TW201216062A - Data protection method for damaged memory cells - Google Patents

Abstract

A data protection method for damaged memory cells is provided. In the method, a power-on self test (POST) is executed and an initial backup memory in the memory is reserved for the interrupt handler to use as backup pages. Then, an operating system is executed, in which a memory controller detects and corrects a correctable error occurred in each of the pages, accumulates the number of the detected correctable errors, and compares the accumulated number with a threshold. When the accumulated number is larger than the threshold, the corresponding page is determined as a damaged page and a backup page is statically obtained from the initial backup memory or dynamically obtained from the operating system, so as to backup data in the damaged page to the backup page and record the mapping addresses of the damaged page and the backup page in a page mapping table. Therefore, when the operating system accesses the damaged page, the memory controller changes to access the backup page according to the page mapping table.

Description

201216062 w 34904twf.doc/n VI. Description of the Invention: [Technical Field] The present invention relates to a data protection method and a data protection method for damaged memory cells. [^ Previous] With the development of electronic technology, all kinds of information exchanges are all kinds of storage media, and the memory of the towel is fast because of its fast operation. However, due to the t-transmission volume, the requirements for the transmission rate are also increasing, resulting in a gradual increase in the memory capacity and speed requirements of ^(4). However, while increasing the memory capacity, the density of the memory cells in the memory is also high. As the density of the hybrid circuits on the semiconductor memory elements increases, the size of each memory cell must be reduced. Therefore, in the case where a memory cell needs to obtain a smaller ruler and obtain its original capacitance, not only the process becomes complicated, but also the probability of a memory cell error increases due to the high-speed access of the system. The error of the memory can be divided into Correctable Error (CE) and Uncorrectable (Uncorrectable)

Error ' UE). Correctable Error (CE) is generally a Single Bit Error (SBE) or a multi-bit error in the number of bits in which the error occurred and the position within the correctable range of the electronic device (Multi-BitError, MBE). Among them, the error correction and error correction (EDC) can still be corrected by the error detection and correction (EDC) 201216062 100427.TW 34904twf.doc/n. For such errors, the conventional technique is to use a method of storing a S (Spare) memory module in the memory to use as a data backup when the memory is corrected. However, this spare memory consumes a large amount of memory space and causes a waste of memory, and it takes a considerable amount of time to use the memory for the data, resulting in a decrease in overall performance. Miscellaneous 1 aspect For the uncorrectable error, the conventional technique uses memory = half as a repetition (10) bribe to record the dragon, in order to make corrections to the riding when the texture is not correct. However, the result of using this half of the money will result in a waste of the journal and the overall effectiveness of the payment system. SUMMARY OF THE INVENTION (5) A data protection method for a damaged memory cell, by backing up the backup page to backup the data in the damaged memory cell can protect the data of the vine. The invention proposes that the (four) protection method of the damaged memory cell includes the electronic decoration of the memory, the unit of the pity, the body of the 愔罝H body including the memory controller, and the recording of the early morning 疋 Alt 7L consists of a plurality of memory cells. Divided into multiple pages "The cell system is divided into four parts. The four pages each include a part of the above memory cells. = The law system first executes the boot process and retains in the memory - = read 'body to provide for interrupt processing The module is used, the module " == coffee as the backup page of the beginning, this is = to the >, a backup page and is a multiple of the page. Then, 5 34904twf.doc/n 201216062 is ordered to operate the system, and when the operating system reads the memory, the memory of each page is corrected by recording the detection, and the current error can be corrected. After the accumulating, the number of times the above-mentioned pages can be corrected is compared with the value of ^, and when the number of correctable errors of the page is greater than the facade, the page is damaged and the backup memory is obtained: the backup of the page is ^ (r-snap backup or dynamic from the operating system to use up the initial backup memory page), back to the backup page, and the damage = backup page (four) mapping The page recorded in the memory controller is changed by the memory controller according to the mapping address in the page mapping table to access the data of the backup page. In addition, in the embodiment, the above-mentioned steps of executing the booting procedure are further included in the system data area of the miscellaneous, and the (four) section is reserved for the power: the interrupted module of the device and the work (4) In the embodiment, the above-mentioned step of executing the booting procedure further interrupts the processing module, and enables the step of executing the operating system when the correctable error occurs. I: Using the =, moving to the operating system / the above-mentioned execution of the operating system further includes: obtaining the base address and size of the initial memory of the memory segment of the memory of the reserved memory from the IOS, so the operation The system will not be able to use this memory as well. 201216062 100427.TW 34904twf.doc/n. In the embodiment of the present invention, the step of correcting the correctable error occurring in the memory cell in each page by the memory controller pre-measure includes including the correctable correction when the memory controller integrates the correction to the correctable error The wrong page address is recorded in the scratchpad and a system interrupt is issued to execute the processing module. In the embodiment of the present invention, the step of accumulating the number of times that each page can be repaired φ f error includes reading, by the interrupt processing module, the occurrence of the correctable error in the register of the memory controller. Page, and accumulate the number of times this page appears to correct the error. In an embodiment of the present invention, the dynamics obtains a backup from the operating system, and the step of restoring the body includes writing a page request by the interrupt processing module to the memory area of the data area (4), and then responding to the memory. The module examines the data of the memory segment of the data area of the system to view the page request sent by the interrupt processing module, and then obtains the backup memory from the operating system to provide the φ + break group, the situation processing module贱Backup memory is used as a backup page. In an embodiment of the present invention, the step of viewing, by the application module, the data of the memory segment of the system 4 to view the page request sent by the interrupt processing module includes using an inquiry by the application module ( The p〇mng) method examines the data of the memory segment of the system data area to view the page request issued by the interrupt processing module. In an embodiment of the present invention, the step of using the application module to the operating system to back up the memory for use by the interrupt processing module includes 201216062 luwz/.i W 34904twf.doc/n being backed up by the application module The base address and size of the memory are recorded in the memory segment reserved in the system data area for use by the interrupt processing module. In an embodiment of the present invention, the step of obtaining the backup memory by the application module for use by the interrupt processing module includes querying, by the interrupt processing module, a page usage table to obtain a vacancy from the backup memory. The page is used as a backup page, and the usage status of this backup page is updated to be in use, and the usage status of the damaged page is updated to be damaged. In an embodiment of the present invention, in the step of querying the page usage table by the interrupt processing module to obtain the vacant page from the backup memory as the backup page _, the backup memory is found when the interrupt processing module queries the page usage table. If there is no vacant page, the system interrupt function when the error occurs can be disabled. In an embodiment of the present invention, in the step of querying the page usage table by the interrupt processing module to obtain the vacant page from the backup memory as the backup page, if the interrupt processing module queries the page usage table, the backup memory is found. There are not many vacant pages, and a memory space can be dynamically obtained from the operating system as backup memory. In an embodiment of the present invention, in the step of querying the page usage table by the interrupt processing module to obtain a vacant page from the backup memory as a backup page, if the interrupt processing module queries the page usage table, the backup memory is found. If there is no vacant page, an IPMI OEM command is sent to the baseboard management controller to send a warning message to the baseboard management controller. In an embodiment of the present invention, the electronic device includes a basic input/output (Legacy Basic Input/Output System, Legacy BIOS) or a extended firmware interface basic input/output system. The system (Extensible Firmware Interface BIOS, EFI BIOS) is powered on. In an embodiment of the invention, the correctable error includes a unit error, or a number of bits of the error and a multi-bit error of the position within the correctable range of the electronic device. φ Based on the above, the data protection method for the damaged memory cell of the present invention uses the § replied address remapping method of the smear group to transfer the address of the damaged page to the address of the normal page, Data protection of damaged memory cells can be achieved without modifying the operating system and without preserving spare memory modules or repetitive memory. The above described features and advantages of the present invention will become more apparent from the description of the appended claims. [Embodiment] 曰Because the operating system can see the usage status of the entire memory, in order to improve the system performance, it will be managed in a way that is divided into pages, and the memory can be corrected in each page (Correctable When the number of Error's CE exceeds a certain amount, only this page is removed from the memory. In this way, in addition to preventing uncorrectable errors (Unc〇rrectable Error, UE), more memory can be reserved for the operating system, thereby increasing the overall performance of the system. The present invention is a data protection method for a genus s&'s memory that has been developed according to the above concept. The following is a detailed description of a detailed implementation of the method of 201216062 1UU42/.1 w 34904twf.doc/n. FIG. 1 is a schematic diagram of a system architecture of an electronic device according to an embodiment of the invention. Referring to the embodiment of the present invention, the processor 110, the memory 120, and the interrupt processing module 130 are included, and the memory 120 includes a memory controller 140 and a memory unit 15A composed of a plurality of memory cells. The processor 110 is, for example, a central processing unit (CPU), and the memory unit 150 is, for example, a dynamic random access memory (DRAM) or other kinds of memory, and is not provided here. limit. Further, the interrupt processing module 13 is, for example, a system management interrupt handler (SMI handler), and is used to receive a system management interrupt (SMI) received by the processor 110. In detail, when the component in the electronic device 100 triggers the system management interrupt, the system management interrupt message is sent to the processor 11〇, and the processor U〇 enters the system management mode when receiving the signal.

Management Mode (SMM), and the Basic Input Output System ('BIOS) executes the system management interrupt handler in this mode to manage system interrupts. 2 is a flow chart of a method for protecting a damaged memory cell according to an embodiment of the invention. Referring to Fig. 2, the method of the present embodiment is applied to the electronic device 1 of Fig. 1. The detailed steps of the method will be described below with reference to the elements in the electronic device 1A of Fig. 1. First, when the power of the electronic device 100 is turned on, the processor 1 executes the booting process, and an initial backup memory 201216062 100427.TW 34904twf.doc/n is reserved in the memory 120 (this size is a multiple of the page size) For example, if the page size is 8 KB, the initial backup memory can be 1 MB for use by the backup page, and the initial backup memory cannot be used by the operating system (step S202). The backup memory includes at least one backup page. The processor 110 is, for example, a conventional basic input/output system (Legacy Basic Input/Output).

System'Legacy BIOS) or Extendable Firmware Interface BIOS (EFI BIOS) for

Booting 'Reserving the initial backup memory is the E820h function call using ACPI INT 15h for Legacy BIOS and GetMemoryMap() for EFI BIOS. After the booting process is completed, the processor 11 then executes the operating system of the electronic device 1,, and when the operating system reads the memory 12〇, the memory controller 14G detects and corrects the memory of each page. _ can correct the error, (step S2.) The above-mentioned correctable error includes a unit error, or a number of bits in which the error occurred and a multi-bit error in the correct range of the electronic device, and there is no limit here. For example, the processor UG will first enable the system interrupt function when the error interrupt register is positive. When the memory device 142 贞 贞 正 到 可 可 可 可 可 可 可 可 可 可 可 142 142 142 142 Fix the wrong 1st address and issue =read ==module". and the interrupt processing mode = the error recorded in the hundred a 1 temporary storage 11 142 can be known which page has a correctable error Then, the interrupt processing module 130 accumulates the number of times that each page can correct the 201216062 100427.TW 34904twf.doc/n error and compares it with a threshold value, and determines whether the accumulated number of times exceeds the threshold value (step S206). ). The threshold value here is, for example, a preset value or a value set by the user. In detail, when the electronic device 100 is powered on, for example, a setting interface is displayed on the boot surface of the BIOS for use. When the interrupt processing module 13 determines that the number of correctable errors of the page is greater than the threshold value, the page is determined to be a damaged page, and a backup page is obtained by the backup memory to be in the damaged page. Backup of data to backup

Injury, face, it is known that the backup memory is obtained for the initial backup memory or dynamically by the operating system (step S208). After the data backup is completed, the interrupt processing module 130 records the mapped address of the damaged page and the backup page in the page=table 147 in the memory controller 140 (step S21G). The interrupt processing module 13 〇 〇 〇 疋 疋 疋 δ δ δ δ δ δ δ δ δ δ δ δ δ δ δ δ δ δ δ δ δ δ δ δ δ δ δ δ δ δ δ δ δ δ δ δ δ δ Fill in the damaged page and the position of the backup page in the white block

Controller, two == when the damaged page is taken, the memory control page; the address is changed, and the effect of accessing the material is changed. _ Protection of the memory cell's memory address remapping (rema_g) method, the processor 14. The weight of the backup page is the final access to the address 'but via the memory control map'

12 201216062 100427-TW 34904t^.doc/n address = The program of the operating system in this embodiment does not need to be changed. The processor is, in another example, 'the memory segment is reserved in the system data area during the process of executing the booting process in the handcuff device 110: 乍 is: 2 0 = group, and the operating system The information exchanged by the group should be described in detail with reference to an embodiment.

3 is a flow chart of a method of protecting a material according to an embodiment of the invention. Referring to Figure 3, and the electronic device 100' of Figure 1 below, the detailed steps of the method are illustrated in conjunction with the elements of Figure i. First, when the power of the electronic device (10) is turned on, that is, by the process of writing 11 〇 = 丁, the sequence (step S302), when executing the booting process, the processor (10), for example, is retained in the system data area of the memory 12 — - The memory segment is used as the interrupt processing module m of the electronic device 100 to exchange data with the application module of the system (step S304), wherein the device 110 is, for example, a reserved address between 9FC00h and 9FFFFh. ^ Memory section of the J ^ 1 kilobyte (KB). Dong Wei is worth noting that the above processor 110 is, for example, a memory area that executes the Legacy BIOS to boot, and the processor u〇 remains. The segment is, for example, located within an extension of the basic input/output system (Extende (j bios Data Area, EBDA). In another embodiment, the processor 11 is, for example, executing an extended firmware interface BIOS (EFI BIOS). Power on, and use this extended firmware interface to retain the memory segment. The reserved memory segment is used in the Legacy BIOS when using ACPI INT 15h 201216062 i \jwj. /. i vy 34904twf.doc/n E820h feature call And when using EFI BIOS is to use EF I BIOS interface

GetMem^yMapO. The above two are merely illustrative and are not limited herein. Then, the processor 110 retains the initial backup memory provided to the processing module 13 in the memory 12 as a backup page (step S306). It should be further noted that the interrupt processing module 13 uses the initial backup note as a backup page. When the page of the initial backup memory is fast, the interrupt processing module 13 can pass through the application module 17 Dynamically, the operating system 18G obtains an appropriate amount of backup memory (for example, 诵). When the page of the backup memory obtained by the operating system is also (4) finished, the backup memory can be taken from the operating system, and so on. The backup memory includes at least one backup page, and the address of the backup (four) block may be a continuous or discontinuous memory block. That is to say, the backup memory is a multiple of the single page capacity. For example, if the page capacity is eight kilobytes (8 KB), the backup memory capacity is a multiple of eight kilobits (for example, 1 MB). . Then 'processing n 1 ίο will start the interrupt processing module (step S3 〇 8). = The interrupt handling module is, for example, a system management interrupt handler that can be used to interrupt events. Further, the processor 11 is also enabled to correct the error of the system interrupt function (step S31G). That is, whenever the memory detection correction is corrected to the occurrence of the correctable error, the system management interrupt message is sent to the processor 11G, and the processing module 13G is processed. Among them, the processor 11G activates the system interrupt function when the error 00 is present, for example, by enabling the error interrupt register 141 of the controller ho. 201216062 100427.TW 34904twf.doc/n It should be noted that, in the process of executing the booting process by the processor 11G, the setting program of the Bluss (u bowl y) or other software tools can be executed according to the user's operation. For the user to modify the threshold for correcting the number of errors (for example, 5 times), or to modify the page size used by the memory controller (for example, 8 KB), the initial backup memory size is: 128 The memory of 8KB page size, which is 1MB) and the size of the backup record to the operating system each time (for example, the memory of 128 running page sizes, which is 1MB). After the processor 110 executes the booting process, the operating system is executed (step S312). At this time, the operating system can know the memory segment retained by the processor 11G in the memory 12G by the interface of the terminal 8. Back up the memory of the county and the big scorpion, so the operating system will not be able to use these memory blocks. When the operating system reads the memory 120, the processor no passes the memory control S14 (H is correcting the error that occurs in the memory of each of the f-faces (step S314). In detail, when the memory controller 14〇 When the detection is corrected to the correctable error, the page address where the correctable error occurs is recorded in the error page register 142, and a system interrupt is issued to execute the interrupt processing module 13〇. The interrupt processing module is executed. 13〇 also reads the page that can correct the error recorded in the error page register 142, and knows which page has a correctable error. Then 'the interrupt processing module 130 accumulates each page to correct the error. The number of illnesses ' is compared with a threshold value, and it is judged whether the accumulated number of times exceeds the threshold value (step S316). 15 201216062 34904twf.d〇c/n Field interrupt processing template address 13 〇 judges that there is a page correctable error When the number of times is greater than the door shuttle value, 'this page is determined to be a damaged page, and the backup page is obtained by the backup memory (step S318) to back up the data in the damaged page to the backup. The interruption is based on (10) using the initial η hidden body as the backup page 'but when the backup page is running out, the system must obtain the backup memory. In detail, the interrupt processing module γο The read and body segments retained in the system data 1 of the memory 12 by the processor 11 are exchanged with the module for the (4) system to obtain backup memory. 轳: Interrupt processing module 130 When the query page uses the table, it is found that the backup memory F vacant page is running out, and then the data is exchanged with the application module in the operating system through the tongue data section of the system data of the memory 120. From the operating system, the memory space is used as the backup memory. [4] FIG. 4 is an example of obtaining a backup page by the dynamic system according to the present invention. FIG. 5 is a view of the present invention. The method flow for obtaining the backup page from the operating system is not shown in FIG. 4 and FIG. 5 'This embodiment is obtained by the interrupt processing, the system data area 160, the application module 170, and the operating system 180. memory Crying 2 In this embodiment, the interrupt processing module 13G writes a page request in the above-mentioned memory stored in the system data area 160 of the memory unit 120 (step S502). The module 70 checks the memory of the system data area 160 to view the page sent by the interrupt processing module 13G (step S504). The application module 170 For example, the data of the memory segment of the system data area 160 is viewed by a polling method and the interrupt processing module 130 is checked whether the backup memory is being requested. When the application module 170 views the page request, the operating system 180' is called to obtain the backup memory (step S506), and the base address and size of the obtained backup memory are recorded in the system data area 16〇. In the stunning section (step S508). Finally, the interrupt processing module 13 obtains the backup memory by accessing the memory segment of the cryptographic region 160 (step S510). FIG. 6 is a diagram showing an example of a memory segment retained in a system data area according to an embodiment of the invention. The system data area 16 includes four fields, such as the request size 161, the request status 162, the response size 163, and the response base address 164, which are shown in FIG. The interrupt processing module 13 is, for example, writing the size of the backup memory to be requested in the request size 161 and writing 1 in the request state 162 to trigger the request of the backup memory. After the application module 170 obtains the backup memory from the operating system 18, the base address and the size record of the obtained backup memory are recorded in the response size 163 and the response base address 164, respectively, and are recorded after the recording is completed. The request status 162 is written to 完成 to complete the request. Thereby, the interrupt processing mode = bo obtains the base address and size of the requested backup memory by reading the response size 163 and the response base address 164. Returning to step S318 of FIG. 3, after the interrupt processing module 13 obtains the backup memory, it further queries a page usage table to obtain a vacant page from the backup memory as a backup page, and the backup page 17 The usage status of 201216062 100427.TW 34904twf.doc/n is updated to be in use, and the usage status of the damaged page is updated to be damaged (step S320). The page usage table is, for example, stored in a random access memory (RAM) of a System Management Mode (SMM), and is not limited herein. For example, FIG. 7(a) and FIG. 7(b) are examples of an updated page usage table according to an embodiment of the invention. Referring to FIG. 7(a) and FIG. 7(b), the page usage tables 710 and 720 of the embodiment include two address blocks: a page address and a page usage state, and the page usage table 710 records the damaged page and the backup page. The original state, and the page usage table 720 records the updated state of the damaged page and the backup page. In this embodiment, 〇〇 represents that the page is in a normal state; 01 represents a page in use state; 10 represents a damaged state; and 11 is a reserved state. 7(4) and 7(b), the number of digits of the page X in the page usage status field is changed from 01 to 10, which means that the state of the page X changes from the in-use state to the damaged state, and the page y is in the page usage state. The transition from 00 to 01 in the block indicates that the state of page y transitions from the normal state to the in-use state. It should be noted that if the interrupt processing module 130 queries the page usage table and finds that the backup memory has no vacant pages, and can no longer obtain the memory space from the operating system, it means that the backup memory cannot be used to back up the damaged pages. At this time, the interrupt processing module 130 disables the system interrupt function when the correctable error occurs, or transmits the IPMI OEM command to the Baseboard Management Controller (BMC) to be controlled by the baseboard management. The device issues a warning message and informs the system administrator to perform a replacement of the memory unit 150 or other suitable 201216062 100427.TW 34904twf.doc/n action. After updating the page usage table, the interrupt processing module 13 备份 backs up the data in the damaged page to the backup page (step S322). The interrupt processing module 130 writes, for example, the size of the page into the page size register 143 in the memory controller 14 , and writes the addresses of the damaged page and the backup page into the memory controller 140 respectively. The source page register 144, the destination page register 145' and enable (set to 1) the backup register 146, and control the memory controller 14 to back up the data in the damaged page to the backup page. By checking whether the status of the backup register 146 is replied (set to 〇), it can be known whether the data backup is completed. After the data backup is completed, the interrupt processing module 13 records the damaged area and the backup page © the mapping table 147 of the memory controller (steps). The 'interrupt processing mode (four) 〇; ==:; = rping) method, the data of the memory control page will be damaged (step S32 = the mapping address in the table, the effect of accessing the preparation material) The data protection method that protects the damaged memory cell and divides the memory into the data = the number of times the error can be corrected is more than - when the quantitation occurs in the cell τ 〇 this page from the memory 201216062 100427.TW 34904twf.d〇c/n Removed. The method of removing the page by using the asymptotic method can not only prevent the occurrence of the unrepairable error, but also use the dynamic mode. When the page Φ of the backup memory is used up, the operating system is required to block the memory block. In the progressive mode, the operating system requires memory blocks, so more can be provided. The use of the memory system for the operating system can therefore increase the overall efficiency of the system. The present invention has been disclosed in the above embodiments, and it is not intended to limit the invention to those skilled in the art, and it is possible to make some modifications and refinements without departing from the spirit and scope of the invention. The scope of the invention is defined by the scope of the appended claims. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a schematic diagram of a system architecture of an electronic device according to an embodiment of the invention. 2 is a flow chart of a method for protecting a damaged memory cell according to an embodiment of the invention. 3 is a flow chart of a method for protecting a damaged memory cell according to an embodiment of the invention. FIG. 4 is a diagram showing an example of dynamically obtaining a backup page from an operating system according to an embodiment of the invention. FIG. 5 is a flow chart of a method for dynamically obtaining a backup page from an operating system according to an embodiment of the invention. FIG. 6 is an illustration of a memory segment retained in a system data area 16A according to an embodiment of the invention. 7(a) and 7(b) are diagrams showing an example of an update 20 201216062 100427.TW 34904twf.doc/n page usage table according to an embodiment of the invention. [Main component symbol description] 100: electronic device 110: processor 120: memory 130: interrupt processing module 140: memory controller 141: error interrupt register 142: error page register 143: page size temporary storage 144: Source page register 145: Destination page register 146: Backup register 147: Page mapping table 150: Memory unit 160: 育 cultivating area 161: Request size 162: Request status 163: Response size 164: response base address 170: application module 180: operating system 710, 720: page usage table 21 201216062 ΐυϋ 42 /.1 W 34904twf.doc / n S202 ~ S212: data of damaged memory cells according to an embodiment of the present invention Steps S302 to S326 of the protection method: Steps S502 to S510 of the data protection method for the damaged memory cell according to an embodiment of the present invention: a method for dynamically obtaining a backup page from the operating system according to an embodiment of the present invention

twenty two

Claims (1)

201216062 100427.TW 34904twf.doc/n VII. Patent application scope: 1. A data protection method for damaged memory cells, which is applicable to including a memory = memory device and a memory controller and a division = The pages, each of which includes the memory cells: The method includes the following steps: an execution-start procedure, and the memory is provided in the memory to provide an initial interrupt to the interrupt processing mode. At least - backup page; mode touch with 'such as read backup memory package execution - operating system, and in the operating system read == side test correction each of these pages memory two pages each of these pages appear the correctable error The number of times, and when it is the correct value of the shot, it is determined that the page is - alleged; ^ number is greater than the threshold or dynamic from the difficult system to obtain the 2 backup memory backup page... Obtaining at least one page in the remaining 5; and the data in the damaged page to the backup memory:===-the mapping address is recorded by the memory controller . The mapping address 'is changed to the backup page 23 201216062 I00427..1-VV 34904twf.doc/n 2. The data protection method for the damaged memory cell as described in the scope of the patent application, wherein the booting is performed The steps of the program further include: retaining a memory segment in a system data area of the memory to exchange data between the interrupt processing module of the electronic device and the application module of the operating system. ~ 3. The method for protecting data of damaged memory cells as described in claim 2, wherein the step of executing the operating system further comprises: 'taking the memory area of the memory retained by the operating system A base address and size of the segment and the initial backup memory. " 4. The data protection method for damaged memory cells according to claim 3, wherein the memory controller detects and corrects the correctable error occurring in the memory cells in each of the pages, and accumulates The step of correcting the number of errors in each of the faces includes: when the memory controller detects the correction to the correctable error, the page in which the correctable error occurs is in a register and issues a == Executing the interrupt processing module; ... - reading, by the interrupt processing module, the page recorded in the memory controller where the correctable error occurs; and accumulating the number of times the page has the correctable error. (4) The data of the 4 rhyme green cell == method, wherein the operating system obtains the backup memory by the interrupt processing module to write the person-page request reserved in the memory segment; 201216062 100427.TW 34904twf.doc/n The data of the observatory section of the department and the Kasuga area is inspected by the group to view the request of the page issued by the group of the towel and the application model The group obtains the backup memory from the operating system for use by the interrupt processing module. 6. If you apply for the information on the damaged memory cells mentioned in item 5 of the full-time division, the method of compliance is as follows: The data of the body segment's step of reviewing the page request issued by the processing module includes: (4) by the application module (4)-interrogation method or the interruption method, the system data 5 is found (4) View the page request sent by the towel processing module. 7. If you apply for the data protection method for damaged memory cells as described in item 5 of the full-time division, the + (4) should be modeled. (4) The operation is to obtain the backup memory to provide the steps for the interrupt processing module. The method includes: recording, by the application module, a base address of the backup memory and a memory segment reserved in the data area of the data system for use by the interrupt processing module. 8. The data protection method for damaged memory cells according to claim 5, wherein the step of obtaining the backup memory by the application module for use by the interrupt processing module comprises: processing by the interrupt The module queries a page usage table to obtain a vacant page from the backup memory as the backup page, and updates a usage status of the backup page to be in use, and updates the usage status of the damaged page to be damaged. . 25 201216062 iOWUA w 34904twf.doc/n 9. The method of claim s, wherein the vacant page is obtained from the data memory of the damaged memory cell described in this item as the page The step further includes: when the interrupt processing module queries the page usage table and finds that the number of the vacant pages of the backup memory is less than a preset value, dynamically obtaining from the operating system - memory empty _ as the backup memory body. 10. The data protection method for damaged memory cells according to claim 1, wherein the electronic device comprises a legacy basic input/output system (Legacy BIOS) or an extended firmware interface. The basic input/output system (Extensible Firmware Interface BIOS, EFI BIOS) is turned on. 26