have one or more regions, is configured to use lapping magnetic recording (SMR) to store data on lapping data track.SMR is that register system is to improve the data recording density on disk, the adjacent data tracks thereby at least one data-track partly overlaps.Referring to Fig. 2-4, discuss SMR in detail.

SMR is the scheme of write operation of radially carrying out on disk, and wherein class of track is similar to roofing board and partly overlaps each other.With reference to Fig. 2, if supposition is being carried out and write by arrow direction indication in roofing board writing scheme, fashionable when writing in track N execution, adjacent orbit N-1 is partly overlapping.In addition, when track N+1 is carried out, write fashionablely, adjacent orbit N is partly overlapped.Than each track part is written into without any having a mind to overlapping recording method, SMR can cause increasing the recording density in the radial direction that per inch (TPI) characteristic of track is storage medium.

In addition, SMR can produce magnetic flux in a direction.Therefore, following constraint is met: after writing N track, N-1 track can not be written into.As shown in Figure 3, after writing on track N, if the reverse direction that track N-1 records direction with lapping writes, because overlapping or adjacent orbit disturb (ATI), track N can become and can not read.Therefore, the data that change is recorded to the data of track N-1 or change is recorded to track N after track record N+1 after track record N need to be different from the write-in policy of non-lapping track, and it can cover simply at any time.

Forward now Fig. 4 to, illustrate for sectional type and read-revise-another exemplary embodiment of the system of the lapping of write operation record.Rotating disc medium 402 can be divided into a plurality of regions (for example region 1, region 2, etc.), and each region can comprise a plurality of data-tracks.

Due to SMR singly write Inbound, after writing track N, write given track N-1 can rewrite all lapping tracks after track N – 1 (that is, and track N, track N+1, track N2, etc.).In order to accomplish truly this point, track collection can be grouped into " band ", and the bar end of tape is at protection track.In certain embodiments, protection track can be non-lapping track, or is not used for the lapping track of record data.When track N-1 need to be written into, by protecting the track N-1 of track to be rewritten, and do not affect the track of other band.Therefore, the track in each region can be divided into a plurality of bands.As the example embodiment of Fig. 4 is described, region 1 can comprise band band 0(B_0) to band k(B_k), and region 2 can comprise band k+1(B_k+1), band k2(B_k+2) etc.In example embodiment, each region can comprise 100 data-tracks, and 100 data-tracks can be divided into 10 bands that each contains 10 tracks.Each track can be by the data sector that Further Division becomes and a plurality of LBA (Logical Block Addressing) (LBA) is associated, and data can be stored in each LBA.

In one embodiment, the DSD of employing SMR can receive and write order, to cover the one or more LBA corresponding to the data sector of lapping track in band.The data that will write can be from medium buffer memory or other positions of main process equipment, DSD.In order carrying out, to write order, to be necessary to read a plurality of tracks of band, according to writing order, revise reading out data, and then write amended data and return to band.This can be called and reads-revise-write operation (RMW), or for SMR equipment more particularly, band reads write operation (BRO).The data of reading from track can be temporarily stored in volatile memory (such as, buffering or register) in, before writing back band, it is modified.

If write order, comprise the LBA being associated with the data sector of band 1 middle orbit, for example, data can read into volatile memory from band 1, and modify according to writing order.If DSD starts to write amended data back band 1, and the power failure that meets accident, some data may forever be lost.For example, if track N has partly been write fashionable generation power fail, because adjacent orbit disturbs, track N+1 may be not readable.Because the content of volatile memory middle orbit N+1 also may be lost, may have no idea to recover the data of track N+1.

Avoid a solution of this problem to carry out reading-revise-write (RMWW), but not RMW operation.With reference to Fig. 5, illustrate and read-revise-process flow diagram of the exemplary embodiment of write operation, and its summary is appointed as to 500.The method can relate to: 502, from being positioned at band 1 reading out data main storage device, and be stored to buffer zone.Main storage device can be lapping jukebox storage, needs the non-volatile solid state memory of RMW operation, maybe can use the storer of any other type of RMW operation.Buffer zone can be volatile solid-state (such as, DRAM) or non-volatile solid state memory (such as, flash memory).504, once data in buffer zone, it can be modified or be incorporated to new or other data.Can carry out the data manipulation that includes, but is not limited to write new data, modification available data, band cleaning or their any combination.

506, once any modification completes, amended data can be written to from buffer zone available nonvolatile storage.In an embodiment, this is the working area on disk, different nonvolatile memory (such as, flash memory), or their combination in any.Further, as described below, a plurality of working areas can be used, can be had different size, and can or can not be on identical storage medium.

508, amended data also can be written back to main storage device, to the same band that reads raw data, maybe can be written to different bands.Carrying out RMWW operation has increased one deck redundancy, to help prevent loss of data.For example, if there is power-off during writing working area, raw data remains safe in band 1.If there is power-off during writing band 1, amended data should be safe in working area.

When data are transferred to working area from primary storage medium, this RMWW method depends on buffer stores data.It is large not that problem may appear at buffer zone, cannot hold in band all data that shift.These problems can occur, because the change of any data slice can read or write whole band on band.For example, during 2 kbytes of data sheet on changing 100 megabyte bands, whole 100 megabyte may need to be read into or to write from buffer zone.Stripe size can be reduced or buffer size increases to avoid buffer pool size problem.

The solution of buffer size problem can be band to be divided into the segmentation of reading of selected track, for carrying out RMWW operation.Each segmentation can comprise a plurality of tracks, rail portion, and it can be that size is fixing or dynamic, and with ascending order LBA, sequentially specifies in whole band.When needs executing data business, only segmentation (rather than whole band) can be modified, thereby saves time and reduce buffer size.

The shortcoming that band is divided into segmentation is: in band, there is no physical protection track.Do not protect track, when revising track N-1, the destruction of track N can occur.In order to overcome this problem, virtual protection track can replace physical protection track in the end of segmentation, allows control and dirigibility when writing multilayer lapping data.Virtual protection track can be at least one track in width and can in the user-accessible district of band, move.Virtual protection track can float, and when segmentation makes progress, it can move, and further, can in band, use a plurality of virtual protection tracks.

When the data in segmentation (segmentation 1) (comprise N track and start at track A) read buffer zone from primary memory, read-revise-segmentation method of write operation can work.In various embodiments, track A can be the track that in band, the first physics is positioned at, or can be another track in band, because can need this part of the strap end of rewriting can carry out RMW operation.For example, if all merging are in last 3 tracks of band, only have these three magnetic tracks to be rewritten, because write these tracks that do not disturb previous generation (, if overlapping direction occurs, wherein last Orbital Overlap penultimate track).Yet if the data in the first track of band will be changed, all tracks of this band are in requisition for being rewritten in RMW operating period.

At selected data, after buffer zone, can operate Update Table by executing data, such as merging any new data and available data.Once data manipulation completes, amended data (comprising that track A is to N) can be written to the first working area (spA).Then, revise track A to N-1(now in spA) can be written to each band of primary memory.Owing to writing track N-1, can destroy track N, track N can consider the virtual protection track in primary storage, because the data of track N also also can currently be stored in spA in buffer zone.In some cases; when carrying out read operation during this process; this reads and can point to each working area, because such as the data when on virtual protection track, on track N, in primary storage medium can be damaged time, the data in working area be revised or unique effective copy recently.

In addition, the second segmentation can be selected for the track of the next generation that is not yet modified to current RMWW operation, in given example, can comprise track N+1 to M.The data of the second segmentation can read buffer zone and revise from primary memory, and this buffer zone can comprise track N.The second segmentation can be written to spB, and can produce the spB that comprises track N to M, because track N also can be used for buffer zone.After the second segmentation is stored in spB, the track N to M-1 of modification can be written to primary memory.Now, the track A to M-1 on primary memory can comprise valid data, and M can be virtual protection track.This process can be repeatedly, until RMWW operation has all been carried out in all segmentations in band, thereby provide the complete RMWW operation of whole band, and when the data capacity of band is greater than when carrying out associated buffer big or small of RMWW operation, this is useful.

With reference to Fig. 6, illustrate and read-revise-process flow diagram of the exemplary embodiment of the segmentation method of write operation, and its summary is appointed as to 600.In certain embodiments, the method 600 can allow system to carry out RMW operation to the data of storing in primary memory, and these data have larger size of data with comparing of can storing in associated buffer.With reference to the magnetic track with the band of lapping track, in the example of this system shown in above accompanying drawing.The method 600 each band of permission are divided into and comprise one or more tracks; wherein each band can comprise the more data of storing than in the cushion space that is associated, and the method 600 each segmentation is carried out read-revise-write (RMWW) to operate and implement virtual protection band.

In certain embodiments, band can be split into a plurality of virtual segmentation, for whole band being carried out to RMWW operation.In method 600, in 602, the first segmentations, can be read and can comprise preliminary orbit, track X is to track Y, and it can comprise N track.At track X to Y(, comprise X and Y) in the data volume of distributing should not surpass the size of the buffer zone being associated.In addition, at track X to Y(, comprise X and Y) should be over the size in the first working area (spA).604, can be to the data executing data operation in buffer zone, such as using any new data receiving from main frame to revise the data of selected track.

606, can carry out inspection, to determine track whether be physical protection track, this, in this case, can not need virtual protection track if also can be considered to the last track in band.In some examples as herein described; the last track of band can be assumed that physical protection track; but we can define the last track of band is non-protection track; do not comprise that physical protection track is for a part for definition band; in this case, can make and realize identical object to providing the subtle change of example herein.

Continue, 608, when Y is not physical protection track, the data in track X to Y can be stored in the spA of buffer zone.610, when Y is physical protection track, the data of track X to Y-1 can store spA into.This is because the data (if any) on physical protection track are readable or unavailable, so physical protection track can not comprise the memory location of mapping address.In the future of this process iteration, selected data can be stored in spB.

612, once selected data has been stored in spA or spB, track X to Y-1 can be stored in main storage device.In lapping storage system, write Y-1 and can destroy the data in track Y, thereby cause track destroy, thereby when the data trnascription of track Y is remained on to spA or spB(or needs, in both) virtual protection track segmentation band can be created.Before track Y-1 data being write in respective strap, fault writes the data of magnetic track Y to spA or spB(or other positions) can cause the loss of data on track Y.

Once current segmentation has been written to primary data store equipment, the method 600 can determine whether to exist a plurality of segmentations of the selected band also not operating through segmentation RMWW.If all segmentations corresponding to selected section are written into, and do not have more segmentations of selected section to write, that is, RMWW operation is carried out in the applicable segmentation of whole band or band, and 616, this process can finish.When needing another segmentation of track, 618, with Zx, increase progressively X, and increase progressively Y with Zy.In an embodiment, Zx can equal Y-X+1, and can be more than or equal to Zx with Zy.When Zx equals Y-X+1, the new starting point track of next segmentation should be a track higher than virtual protection track, previously the data of virtual protection track Y-1 can be in buffer zone or can before store spA or spB into.

For example, the segmentation that contains 6 tracks can start at track 100 and finish, at track 105, at track 105, to have virtual protection track.This can cause Zx=6, thus the first track of mobile the first segmentation, and track X, from track 100 to track 106.When Zy=Zx, the last track in segmentation (track Y) six tracks that have also been shifted, move to track 111 from track 105.Set Zx=Zy desirable, because fragment size remains unchanged in each iteration of this process.But in certain embodiments, during some or all iteration of process, it is favourable making segmentation increase track size, in this case, be necessary to arrange Zy to value be greater than or less than Zx.For example, the segmentation of considering wherein to comprise six tracks can start at track 100 and end at the situation of track 105, increases progressively as follows: X=X+Zx and Y=Y+Zy, wherein, Zx=Y-X+1, Zy=Y-X+2.In this escalation regimens, the first track in segmentation can move to track 106 from track 100, and the last track in segmentation can move to track 112 by track 105, and the fragmented packets of generation is containing 7 tracks, rather than 6.Therefore, Zx and Zy can select to change according to system the value of each iteration, and it can be corresponding to the efficiency maximizing in buffer zone or working area.

After X and Y are incremented respectively Zx and Zy, can determine that track Y is virtual protection track, in this case, can not need physical protection track.This process can be repeated, until arrival physical protection track or system have completed RMWW operation for all selection bands.Therefore, working area spA and spB can replace in whole process, no matter have how many segmentations.Yet, do not exist working area number to be restricted to 2 rule, if needed, can use more working area.

Fig. 7 reads-revises-exemplary embodiment of the segmentation method of write operation.Example process starts from track t_begin and finishes at t_last.In this example, t_last is last track or the physical protection track of selected band.When band operates through RMW, 702, data can read from track t_begin to t_N the buffer zone of executing data operation.Also 702, corresponding to the Update Table of track t_begin to t_N, can be stored in the first working area (spA).706, this can guarantee when track t_N-1 is rewritten, even if it damages on primary memory, the data trnascription of track N is stored in working area; The data of track N are considered to virtual protection track now.Then, 708, track t_N+1 to T_R can read in buffer zone from primary storage, revises, and stores SPB into.Then, data can be written to primary memory, and this can cause the track t_begin to T_R-1 that stores in primary data store medium as a part for current RMW operation, thereby track T_R becomes virtual protection track.

The method can expand to hold the band of different sizes at an easy rate.In the present embodiment, this process proceeds to the end of band.714, the data of track t_last – x+1 to t_last-1 are read into buffer zone and revise.Because t_last-x also should be in buffer zone, track t_last – x to t_last-1 can be stored in working area spLAST.But working area spLAST spA, spB or diverse working area.Track t_last-x to t_last-1 can be written to the respective carter t_last-x to t_last-1 on primary data store medium.When process arrives the last segmentation of band, may not need logical track protection, because physical protection track may exist.In provided example, last track t_last can be physical protection band, and does not need to be stored in any working area.

With reference to figure 8, illustrate and read-revise-diagrammatic representation of the dividing method of write operation, and refer generally to be decided to be 800.Shown in particular example in, system can write two data sectionals, each segmentation comprises nearly six tracks, still, can write as required more segmentations or different fragment size.In the example shown, the track in the first segmentation can comprise track 100 to 105, and track in the second segmentation can comprise track 106 to 111.

Method 800 can comprise two iteration, or the stage, and each stage shows the processing of carrying out in can read in sectional type-modification-write operation.In the first stage (pass A) 802, track 100 to 105 can read in buffer zone from main data storage 806.Method 800 can be carried out modification, merge or other data manipulation, and data are in buffer zone.Next, 808, corresponding to the data of track 100 to 105, can store spA into.812, then track 100 to 104 can be written to primary data store equipment.Then track 105 virtual protection track, because corresponding data trnascription can be in buffer zone with it, and copy can be stored in spA, and respective rail 105 on primary data store equipment can be damaged due to writing of track 104.

Move to subordinate phase (pass B) 804, in 814, the second segmentations, can comprise that track 111 is to track 106, and can read from main data storage.818, corresponding to track 106(or 105, because it also should be still in buffer zone) data to 111 can store spB into.820, track 105 to 110 can be written to the respective rail of primary data store equipment subsequently.

The valid data of recalling the track 105 of primary data store equipment still can be in buffer zone, also still can be in spA.Therefore,, when data-track 105 to 110 is in 820 respective rail that are written on primary data store equipment, track 105 will be used valid data to rewrite, and be no longer virtual protection track, make track 111 be called now virtual protection track.

In certain embodiments, can there be a plurality of stages, such as phase III (pass C) (not shown), fourth stage (pass D) (not shown), etc.Can repeatedly use working area during the course.For example, spA can be for process A, process C, etc.But the size of spA needn't be identical for each stage, it can be dependent on fragment size.If the fragmented packets in stage C is containing eight tracks, in stage A, the large I of spA is different from stage C.

According to various embodiment, method described herein may be implemented as the one or more software programs that move on computer processor or controller.According to another embodiment, method described herein may be implemented as the one or more software programs in the upper operation of computing equipment (personal computer driving such as, use dish).Specialized hardware implementation includes, but is not limited to special IC, programmable logic array and can be configured to equally realize other hardware device of method described herein.Further, method described herein may be implemented as the computer-readable medium that comprises instruction, makes processor carry out the method when carrying out instruction.

The explanation of embodiment described herein aims to provide the general understanding to the structure of various embodiment.This explanation is not intended to as the complete description that utilizes the device of structure described herein or method and all elements of system and feature.Those skilled in the art are after browsing the disclosure, and many other embodiment can be apparent.For example, the segmentation of the RMW operation of discussing herein can be used for using RMW operation or data storage device or storage medium that can compare operation, such as the programming-erase cycle of solid-state memory.Other embodiment can be utilized and be derived from present disclosure, makes can carry out without departing from the spirit and scope of the invention structure and logic and replaces and change.And, although specific embodiment has been illustrated and has described, should be appreciated that design anyly configures the specific embodiment shown in alternative subsequently with what realize identical or similar object.

Present disclosure is intended to contain any and all follow-up modifications or the change of various embodiment.After those skilled in the art browse this description, the combination of above-described embodiment and herein not specifically described other embodiment will be apparent.In addition, this diagram is only representational, and can not to scale (NTS) draw.Some ratio in figure can be exaggerated, and other ratio can reduce.Therefore, the disclosure and accompanying drawing should be regarded as illustrative and not restrictive.

Claims (20)

1. a device, comprising:

The data storage medium that comprises the track band with a plurality of data-tracks that configure in lapping mode, the wherein overlapping adjacent orbit of at least one rail portion;

Controller, is configured to:

Virtual division track band becomes at least the first segmentation and the second segmentation, and each fragmented packets is containing at least one track;

Described the first segmentation is carried out to first read-modification-write operation; And

Described the second segmentation is carried out to second read-modification-write operation.

2. device as claimed in claim 1, comprising:

Described controller is further configured to:

Read first and be fragmented into buffer zone;

Be modified in the data of the first segmentation in described buffer zone;

Write first and be fragmented into the first working area; And

Write and be less than the part of all the first segmentations to the physics adjacent orbit of the first segmentation corresponding on data storage medium.

3. device as claimed in claim 2, comprising:

Described controller is further configured to:

Read second and be fragmented into buffer zone;

Be modified in the data of the second segmentation in buffer zone;

Write comprise the first segmentation at least last track second be fragmented into the second working area, and

Write and be less than the part of all the second segmentations to the physics adjacent orbit of the second segmentation corresponding on data storage medium.

4. device according to claim 3, wherein, the described part that is less than all described the first segmentations is the last adjacent orbit that the first segmentation deducts the first powder fist.

5. device as claimed in claim 4, wherein, the described part that is less than all described the second segmentations is the last adjacent orbit that the second segmentation deducts the second segmentation.

6. device as claimed in claim 1, wherein:

Described controller is further configured to:

Repeat reading-revise-write-write operation of any extra segment, until all tracks in band are by reading accordingly-revise-write-write operation; And

When the last track of segmentation is while being separated band with the physical protection band of the track of other in band not, stop band carrying out and read-revise-write-write operation.

7. device as claimed in claim 1, further comprises:

The first working area;

The second working area; With

Described controller is further configured to: first, read-revise-write first during write-write operation to be fragmented into the first working area, and read-revise second-write second during write-write operation to be fragmented into the second working area.

8. device as claimed in claim 7, comprising:

Described controller is further configured to: to further the reading-revise of other segmentations of band-write-write operation operation, alternately write the first working area and the second working area.

9. device as claimed in claim 1, further comprises:

The first segmentation comprises the first scope of the physics adjacent orbit of band;

The second segmentation comprises the second scope of the physics adjacent orbit of band, and

The last track of contiguous described first segmentation of the first track of described the second segmentation.

10. a method, comprising:

The lapping band of the track of data storage medium is divided into at least the first segmentation and the second segmentation, and each fragmented packets is containing at least one track;

Carry out first of the first segmentation and read-revise-write-write operation process, and

Carry out second of the second segmentation and read-revise-write-write operation process.

11. methods as claimed in claim 10, further comprise: described data storage medium comprises a plurality of lapping bands of the track with a plurality of overlapped data tracks that configure with following form: the overlapping adjacent orbit of at least one rail portion.

12. methods as claimed in claim 10, further comprise: described first, read-revise-write-write operation and second reads-revise-write-write operation between actuating logic protection band.

13. methods as claimed in claim 10, further comprise:

Any additional segments is carried out and read-revise-write-write (RMWW) operation, until all tracks in lapping band are through reading accordingly-revise-write-write operation; With

When the last track of segmentation is while being separated band with the physical protection band of the track of other in band not, stop band carrying out and read-revise-write-write operation.

14. methods as claimed in claim 13, further comprise: between corresponding to two continuous reading-revise-write-write operation section, implement the virtual protection band that floats.

15. methods as claimed in claim 13, further comprise: implement to float virtual protection band and the mobile virtual protection bar that floats take last processing to through the segmentation of RMWW operation and process between the next segmentation through RMWW operation.

16. methods as claimed in claim 10, further comprise:

First read-revise-write described first in write-write operation process to be fragmented into the first working area, and

Second, read-revise-write described second in write-write operation process to be fragmented into the second working area.

17. methods as claimed in claim 16, further comprise: for other of other segmentations corresponding to lapping band, read-revise-the alternately iteration of write-write operation, be alternately written to the first working area and the second working area.

18. 1 kinds of devices, comprising:

The computer-readable data storage medium of storage instruction, when being carried out by processor, makes described processor manner of execution, comprising:

The lapping band of the track of dividing data storage medium becomes at least the first segmentation and the second segmentation, and each fragmented packets is containing at least one track;

Described the first segmentation is carried out to first read-modification-write operation; And

Described the second segmentation is carried out to second read-modification-write operation.

19. devices as claimed in claim 18, wherein:

Described computer-readable data storage medium storage instruction, when being carried out by processor, makes described processor carry out described method and further comprises:

First, read-revise-write-write operation and second reads-revises-implement virtual protection band between write-write operation, wherein, second, read-revise-after write-write operation completes, described virtual protection band is not physical protection band.

20. devices as claimed in claim 18, is characterized in that:

Described computer-readable data storage medium storage instruction, when being carried out by processor, makes described processor carry out described method and further comprises:

First, read-revise-write described first in write-write operation process to be fragmented into the first working area, and

Second, read-revise-write described second segment to the second working area in write-write operation process.