JP2007328824A - Memory card, nonvolatile memory, data writing method of nonvolatile memory and data writing device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To surely prevent the lifetime of a nonvolatile memory from becoming short by preventing management data from being rewritten when determining a defective block resulting from a systematic factor. <P>SOLUTION: This nonvolatile memory card comprises the means for: updating a block use management table for managing a data writing state of a memory and inhibiting rewriting to the block when data writing of a block unit is normally executed; selecting an empty block and retrying data with reference to an error management table (143) in which error management data is registered about the block when writing of the data is not normally executed and the error management table according to a rewriting request from a host; and clearing the error management data registered in the error management table when data writing is normally executed by retrying. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention determines a defective block based on an error when writing data to a nonvolatile memory, confirms that the subsequent writing process has been executed correctly, and manages the preceding block as a defective block, a nonvolatile memory, The present invention relates to a data writing method and a data writing apparatus for a nonvolatile memory.

  In recent years, memory cards equipped with a rewritable nonvolatile memory that does not have a risk of losing stored information even when the power is turned off are spreading as a storage device for storing various digital information represented by image data and music data. ing.

  A typical type of rewritable nonvolatile memory is a NAND flash memory. This type of flash memory is managed in units of blocks, and “1” is written as management data for unused blocks. When data is written, the management data is rewritten from “1” to “0”. Further, when data is written, it is determined whether or not the data has been normally written. When an error occurs, the management data of the block is rewritten from “1” to “0”. When a write error occurs, retry processing is performed on different blocks of the nonvolatile memory. In this retry process, if an error occurs, the management data of the block is rewritten to “0” (see, for example, Patent Document 1).

As described above, whenever an error occurs during data writing, the management data of the corresponding block is rewritten from “1” to “0”. Note that there are two states for writing data to the nonvolatile memory, and there are a single write that is completed by a minimum writing unit and a multiple write that is completed by performing this multiple times.
JP-A-6-3232806

  When data is written as described above, it is determined whether or not the block is a bad block. The determination that the block is a bad block is not only a factor of the memory itself but also a systematic There may be cases caused by factors. Therefore, when it is determined that the block is defective when data is written as in the conventional case, the management data of the corresponding block is rewritten from “1” to “0”, and the block determined to be defective cannot be used immediately. By doing so, the block cannot be used even if the memory itself is not defective. Therefore, the conventional nonvolatile memory data writing method has a problem that the life of the memory card is shortened.

  The present invention has been made in order to solve the above-described problem. For the determination of a bad block due to system factors, rewriting of management data is prevented, and the lifetime of the nonvolatile memory is shortened. It aims to be able to prevent reliably.

  According to one embodiment of the present invention, in a memory card equipped with a rewritable nonvolatile memory, a block usage management table for managing a data writing state of the nonvolatile memory, and data in a block unit to the nonvolatile memory. The writing means and whether or not the data writing has been normally executed are determined. When the data writing has been normally executed, the block use management table is updated, and rewriting to the corresponding block is prohibited. Means, an error management table in which error management data is registered for the block when the data writing is not executed normally, and an empty block by referring to the error management table by a rewrite request from the host device. The means to select and retry writing data and the above retry What the data writing is intended and means for clearing the error management data registered in the error management table when it is executed successfully.

  According to another embodiment of the present invention, a first step of writing data to the nonvolatile memory, a second step of determining whether or not the data writing has been normally executed, and the data writing is normal The third step of updating the block usage management table when it is executed and prohibiting rewriting to that block, and error management data for the block if the data writing is not executed normally A fourth step of registering in the management table, a fifth step of selecting and retrying a free block by referring to the error management table in response to a rewrite request from the host device, and writing of data normally by the retry When executed, the error management data registered in the error management table is cleared. Data writing method for a nonvolatile memory comprising the steps is provided.

  According to another embodiment of the present invention, a block use management table for managing a data write state of the nonvolatile memory, means for writing data to the nonvolatile memory in units of blocks, and the data writing is executed normally. Means for updating the block usage management table when data writing is executed normally and prohibiting rewriting to the corresponding block, and the data writing is not executed normally. An error management table in which error management data is registered for the block, a means for selecting an empty block by referring to the error management table in response to a rewrite request from the host device, and retrying data writing, and the retry If the data is successfully written by Non-volatile memory and means for clearing the error management data registered in the table is provided.

  According to another embodiment of the present invention, in a device for writing data to a rewritable nonvolatile memory, a block usage management table for managing a data writing state of the nonvolatile memory, and data in blocks to the nonvolatile memory , And whether or not the data writing has been executed normally, and if the data writing has been executed normally, the block usage management table is updated and rewriting to the corresponding block is prohibited. An error management table in which error management data is registered for the block when the data writing is not executed normally, and an empty block referring to the error management table by a rewrite request from the host device. Means to select data and retry writing data, and the retry Therefore and means for clearing the error management data writing of data is registered in the error management table when it is executed successfully.

  According to the present invention, it is possible to prevent management data from being rewritten and to prevent the lifetime of the nonvolatile memory from being shortened with respect to the determination of a defective block due to system factors.

  Hereinafter, an embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing the overall configuration of a memory card according to an embodiment of the present invention.

  The memory card shown in FIG. 1 is used by being mounted on various electronic devices such as a personal computer, an electronic camera, and a game machine, and is rewritable with a host interface 11, a controller 12, a ROM 13, and a RAM 14. A non-volatile memory such as a flash memory 15 is provided.

  The host interface 11 serves as an interface with an electronic device main body (hereinafter referred to as a host device) to which the memory card is mounted. The controller 12 controls the entire card, and has a microprocessor (MPU) 121 and a register (REG) 122. The microprocessor 121 interprets and executes a command received via the host interface 11 in accordance with a control program (firmware) stored in the ROM 13.

  The ROM 13 is a read-only nonvolatile memory. The ROM 13 is used to store the control program and management fixed data in advance. The RAM 14 is a volatile memory and includes a work area 141 of the controller 12, a block use management table 142, an error management table 143, a write data buffer 144, and the like. As shown in FIG. 2, the block usage management table 142 stores an unused state and a used or defective block state for each block of the flash memory 15 to manage the block usage state. In the block usage management table 142, as shown in FIG. 3A, when each block of the flash memory 15 is unused, “1” is written to each block. When data is written in units of blocks, “1” is rewritten to “0” as shown in FIG. 3B, indicating that the block is in use. Further, “0” is also written to the defective block, and the subsequent use is prohibited.

  The error management table 143 is used in the case of multiple write processing, and is registered as defective data when an error occurs during data writing.

  Each block of the flash memory 15 is composed of a plurality of sectors of a certain size. The area of the flash memory 15 includes a management area 151 for storing system management information and the like, and a data area 152 for storing digital contents such as image data and music data. The management area 151 stores important information such as a content decryption key, copy control information, and movement control information necessary for using the digital contents stored in the data area 152.

  Next, a data writing process for the memory card will be described. First, the single write process will be described with reference to the flowchart shown in FIG. When performing a single write process according to a command from the host device, the controller 12 first calculates a write address (step A1). That is, a writable area is found by referring to the block usage management data written in the block usage management table 142. In this case, since the bad block is managed as used, it is not used again.

  Then, data is written to the write address (step A2). In this case, data sent from the host device is held in block units in the write data buffer 144 of the RAM 14 and written to the write address. Thereafter, it is determined whether or not the write status, that is, the data write processing status is normal (step A3).

  Even if an abnormal status is found in step A3, the defective block is not immediately registered in the block usage management table 142. That is, if it is determined in step A3 that the write status is abnormal, the preceding state is determined to be abnormal (step A4), and it is determined from the number of retries whether the retry is continued or terminated (step A5). ). In this step A5, it is determined whether or not to continue the retry depending on whether or not the preset number of retries has been reached. If the number of retries has not reached the set value, the process returns to step A1 to find a different block. To rewrite the data. If it is determined in step A5 that the number of retries has reached the set value, the process proceeds to step A6 to end the error, and the single write process ends.

  If it is determined in step A3 that the write status is normal, it is determined whether or not the data writing is a retry process (step A7). The state, that is, whether or not the previous writing is normal is determined (step A8). In the case of writing abnormality, it is determined as defective data, the address used at the time of the previous write is set as a defective block (step A9), and after the preceding state is made normal at step A10, the block usage management table 142 is updated (step A11). For example, as shown in FIG. 3A, when all of the block usage management table 142 is in an unused state, the first data write results in an error and retry processing is performed, and as a result, data can be written correctly. As shown in FIG. 3B, the management data for the block is rewritten to “0” to be used, and the block written immediately before is rewritten to “0” as a defective block.

  Further, when data is normally written by retry processing as described above, the previous block is already in use as shown in FIG. 3C, and the previous block is the previous one. If the block has been written, as shown in FIG. 3 (d), "0" is written to the current block that has been normally written, and the previous block to which the previous writing was performed is defective. Write “0” as a block.

  If it is determined in step A7 that it is not a retry process, or if it is determined in step A8 that the preceding state is normal, the block use management table 142 is set after the preceding state is made normal in step A10. Is updated (step A11). In this case, “0” is written to the block in which the data has been normally written to make it in use.

  Thereafter, the single write process is terminated as normal termination (step A12).

  As described above, in the single write, even if an error occurs when data is written, the bad block is not registered in the block use management table 142, and then the process is normally terminated by the retry process. Since the block usage management table 142 is updated with the address used at the previous writing as a bad block, the block usage management table 142 is updated only when an error occurs due to the cause of the memory itself. Therefore, in the case of an error due to a system factor, the block usage management table 142 is not updated, and even when retry is repeatedly performed, it is possible to avoid bad block determination for the rewrite execution attempt. It is possible to reliably prevent the lifetime of the nonvolatile memory from being shortened.

  Next, data write processing by multiple write will be described with reference to the flowchart shown in FIG.

  When a write by multiple write is instructed from the host device, the controller 12 first selects an available writable block in the flash memory 15 using the management data written in the block use management table 142 (step B1) Data is written to this empty block (step B2). In this case, data sent from the host device is sequentially written to the write data buffer 144 of the RAM 14, and this data is written to the empty block. Thereafter, it is determined whether or not the write status, that is, the data write processing status is normal (step B3).

  If it is determined in step B3 that the write status is abnormal, it is determined that the preceding state is abnormal, and error data is registered in the error management table 143 of the RAM 14 (step B4). When a write error occurs, data indicating the occurrence of the write error is sent from the controller 12 to the host device via the host interface 11. Thereby, a rewrite request is instructed from the host device (step B5).

  In response to the rewrite request, the controller 12 refers to the error management table 143, selects another free block (step B6), and performs a retry process on the free block (step B7). That is, the data sent from the host device is sequentially held in the write data buffer 144 and written to the free block.

  Then, it is determined whether or not the data has been normally written for the retry process (step B8). If a write error has occurred, the process returns to step B5 and the above process is repeated.

  If it is determined in step B8 that the writing process has been completed normally, the error management data registered in the error management table 143 is cleared and the block usage management table 142 is updated (step B9). Then, the multiple write process is terminated as normal termination (step B10).

  When an error occurs during multiple write, only a part of the write data is held in the write data buffer 144 of the RAM 14. For this reason, the retry based on the data held in the write data buffer 144 cannot be performed, and the block that was previously written by the act of being able to write normally by the retry as in the case of the single write is defective. It cannot be determined to be a block. Further, there may be a situation in which a defective block cannot be formed again due to a write re-request from the host device.

  Therefore, in the multiple write processing according to the present invention, as described above, the block in which a write error has occurred is registered in the error management table 143 and is managed separately from the block use management table 142. Thereafter, in response to a rewrite request from the host device, a retry operation can be performed by selecting an empty block based on the error management data registered in the error management table 143. The error management data registered in the error management table 143 is cleared when the retry operation is correctly performed.

  In the case of the multiple write, if an error occurs, it cannot be determined whether the error is caused by the memory itself or caused by a system factor, and therefore the block usage management table 142 is not updated. Even if a block in which an error has been determined to be left is left as an unused area as it is, in the single write processing described with reference to FIG. There is no problem because it can be judged whether it is caused by the factor. As described above, in the case of multiple write processing, the block usage management table 142 is not updated even if an error occurs, and it is possible to reliably prevent the lifetime of the nonvolatile memory from being shortened.

  In the above embodiment, the case where the present invention is applied to a memory card including a nonvolatile memory has been described. However, for example, the present invention can also be applied to a nonvolatile memory built in a computer in the same manner as in the above embodiment. It is.

  As described above in detail, according to the present invention, even if an error occurs when data is written to the nonvolatile memory by the single write process, the defective block is not immediately registered in the block use management table. When the writing is successful by retry processing, the block usage management table is updated with the address used at the previous writing as a bad block, so the block is used only when an error occurs due to the cause of the memory itself. The management table is updated. Therefore, in the case of an error due to system factors, the block usage management table is not updated, and it is possible to reliably prevent the lifetime of the nonvolatile memory from being shortened. In the multiple write processing, when a write error occurs, error data is registered in the error management table, and a free block is created based on the error management data registered in the error management table by a rewrite request from the host device. The error management table is cleared when the retry operation is performed correctly, and the block usage management table is not updated due to an error due to system factors. The memory can be used efficiently to extend the life.

  Note that the present invention is not limited to the above-described embodiment as it is, and can be embodied by modifying the constituent elements without departing from the scope of the invention in the implementation stage. In addition, various inventions can be formed by appropriately combining a plurality of constituent elements disclosed in the embodiment. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, you may combine suitably the component covering different embodiment.

1 is a block diagram showing an overall configuration of a memory card according to an embodiment of the present invention. The figure which shows the structural example of the block use management table in the same embodiment. The figure which shows the management data example of the block use management table in the embodiment. 6 is a flowchart showing an operation of single write processing in the embodiment. 6 is a flowchart showing the operation of multiple write processing in the embodiment.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 11 ... Host interface, 12 ... Controller, 121 ... Microprocessor (MPU), 122 ... Register (REG), 13 ... ROM, 14 ... RAM, 141 ... Working area, 142 ... Block use management table, 143 ... Error management table 144: Write data buffer, 15 ... Flash memory, 151 ... Management area

Claims (4)

In memory cards with rewritable nonvolatile memory,
A block use management table for managing the data write state of the nonvolatile memory;
Means for writing data in block units to the non-volatile memory;
Means for determining whether or not the data writing has been normally executed, and updating the block use management table when the data writing has been normally executed, and prohibiting rewriting to the corresponding block;
An error management table in which error management data is registered for the block when the above data writing is not executed normally;
Means for referring to the error management table in response to a rewrite request from the host device, selecting a free block, and retrying data writing;
A memory card comprising: means for clearing error management data registered in the error management table when data writing is normally executed by the retry. A first step of writing data to the non-volatile memory;
A second step of determining whether or not data writing has been executed normally;
A third step of updating the block use management table when the data writing is normally executed and prohibiting rewriting to the block;
A fourth step of registering error management data for the block in the error management table when the data writing is not executed normally;
A fifth step of selecting a free block by referring to the error management table in response to a rewrite request from the host device and retrying;
And a sixth step of clearing the error management data registered in the error management table when the data writing is normally executed by the retry. A block usage management table for managing the data write state of the nonvolatile memory;
Means for writing data in block units to the non-volatile memory;
Means for determining whether or not the data writing has been normally executed, and updating the block use management table when the data writing has been normally executed, and prohibiting rewriting to the corresponding block;
An error management table in which error management data is registered for the block when the above data writing is not executed normally;
Means for referring to the error management table in response to a rewrite request from the host device, selecting a free block, and retrying data writing;
A non-volatile memory comprising: means for clearing error management data registered in the error management table when data writing is normally executed by the retry. In a device for writing data to a rewritable nonvolatile memory,
A block use management table for managing the data write state of the nonvolatile memory;
Means for writing data in block units to the non-volatile memory;
Means for determining whether or not the data writing has been normally executed, and updating the block use management table when the data writing has been normally executed, and prohibiting rewriting to the corresponding block;
An error management table in which error management data is registered for the block when the above data writing is not executed normally;
Means for referring to the error management table in response to a rewrite request from the host device, selecting a free block, and retrying data writing;
A data writing apparatus comprising: means for clearing error management data registered in the error management table when data writing is normally executed by the retry.

Images