CN101493794B - Flash memory data processing method and device - Google Patents

Abstract

The embodiment of the invention discloses a flash memory data processing method and device for improving random write performance. The method in the embodiment of the present invention includes: querying the source data to be updated according to the obtained target data; writing the target data into the first data page of the preset blank data block, and marking the data page where the source data is located as An invalid page: replacing the entry of the invalid page in the preset page-level logical mapping table with the entry of the first data page of the preset blank data block. The embodiment of the present invention also provides a flash memory data processing device. The embodiments of the present invention can effectively improve random write performance.

Description

A flash memory data processing method and device

technical field

The invention relates to the field of data processing, in particular to a flash memory data processing method and device. the

Background technique

Solid State Disk (SSD, Solid State Disk) is composed of a control unit and a storage unit. Among them, the storage unit is mostly composed of a non-volatile flash memory (Flash) chip. Since it does not have a mechanical rotation device, the SSD has strong shock resistance. It has the advantages of small overhead and is widely used. the

Flash memory is physically composed of several data blocks, and each data block is composed of several pages. In the prior art, the management of flash memory is based on the management of data blocks, that is, the smallest management unit of flash memory is data block, for example , when a page of data in a certain data block needs to be changed, the data block will be completely erased, and then all rewritten. the

It can be seen from the above process that in the scheme of the prior art, when writing a certain page of data, the data block where the page is located will be completely erased, and then the data block is rewritten, so that when updating data A large number of data erasing and writing operations are required, thereby reducing the performance of random data writing. the

Contents of the invention

The embodiment of the present invention provides a flash memory data processing method and device, which can improve the performance of random data writing. the

The flash memory data processing method provided by the embodiment of the present invention includes: querying the source data to be updated according to the obtained target data; writing the target data into the first data page of the preset blank data block, and writing the source data to The data page is marked as an invalid page; the entry of the invalid page in the preset page-level logical mapping table is replaced with the entry of the first data page of the preset blank data block. the

The flash memory data processing method provided by the embodiment of the present invention includes: obtaining the length information of the data to be written; splitting the data to be written into block levels according to the length information and the number of sectors contained in a data block data and page-level data; write the block-level data in a block-writing manner, write the page-level data in a page-writing manner, and update the preset page-level logical mapping table. the

The flash memory data processing device provided by the embodiment of the present invention includes: a source data query unit, configured to query source data to be updated according to the acquired target data; a writing unit, configured to write the target data into preset blank data In the first data page of the block; the marking unit is used to mark the data page where the source data is located as an invalid page; the replacement unit is used to replace the entry of the invalid page in the preset page-level logical mapping table It is an entry of the first data page of the pre-set blank data block. the

The flash memory data processing device provided by the embodiment of the present invention includes: a length information acquisition unit, configured to acquire the length information of data to be written; Split the data to be written into block-level data and page-level data; the data writing unit is used to write the block-level data according to the block writing method, and write the page according to the page writing method Level data and update the preset page-level logical mapping table. the

As can be seen from the above technical solutions, the embodiments of the present invention have the following advantages:

In the embodiment of the present invention, after the target data is acquired, the target data is written into the first data page of the preset blank data block, and the data page where the source data is located is marked as an invalid page, and the preset page level The entry of the invalid page in the logical mapping table is replaced by the entry of the first data page. Therefore, in the process of data updating, there is no need to erase the data block where the invalid page is located, and directly use the first data block in the blank data block. A data page replaces the invalid page for data storage, thus saving the operation of data updating, thereby improving the performance of random writing of data. the

Description of drawings

Fig. 1 is the flowchart of the first embodiment of the flash memory data processing method in the embodiment of the present invention;

Fig. 2 (a) is the flow chart of the second embodiment of flash memory data processing method in the embodiment of the present invention;

Figure 2 (b) is a schematic diagram of a page-level logical mapping table in an embodiment of the present invention;

Fig. 3 is the flowchart of the third embodiment of the flash memory data processing method in the embodiment of the present invention;

Fig. 4 is a schematic diagram of the page writing method in the embodiment of the present invention;

Fig. 5 is a schematic diagram of block writing mode in the embodiment of the present invention;

Fig. 6 is a schematic diagram of the first embodiment of the flash memory data processing device in the embodiment of the present invention;

FIG. 7 is a schematic diagram of a second embodiment of a flash memory data processing device in an embodiment of the present invention. the

Detailed ways

Embodiments of the present invention provide a flash memory data processing method and device, which are used to improve the performance of random writing of data. the

Please refer to Fig. 1, the first embodiment of flash memory data processing method in the embodiment of the present invention comprises:

101. Query the source data to be updated according to the acquired target data;

In this embodiment, when the data of a certain page in the flash memory data block needs to be updated, new data, that is, target data, is first obtained, and then the source data to be updated is queried according to the storage location of the target data. the

For example, when the user needs to replace a file in the operating system (specifically, the user can paste a file with the same file name and select "overwrite"), the operating system can learn from the pasted position of the user that the file is in the flash memory The corresponding location, and read the data originally stored in this location in the flash memory, the source data to be updated can be obtained. the

102. Write the target data into the first data page of the preset blank data block, and mark the data page where the source data is located as an invalid page;

After the source data to be updated is determined, a blank data block is obtained from the preset blank data block group, and the target data is written into a certain data page of the blank data block, which is recorded as the first data page. the

And mark the data page where the source data is located as an invalid page. The specific marking method can be: insert an invalid mark in the label or attribute field of the data page, and the invalid mark is used to indicate that the data page is invalid, that is, it will not be The data page performs read and write operations. It should be noted that in practical applications, there may be many ways to specifically mark invalid pages, which are not limited here. the

103. Replace the entry of the invalid page in the preset page-level logical mapping table with the entry of the first data page. the

Since the data page where the source data is located is marked as an invalid page in step 102, in order to ensure that the data in the data page associated with the data page remains intact, the invalid page can be stored in the preset page-level logical mapping table The table entries of the first data page are replaced with the table entries of the first data page. the

The page-level logical mapping table in this embodiment records the logical address and physical address of each data page storing data, and the association relationship between each data page. the

In this embodiment, the entry of the invalid page in the page-level logical mapping table is replaced with the entry of the first data page, so that the data in the first data page can replace the data in the invalid page, and each data page The association relationship between remains unchanged, that is, the data update is realized. the

In this embodiment, after the target data is acquired, the target data is written into the first data page of the preset blank data block, and the data page where the source data is located is marked as an invalid page, and the preset page-level logic The entry of the invalid page in the mapping table is replaced by the entry of the first data page. Therefore, in the process of data updating, there is no need to erase the data block where the invalid page is located, and directly use the first data block in the blank data block. The data page replaces the invalid page for data storage, thus saving the operation of updating data, thereby improving the performance of random writing of data.

Please refer to Fig. 2 (a), the second embodiment of the flash memory data processing method in the embodiment of the present invention includes:

201. Search for blank data blocks, and establish a page-level logical mapping table;

In this embodiment, when the system is powered on, each physical data block is traversed, the physical data block with unrecorded data and the physical data block that has been erased are queried, and the physical data block with unrecorded data and the physical data block that has been erased The erased physical data block is used as a preset blank data block. the

Assuming that there are 3 data blocks in total, namely X, Y and Z, where data has been recorded in X and has not been erased, and no data has been recorded in Y and Z, then Y and Z can be found to be blank data blocks by traversing . the

It should be noted that in this embodiment, a page-level logical mapping table can also be established in advance, and the logical address and physical address of each data page storing data are recorded in the page-level logical mapping table, as well as the relationship between each data page. As for the association relationship, the page-level logical mapping table in this embodiment may use a data storage method such as a binary tree or a B+ tree to record data. the

In this embodiment, a binary tree is taken as an example to illustrate the page-level logical mapping table. It can be understood that other types of data storage methods can also be used:

Please refer to Figure 2(b). There are 6 nodes in the figure, namely A, B, C, D, E, and F. Each node represents a data page, and each node is composed of four parts. From From left to right are "left node identifier", "logical address", "physical address" and "right node identifier". Each node in this embodiment can be composed of 16 bytes, and each of the four parts occupies 4 bytes. the

In Fig. 2(b), if the "left node identifier" or "right node identifier" of a node is "NULL", it means that the node has no left node or no right node. the

202. Query the source data to be updated according to the acquired target data;

In this embodiment, when the data of a certain page in the flash memory data block needs to be updated, new data, ie target data, is obtained first, and then the source data to be updated is queried according to the storage location of the target data. the

Assuming that the user wishes to replace the source data "300" with the target data "500", it can be determined that the target data is "500" and the source data is "300". the

203. Write the target data into the first data page of the preset blank data block, and mark the data page where the source data is located as an invalid page;

After the source data to be updated is determined, a blank data block is obtained from the preset blank data block group, and the target data is written into a certain data page of the blank data block, which is recorded as the first data page. the

And mark the data page where the source data is located as an invalid page. The specific marking method can be: insert an invalid mark in the label or attribute field of the data page, and the invalid mark is used to indicate that the data page is invalid, that is, it will not be The data page performs read and write operations. It should be noted that in practical applications, there may be many ways to specifically mark invalid pages, which are not limited here. the

Assuming that the source data "300" is stored in the data page represented by node E (as shown in Figure 2(b)), it can be known that the logical address of the data page is 70, the physical address is 8, the left node is D, and the right node is F , because the source data "300" in the data page is invalid, mark the data page as an invalid page. the

The current blank data blocks are Y and Z, and one can be selected arbitrarily. For example, if Y is selected as a blank data block to write data, then the target data "500" can be written into a certain data page in the blank data block Y (that is, the first A data page), assuming that the physical address of the first data page is 9. the

204. Replace the entry of the invalid page in the preset page-level logical mapping table with the entry of the first data page;

Since the data page where the source data is located is marked as an invalid page in step 203, in order to ensure that the data in the data page associated with the data page remains intact, the invalid page can be stored in the preset page-level logical mapping table The table entries of the first data page are replaced with the table entries of the first data page. the

A specific replacement method may be: query the entry where the invalid page is located, modify the physical address in the entry to the physical address of the first data page, and retain the original associated information of the entry, which is used to represent The association relationship between this entry and other entries. the

That is, the entry where the invalid page is located is node E, obtain the physical address 9 of the first data page, modify the physical address in node E from 8 to 9, and keep the original "left node ID" and "right node ID" of the node E. Node ID", then the replacement of the invalid page by the first data page is realized. the

205. Judging whether the preset update condition is satisfied, if yes, execute step 206, if not, continue to judge;

In this embodiment, because some data pages in the data block will be marked as invalid pages in step 203, then these invalid pages cannot be read and written temporarily. In order to improve the utilization rate of storage resources, in some cases it is necessary to Update the data in these data blocks to eliminate invalid pages. The specific situations can include:

(1), the number of invalid pages in a data block reaches the preset value:

When the number of invalid pages in a data block reaches a certain threshold (the threshold can be set according to user needs, generally set to a number greater than or equal to 2), the storage space may be reduced, In this case, the data block where the invalid page is located can be updated according to the target data in the first data page. the

(2), the system is idle:

When the system is not performing read or write operations, that is, it is in an idle state, the idle time can also be used to update the data block where the invalid page is located according to the target data in the first data page. the

It should be noted that the preset update conditions are described above with only two examples, and it can be understood that in practical applications, other preset update conditions may also be used, which are not limited here. the

206. Update the data block where the invalid page is located according to the target data in the first data page. the

In this embodiment, the specific update process can be:

Save the data in other data pages in the data block except invalid pages to the cache;

Erase the data block;

Obtain the target data in the first data page;

Write the target data into the original invalid page to obtain the modified page;

Write the modified page containing the target data and the data content stored in the cache into the data block after erasing;

The entry of the first data page in the preset page-level logical mapping table is replaced with the entry of the modified page. the

In the above update process, due to the characteristics of flash memory, the minimum erasing unit is a data block. In order to rewrite the target data into the original data page, you can first copy the contents of other data pages in the data block to the cache, and then rewrite New data is written, so as to realize the update of the data block. The process of updating the data block is similar to the process of updating data in the prior art, and will not be repeated here. the

After the data is updated, the target data is stored in the modified page (that is, the original invalid page where the target data is written). At this time, the page-level logical mapping table can be adjusted to replace the original The entry of the first data page, the specific replacement process is similar to the aforementioned replacement process, and will not be repeated here. After the replacement is completed, the physical address in node E is changed to 8 again. the

It should be noted that since the page-level logical mapping table is set in this embodiment, certain system resources may be occupied. In order to reduce resource occupation, when the system is idle, or the resources occupied by the page-level logical mapping table reach the preset When the threshold value is higher, the page-level logical mapping table can be sorted out, and the specific sorting method can be as follows:

First obtain the data page to be sorted belonging to the same data block, and merge the valid data in the data block with the data in the data page to be sorted. The specific merge can be realized through the COPY_BACK operation. It is understandable that in practice In the application, other similar operations can also be used to realize data merging, which is not limited here. the

After the merging is completed, the merging data is copied to a blank block, and at the same time, the entry of the data page to be sorted is deleted in the page-level logical mapping table. the

Through the above sorting operation, the data in the data page and the data in the data block to which the data page belongs can be merged, thereby reducing the number of data pages, that is, reducing the content in the page-level logical mapping table. the

It should be noted that the above sorting manner is only an example of sorting the page-level logical mapping table, and in practical applications, other sorting ways can also be used, which are not limited here. the

In this embodiment, since the data block where the invalid page is located is not updated until the preset update condition is satisfied, it can be ensured that the data block does not need to be erased every time the data is updated, thereby saving the cost of data update. The process improves the performance of random data writing;

Secondly, since the binary tree is used as the storage method of the page-level logical mapping table in this embodiment, the query efficiency of the page-level logical mapping table can be effectively improved;

Thirdly, since the page-level logical mapping table can be organized in this embodiment, the occupation of system resources can be effectively reduced. the

To sum up, in this embodiment, since there is a preset page-level logical mapping table, and the page-level logical mapping table contains the physical address of each data page, the logical address, and the association relationship between each data page , so when updating the content of a data page, it is only necessary to replace the physical address of the data page in the page-level logical mapping table to update the data without erasing the entire Therefore, this embodiment can use the page-level logical mapping table to improve the performance of random writing of data. the

Introduce another embodiment of flash memory data processing method in the embodiment of the present invention below:

First of all, it needs to be explained that in the flash memory data storage process in the prior art, the smallest storage unit is the data page, and it can also be stored in the unit of the data block, but in the prior art, only the inter-block between the data blocks is established. The logical relationship table, that is, the block-level logical mapping table, when the user needs to store a small amount of data multiple times, it may cause each storage to occupy a data block, and most of the space of the data block may still be blank, so it will Waste a lot of storage space. the

An embodiment of the present invention provides a flash memory data processing method to solve the above defects. Please refer to FIG. 3 for details. The third embodiment of the flash memory data processing method in the embodiment of the present invention includes:

301. Obtain the length information of the data to be written;

When the user needs to write data, first obtain the length of the data to be written. The unit of the length can be a sector. Assume that a data block contains 64 data pages, and each data page contains 4 sectors. the

302. Split the data to be written into block-level data and page-level data according to the length information and the number of sectors contained in a data block;

In this embodiment, the length information of the data to be written is converted into the corresponding number of sectors L to be occupied (if the length of the data to be written is originally calculated in units of sectors, no conversion is required), the specific conversion method is as follows without limitation;

Get the number N of sectors contained in a data block;

Perform rounding operation on L and N, and determine the block-level data according to the result B of the rounding operation;

The remainder operation is performed on L and N, and the page-level data is determined according to the result of the remainder operation. the

Suppose the above L=260, N=64*4=256, then L div N=1, L mod N=4, then B=1, then the block-level data is the data that can be stored in a complete blank data block, The page-level data is data that can be stored in 4 sectors (that is, a data page). the

303. Write block-level data in a block writing manner, write page-level data in a page writing manner, and update a preset page-level logical mapping table. the

For ease of understanding, the following first introduces the preset process in this embodiment:

First, four data tables can be created, namely:

Block-level logical mapping table: used to record the logical mapping relationship of data blocks, and the one-to-one correspondence between logical addresses and physical addresses;

Page-level logical mapping table: used to record the logical mapping relationship of the data page, recorded in the form of a binary tree, including the logical address of the data page, the physical address and the relationship between each data page;

List of block erasure times: used to record the erasure times of each data block;

Unerased block list: used to record data blocks that have not been erased. the

It should be noted that the specific block-level logical mapping table can be shown in the following table:

Table 1

Block-level mapping table entry The corresponding logical address Corresponding physical address 0 0*n 0x00000000 1 1*n 0x00000005 2 2*n 0x80000010 ... ... ... N N*n 0xFFFFFFFF N+1 (N+1)*n 0xFFFFFFFF ... ... ...

Each item in the above Table 1 is implemented by 4 bytes, and each item identifies the correspondence between the logical address and the physical address of a data block. the

After the above settings are completed, the data writing operation can be performed. The following describes the block-level data writing and page-level data writing respectively: 

(1) Block-level data writing, as shown in Figure 4:

a. Select B blank data blocks from the preset blank data blocks;

When the system is powered on, the blank data blocks that have not been written in data can be found by traversing each data block, or the data blocks in the unerased block list can be erased to obtain blank data blocks. the

It should be noted that before erasing the data blocks in the unerased block list, you can first query the number of times each data block has been erased through the block erasure times list, and select the data blocks with fewer times to be erased first. Erasing is performed so that each data block is erased approximately the same number of times, thereby increasing the overall lifetime of the flash memory. the

b. Write block-level data into B blank data blocks;

If it is determined above that a blank data block is needed to store data, a part of the data can be read sequentially so that this part of data can just fill a data block, and this part of data can be stored in the data block. the

c. Fill the logical addresses and physical addresses of B blank data blocks into the preset block-level logical mapping table;

After the data filling is completed, the data block contains data, and the logical address and the physical address of the data block are filled into the block-level logical mapping table, which is used to indicate that data has been stored in the data block. the

d. Obtain the logical address range according to the logical addresses of the B blank data blocks;

Each data block corresponds to a logical address, and this step obtains the logical address range corresponding to the data block. the

e. Delete entries with logical addresses within the logical address range in the preset page-level logical mapping table. the

Since the data block is full of data, it is impossible to continue storing data in the data page included in the data block, and the entry whose logical address is within the logical address range can be deleted from the page-level logical mapping table. the

(2) Block-level data writing, as shown in Figure 5:

a. Select a blank data block from the preset blank data blocks;

When the system is powered on, the blank data blocks that have not been written in data can be found by traversing each data block, or the data blocks in the unerased block list can be erased to obtain blank data blocks. the

It should be noted that before erasing the data blocks in the unerased block list, you can first query the number of times each data block has been erased through the block erasure times list, and select the data blocks with fewer times to be erased first. Erasing is performed so that each data block is erased approximately the same number of times, thereby increasing the overall lifetime of the flash memory. the

b. Write the page-level data to the data pages in the blank data block in sequence;

From the above step 302, it can be seen that the space of 4 sectors (that is, one data page) is required to write page-level data, and the specific process of writing page-level data is consistent with the process of writing data in units of data pages in the prior art , is not limited here. the

c. Write the logical address and physical address of the data page containing the page-level data in the blank data block in the preset page-level logical mapping table, as well as the relationship between the data page and other data pages;

After the page-level data is written, the data page contains data, and a node can be added in the page-level logical mapping table to represent the data page. The specific adding process can be: get the data The logical address and physical address of the page, and determine the left node and right node of the data page according to the relationship between the data stored in the data page and other data, and then form a new node and insert it into the page-level logical mapping table. the

d. Modifying the flag of the blank data block in the preset block-level logical mapping table to indicate that the page-level logical mapping table exists in the blank data block. the

In this embodiment, after data is written in a data page in a certain blank data block, the content of the data block cannot be completely read only through the block-level logical mapping table, and the content of the data block needs to be read. Only the page-level logical mapping table can fully read the content of the data block. Therefore, the physical address of the data block needs to be adjusted in the block-level logical mapping table. Specifically, the highest bit of the corresponding physical address of the data block can be Change to 1 for identification. the

For example, the physical address of the data block shown in item "2" in Table 1 is 0x80000010, which is converted into a binary number: 1000 0000 0000 0000 0000 0000 0000 0001 0000, and its highest bit is 1, indicating that the data block contains pages When reading the content of the data block, the page-level logical mapping table needs to be read at the same time to completely obtain the content of the data block. the

In this embodiment, since the page-level logical mapping table is set, certain system resources may be occupied. In order to reduce resource occupation, when the system is idle, or the resource occupied by the page-level logical mapping table reaches a preset threshold value , the page-level logical mapping table can be sorted out, and the specific sorting method is the same as that in the foregoing embodiments, and will not be repeated here. the

In this embodiment, since the data to be written can be split into block-level data and page-level data, data blocks and data pages can be allocated for data writing at the same time during one write process, avoiding a small amount of data occupying a The case of a complete data block, thus saving storage space;

Secondly, in this embodiment, the number of times each data block is erased can be recorded, and the data blocks with a few times of erasing are preferentially selected for erasing, so that the number of times each data block is erased is approximately equal, thereby improving the performance of the flash memory. overall lifespan;

Thirdly, since the page-level logical mapping table can be organized in this embodiment, the occupation of system resources can be effectively reduced. the

To sum up, in this embodiment, since there is a preset page-level logical mapping table, and the page-level logical mapping table contains the physical address of each data page, the logical address, and the association relationship between each data page , so when writing page-level data to a blank data block, corresponding entries can be added to the page-level logical mapping table of the blank data block, so that the page-level data of several data pages can be stored in the blank data block Data, instead of the situation that the data of each data page needs to independently occupy a data block, so the storage space can be effectively saved. the

The two flash data processing methods described in the above embodiments are based on the preset page-level logical mapping table to manage the page-level data (including the page-level data stored in the flash memory and the page-level data written to the flash memory), so The performance of flash memory data processing can be effectively improved. the

The flash memory data processing device in the embodiment of the present invention is introduced below:

Please refer to Figure 6, the first embodiment of the flash memory data processing device in the embodiment of the present invention includes:

The source data query unit 601 is used to query the source data to be updated according to the acquired target data;

Writing unit 602, for writing the target data into the first data page of the preset blank data block;

Marking unit 603, used to mark the data page where the source data is located as an invalid page;

The replacement unit 604 is configured to replace the entry of the invalid page in the preset page-level logical mapping table with the entry of the first data page. the

The flash memory data processing device in the present embodiment can also further include:

The update trigger unit 605 is used to judge whether the preset update condition is met, and if so, trigger the update unit to perform corresponding operations;

The update unit 606 is configured to update the data block where the invalid page is located according to the target data in the first data page under the trigger of the update trigger unit. the

The update trigger unit 605 in this embodiment includes:

The first trigger unit 6051 is used to judge whether the number of invalid pages in the data block reaches a preset value, and if so, trigger the update unit to perform corresponding operations;

or,

The second trigger unit 6052 is used to judge whether it is currently idle, and if so, trigger the update unit to perform corresponding operations. the

The update unit 606 in this embodiment includes:

The data update unit 6061 is configured to save data in other data pages in the data block except the invalid page to the cache, erase the data block, obtain the target data in the first data page, and Write the modified page containing the target data and the data content stored in the cache into the data block after the erasure;

A mapping table updating unit 6062, configured to replace the entry of the first data page in the preset page-level logical mapping table with the entry of the modified page. the

The flash memory data processing device in the present embodiment can also further include:

Blank data block search unit 607, used for traversing each physical data block when the system is powered on, querying the physical data blocks of unrecorded data and the physical data blocks that have been erased, and searching the physical data blocks of the unrecorded data Blocks and erased physical data blocks are used as preset blank data blocks. the

The flash memory data processing device in the present embodiment can also further include:

The mapping table generating unit 608 is configured to establish a page-level logical mapping table, which includes the logical address of each data page storing data, the physical address, and the association relationship between the data page and other data pages . the

The flash memory data processing device in the present embodiment can also further include:

The mapping table organizing unit 609 is configured to organize the page-level logical mapping table. the

In this embodiment, the writing unit 602 writes the target data into the first data page of the preset blank data block, the marking unit 603 marks the data page where the source data is located as an invalid page, and the replacing unit 604 writes the preset page level The entry of the invalid page in the logical mapping table is replaced by the entry of the first data page. Therefore, in the process of data updating, there is no need to erase the data block where the invalid page is located, and directly use the first data block in the blank data block. A data page replaces the invalid page for data storage, thus saving the operation of data updating, thereby improving the performance of random writing of data. the

Please refer to Figure 7, the second embodiment of the flash memory data processing device in the embodiment of the present invention includes:

The length information acquisition unit 701 is used to acquire the length information of the data to be written;

A data splitting unit 702, configured to split the data to be written into block-level data and page-level data according to the length information and the number of sectors contained in a data block;

The data writing unit 703 is configured to write the block-level data in a block writing manner, and write the page-level data in a page writing manner. the

The data writing unit 703 in the present embodiment comprises:

The block writing unit 7031 is configured to select B blank data blocks from the preset blank data blocks, write the block-level data into the B blank data blocks, and write the logic of the B blank data blocks to Fill in the address and physical address into the preset block-level logical mapping table, obtain the logical address range according to the logical addresses of the B blank data blocks, and delete the logical address within the logical address range in the preset page-level logical mapping table entries in the table;

The page writing unit 7032 is used to select a blank data block M from the preset blank data blocks, and write the page-level data into the data pages in the blank data block M in sequence, at the preset page level Write the logical address and physical address of the data page containing the data in the logical mapping table, as well as the association between the data page and other data pages, and modify the blank data block in the preset block-level logical mapping table The flag of M is used to indicate that a page-level logical mapping table exists in the blank data block M. the

In this embodiment, since the data to be written can be split into block-level data and page-level data by the data splitting unit 702, and the block writing unit 7031 and the page writing unit 7032 in the data writing unit 703 respectively Therefore, data blocks and data pages can be allocated for data writing in one write process at the same time, which avoids the situation that a small amount of data occupies a complete data block, thereby saving storage space. the

Those of ordinary skill in the art can understand that all or part of the steps in the method of the above-mentioned embodiments can be completed by instructing related hardware through a program. The program can be stored in a computer-readable storage medium, and the program can be executed when executed , including the following steps:

Query the source data to be updated according to the obtained target data;

Write the target data into the first data page of the preset blank data block, and mark the data page where the source data is located as an invalid page;

and replacing the entry of the invalid page in the preset page-level logical mapping table with the entry of the first data page of the preset blank data block. the

The storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, and the like. the

A flash memory data processing method and device provided by the present invention has been introduced in detail above. For those of ordinary skill in the art, according to the idea of the embodiment of the present invention, there will be changes in the specific implementation and application scope. In summary, the contents of this specification should not be construed as limiting the present invention. the

Claims (21)

1. A flash memory data processing method, characterized in that, comprising: Query the source data to be updated according to the acquired target data; writing the target data into the first data page of the preset blank data block, and marking the data page where the source data is located as an invalid page; replacing the entry of the invalid page in the preset page-level logical mapping table with the entry of the first data page of the preset blank data block; The establishment process of the preset page-level logical mapping table includes: Obtain the logical address and physical address of each data page storing data, as well as the association between the data page and other data pages; Establishing the page-level logical mapping table according to the logical address and physical address of each data page storing data and the association relationship between the data page and other data pages; The replacing the entry of the invalid page in the preset page-level logical mapping table with the entry of the first data page of the preset blank data block includes: Query the table entry where the invalid page is located; modifying the physical address in the entry to the physical address of the first data page; The original association information of the entry is retained, and the association information is used to represent the association relationship between the entry and other entries. 2. The method according to claim 1, wherein after replacing the entry of the invalid page in the preset page-level logical mapping table with the entry of the first data page, the step comprises: Judging whether the preset update condition is satisfied, if yes, updating the data block where the invalid page is located according to the target data in the first data page. 3. The method according to claim 2, wherein said judging whether a preset update condition is satisfied comprises: It is judged whether the number of invalid pages in the data block reaches a preset value, and if so, it is determined that a preset update condition is met. 4. The method according to claim 2, wherein said judging whether a preset update condition is satisfied comprises: It is judged whether it is currently in an idle state, and if so, it is determined that the preset update condition is satisfied. 5. The method according to claim 2, 3 or 4, wherein updating the data block where the invalid page is located according to the target data in the first data page comprises: Save the data in other data pages in the data block except the invalid page to the cache; Erase the data block; Acquiring the target data in the first data page; Writing the target data into the invalid page to obtain a modified page; Writing the modified page containing the target data and the data content stored in the cache into the data block after the erasure; and replacing the entry of the first data page in the preset page-level logical mapping table with the entry of the modified page. 6. The method according to any one of claims 1 to 4, wherein the acquisition process of the preset blank data block comprises: When the system is powered on, traverse each physical data block; Query the physical data blocks of unrecorded data and the physical data blocks that have been erased; The physical data blocks that have not recorded data and the physical data blocks that have been erased are used as preset blank data blocks. 7. The method according to any one of claims 1 to 4, wherein the method further comprises: Organize the page-level logical mapping table. 8. The method according to claim 7, wherein said sorting out the page-level logical mapping table comprises: Determine the data pages to be organized belonging to the same data block; Merge the obtained valid data belonging to the same data block and the data in the data page to be sorted; Copy the merged data into a blank data block, and delete the entry of the data page to be sorted in the page-level logical mapping table. 9. A flash memory data processing method, characterized in that, comprising: Obtain the length information of the data to be written; Splitting the data to be written into block-level data and page-level data according to the length information and the number of sectors included in a data block; Writing the block-level data according to the block writing method, writing the page-level data according to the page writing method and updating the preset page-level logical mapping table; The establishment process of the preset page-level logical mapping table includes: Obtain the logical address and physical address of each data page storing data, as well as the association between the data page and other data pages; The page-level logical mapping table is established according to the logical address and physical address of each data page storing data and the association relationship between the data page and other data pages. 10. The method according to claim 9, wherein the splitting the data to be written into block-level data and page-level data according to the length information and the number of sectors included in a data block comprises: Converting the length information of the data to be written into the number L of sectors corresponding to the required occupation; Obtain the number N of sectors contained in a data block; performing a rounding operation on the L and N, and determining block-level data according to the result B of the rounding operation; A remainder operation is performed on the L and N, and page-level data is determined according to a result of the remainder operation. 11. The method according to claim 9, wherein the writing of the block-level data in a block writing manner comprises: Selecting the B blank data blocks from the preset blank data blocks; writing the block-level data into the B blank data blocks; Filling the logical addresses and physical addresses of the B blank data blocks into a preset block-level logical mapping table. 12. The method according to claim 11, characterized in that, after filling the logical addresses and physical addresses of B blank data blocks into the preset block-level logical mapping table, comprising: Acquiring a logical address range according to the logical addresses of the B blank data blocks; Deleting entries with logical addresses within the logical address range in the preset page-level logical mapping table. 13. The method according to claim 9, wherein the writing the page-level data according to the page writing method and updating the preset page-level logical mapping table comprises: Select a blank data block M from the preset blank data blocks; Writing the page-level data into data pages in the blank data block M in sequence; Write the logical address of the data page of the page-level data in the blank data block M in the preset page-level logical mapping table, the physical address and the association relationship between the data page and other data pages; Modifying the flag of the blank data block M in the preset block-level logical mapping table is used to indicate that the page-level logical mapping table exists in the blank data block M. 14. A flash memory data processing device, characterized in that, comprising: a source data query unit, configured to query the source data to be updated according to the acquired target data; a writing unit, configured to write the target data into the first data page of the preset blank data block; A marking unit, used to mark the data page where the source data is located as an invalid page; a replacement unit, configured to replace the entry of the invalid page in the preset page-level logic mapping table with the entry of the first data page of the preset blank data block, and the preset page-level logic Replacing the entry of the invalid page in the mapping table with the entry of the first data page of the preset blank data block includes: querying the entry of the invalid page; modifying the physical address in the entry is the physical address of the first data page; retain the original association information of the entry, and the association information is used to indicate the association relationship between the entry and other entries; The mapping table generation unit is used to establish a page-level logical mapping table, which includes the logical address and physical address of each data page storing data, and the association relationship between the data page and other data pages. 15. The device according to claim 14, further comprising: The update trigger unit is used to judge whether the preset update condition is satisfied, and if so, trigger the update unit to perform the corresponding operation; An update unit, configured to update the data block where the invalid page is located according to the target data in the first data page under the trigger of the update trigger unit. 16. The device according to claim 15, wherein the update trigger unit comprises: The first trigger unit is used to judge whether the number of invalid pages in the data block reaches a preset value, and if so, trigger the update unit to perform corresponding operations; or, The second trigger unit is used to judge whether it is currently in an idle state, and if so, trigger the update unit to perform a corresponding operation. 17. The device according to claim 14 or 15 or 16, wherein the updating unit comprises: A data updating unit, configured to store data in other data pages in the data block except the invalid page in the cache, erase the data block, obtain target data in the first data page, and store the data in the first data page. Writing the target data into the invalid page to obtain a modified page, writing the modified page containing the target data and the data content stored in the cache into the data block after the erasing; A mapping table updating unit, configured to replace the entry of the first data page in the preset page-level logical mapping table with the entry of the modified page. 18. The device according to claim 14 or 15 or 16, wherein the device further comprises: The blank data block search unit is used to traverse each physical data block when the system is powered on, query the physical data blocks of unrecorded data and the physical data blocks that have been erased, and search the physical data blocks of unrecorded data And the physical data block that has been erased is used as a preset blank data block. 19. The device according to claim 14 or 15 or 16, wherein the device further comprises: The mapping table organizing unit is configured to organize the page-level logical mapping table. 20. A flash memory data processing device, characterized in that, comprising: a length information acquisition unit, configured to acquire length information of the data to be written; A data splitting unit, configured to split the data to be written into block-level data and page-level data according to the length information and the number of sectors included in a data block; A data writing unit, configured to write the block-level data in a block writing manner, write the page-level data in a page writing manner, and update a preset page-level logical mapping table; The preset page-level logical mapping table is established in the following manner: Obtain the logical address and physical address of each data page storing data, as well as the association between the data page and other data pages; according to the logical address, physical address and the relationship between the data page and other The association relationship between data pages establishes the page-level logical mapping table. 21. The device according to claim 20, wherein the data writing unit comprises: The block writing unit is used to select B blank data blocks from the preset blank data blocks, write the block-level data into the B blank data blocks, and write the logical addresses of the B blank data blocks And the physical address is filled into the preset block-level logical mapping table, the logical address range is obtained according to the logical addresses of the B blank data blocks, and the logical address within the logical address range is deleted in the preset page-level logical mapping table table entry; The page writing unit is used to select a blank data block M from the preset blank data blocks, and write the page-level data into the data pages in the blank data block M in sequence, in the preset page-level logic In the mapping table, the logical address of the data page storing the page-level data in the blank data block M, the physical address, and the association between the data page and other data pages are written in the preset block-level logical mapping Modifying the flag of the blank data block M in the table is used to indicate that a page-level logical mapping table exists in the blank data block M.

Images