CN103218275B - Error in data restorative procedure, device and equipment - Google Patents

Abstract

The invention discloses a data error recovery method, device and equipment, belonging to the field of terminal equipment. The method includes: judging whether a preset counter in the internal memory overflows, the preset counter is used to count data errors in the internal memory; if the preset counter overflows, according to the data errors stored in the internal memory The physical address is used to determine the failure type of the memory, so that subsequent repairs can be performed accordingly. The present invention effectively distinguishes the failure type of the memory according to the physical address stored in the preset storage space, and repairs the failure type according to the failure type, avoiding the system hanging or failing to start caused by the accumulation of data errors, and ensuring the normal operation conduct.

Description

Data error recovery method, device and equipment

technical field

The invention relates to the field of computer technology, in particular to a method, device and equipment for repairing data errors.

Background technique

As an essential part of a computer system, memory usually exists in systems of different architectures in the form of memory sticks. During system operation, hard failure or soft failure may occur in the memory. Hard failures refer to data errors that cannot be recovered due to hardware problems, and soft failures refer to data errors that can be recovered by power-on, power-off or restart due to data jumps. In order to maintain the normal operation of the system, data errors introduced by hard failures and soft failures need to be repaired.

The repair method in the prior art is generally carried out by adding an ECC (Error Checking and Correction, error detection and correction) verification chip on the memory stick. When the data in the memory has a data error, the ECC will output the correct data to the user after detecting the data error. .

In the process of realizing the present invention, the inventor finds that there are at least the following problems in the prior art:

ECC only outputs correct data to the user according to data errors, and does not perform any repair actions on the wrong data in the memory. ECC cannot effectively distinguish between hard failures and soft failures, and thus cannot repair erroneous data, making the accumulation of erroneous data likely to cause system hangs, system failures, etc., affecting normal business operations.

Contents of the invention

In order to solve the problem of distinguishing and processing soft and hard failures, the embodiments of the present invention provide a data error recovery method, device and equipment. Described technical scheme is as follows:

In a first aspect, a method for repairing data errors is provided, the method comprising:

judging whether a preset counter in the memory overflows, and the preset counter is used to count data errors in the memory;

If the preset counter overflows, the failure type of the memory is determined according to the physical address of the data error stored in the memory, so as to be repaired accordingly.

With reference to the first aspect, in the first possible implementation of the embodiment of the present invention, if the preset counter overflows, determine the failure type of the memory according to the physical address stored in the memory where a data error occurs, so that subsequent Make appropriate repairs, including:

If the same physical address exists in the physical addresses stored in the internal memory where the data error occurs, it is determined that the failure type of the data error corresponding to the same physical address is a hard failure.

In combination with the first aspect, in the second possible implementation of the embodiment of the present invention, if the preset counter overflows, the failure type of the memory is determined according to the physical address of the data error stored in the memory, so that subsequent Make appropriate repairs, including:

If the same physical address does not exist in the physical address where the data error occurs stored in the memory, memory inspection is performed;

After the inspection is finished, it is judged whether the data error in the memory has been repaired;

If the data error has not been repaired, determining that the failure type of the data error is a hard failure;

If the data error has been repaired, it is determined that the failure type of the data error is a soft failure.

In combination with the first aspect, in the third possible implementation of the embodiment of the present invention, if the same physical address does not exist in the physical address stored in the memory where a data error occurs, memory inspection is performed, including:

If the same physical address does not exist in the physical address where the data error occurs stored in the memory, converting the physical address in the preset counter into a patrol address;

Perform inspection according to the data in the internal memory corresponding to the inspection address.

In combination with the first aspect, in the fourth possible implementation of the embodiment of the present invention, if the preset counter overflows, the failure type of the memory is determined according to the physical address stored in the memory where a data error occurs, so that subsequent After performing corresponding repairs, the method also includes:

When it is determined that the failure type of the memory is a hard failure, acquiring the physical address of the data error whose failure type is a hard failure;

An alarm is triggered so as to prompt the user to replace the memory corresponding to the physical address of the data error whose failure type is hard failure.

In combination with the first aspect, in the fifth possible implementation of the embodiment of the present invention, before judging whether the preset counter in the memory overflows, the method further includes:

When a data error occurs in the memory, obtain the physical address where the data error occurs;

The physical address where the data error occurs is stored in the memory, and the data is written back to the physical address where the data error occurs.

In a second aspect, a device for repairing data errors is provided, the device comprising:

A judging module, configured to judge whether a preset counter in the memory overflows, and the preset counter is used to count data errors in the memory;

The determining module is configured to, if the preset counter overflows, determine the failure type of the memory according to the physical address stored in the memory where a data error occurs, so as to perform subsequent corresponding repairs.

With reference to the second aspect, in the first possible implementation manner of the embodiment of the present invention, the determining module is configured to determine whether the same physical address exists in the physical address stored in the memory where a data error occurs. The failure type of the data error corresponding to the address is a hard failure.

With reference to the second aspect, in a second possible implementation manner of the embodiment of the present invention, the determining module includes:

The inspection unit is used to perform memory inspection if the same physical address does not exist in the physical address where the data error occurs stored in the memory;

a judging unit, configured to judge whether the data error in the memory has been repaired after the inspection is finished;

a determining unit, configured to determine that the failure type of the data error is a hard failure if the data error has not been repaired;

The determining unit is configured to determine that the failure type of the data error is a soft failure if the data error has been repaired.

With reference to the second aspect, in a third possible implementation manner of the embodiment of the present invention, the patrol unit includes:

The patrol address conversion subunit is used to convert the physical address in the preset counter into a patrol address if the same physical address does not exist in the physical address stored in the memory where a data error occurs;

The inspection subunit is configured to perform inspection according to the data in the internal memory corresponding to the inspection address.

With reference to the second aspect, in a fourth possible implementation manner of the embodiment of the present invention, the device further includes:

A hard failure physical address acquisition module, configured to obtain a physical address of a data error whose failure type is a hard failure when the failure type of the memory is determined to be a hard failure;

The triggering module is configured to trigger an alarm, so as to prompt the user to replace the memory corresponding to the physical address of the data error whose failure type is hard failure.

With reference to the second aspect, in a fifth possible implementation manner of the embodiment of the present invention, the device further includes:

The data error physical address acquisition module is used to obtain the physical address where the data error occurs when a data error occurs in the memory;

A storage module, configured to store the physical address where the data error occurs in the memory;

A write-back module, configured to write back data to the physical address where a data error occurs.

In a third aspect, a device for repairing data errors is provided, and the device includes:

Memory, used to store data and the physical address where data errors occur;

a processor, configured to determine whether a preset counter in the memory overflows, and the preset counter is used to count data errors occurring in the memory;

The processor is further configured to, if the preset counter overflows, determine the failure type of the memory according to the physical address stored in the memory where a data error occurs, so as to perform subsequent corresponding repairs.

The beneficial effects brought by the technical solution provided by the embodiments of the present invention are:

The data error recovery method, device and equipment provided by the embodiments of the present invention determine whether the preset counter in the memory overflows, and the preset counter is used to count the occurrence of data errors in the memory; if the preset counter overflows, According to the physical address of the data error stored in the memory, the failure type of the memory is determined, so as to be repaired accordingly. By adopting the technical solution provided by the embodiment of the present invention, the failure type of the memory can be effectively distinguished and repaired according to the failure type, avoiding the system hanging or failing to start caused by the accumulation of data errors, and ensuring the normal operation of the business.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings that need to be used in the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those skilled in the art, other drawings can also be obtained based on these drawings without creative effort.

Fig. 1 is a flow chart of a method for repairing data errors provided in an embodiment of the present invention;

Fig. 2 is a flow chart of a method for repairing data errors provided in an embodiment of the present invention;

Fig. 3 is a schematic structural diagram of a data error repairing device provided in an embodiment of the present invention;

Fig. 4 is a schematic structural diagram of a data error repairing device provided in an embodiment of the present invention.

Detailed ways

In order to make the object, technical solution and advantages of the present invention clearer, the implementation manner of the present invention will be further described in detail below in conjunction with the accompanying drawings.

In the embodiment of the present invention, a terminal device refers to a device that provides a data processing function, voice and/or data connectivity to a user, including a wireless terminal or a wired terminal. The wireless terminal may be a handheld device with a wireless connection function, or other processing device connected to a wireless modem, and a mobile terminal that communicates with one or more core networks via a wireless access network. Wireless terminals may be, for example, mobile telephones (or "cellular" telephones) and computers with mobile terminals. As another example, the wireless terminal may also be a portable, pocket, hand-held, computer built-in or vehicle-mounted mobile device. For another example, the wireless terminal may be a mobile station (English: mobilestation), an access point (English: accesspoint), or user equipment (English: userequipment, UE for short), and the like.

Fig. 1 is a flow chart of a method for repairing data errors provided in an embodiment of the present invention. The execution subject of the embodiment of the present invention is a terminal device. Referring to Fig. 1, the method includes:

101: Determine whether a preset counter in the memory overflows, and the preset counter is used to store a physical address where a data error occurs in the memory;

Wherein, the preset counter is a space preset in the memory, and the size of the preset counter is set by technicians during the design process, which is not specifically limited in the embodiment of the present invention.

Preferably, the preset counter reads the ECC register at regular time intervals, and when the flag bit read in the ECC register indicates that there is an error in the data in the memory, the value of the preset counter is increased by 1. Further, the value of the preset counter is decremented by 1 every preset time length, and the preset time length is longer than the time interval for reading. When the value of the preset counter exceeds the overflow threshold, the preset counter overflows. Parameters such as the time interval for reading, the preset duration, and the overflow threshold involved in the preset counter can be set by technicians, which are not specifically limited in this embodiment of the present invention.

When the terminal device judges whether the preset counter in the memory overflows, it can trigger the corresponding instruction when the value of the preset counter exceeds the overflow threshold, and when it receives the corresponding instruction when the preset counter overflows, it determines that the preset counter overflows, otherwise , making sure that the preset counter has not overflowed.

Preferably, the data error is a single-bit error. When the preset counter overflows, it is necessary to judge the failure type of the data corresponding to the physical address of the single-bit data error stored in the memory and process it to prevent multi-bit data errors. occurs; when the preset counter does not overflow, the number of physical addresses where single-bit data errors occur in the memory is relatively small, and the data corresponding to the physical addresses stored in the memory may not be processed.

102: If the preset counter overflows, determine the failure type of the memory according to the physical address stored in the memory where a data error occurs, so as to perform corresponding repairs subsequently.

Among them, memory failure types are divided into soft failure and hard failure. The data error of soft failure can be written back, that is, the correct data is written to the physical address corresponding to the soft failure; the data error of hard failure cannot be written back, and the corresponding memory can only be replaced manually.

If the preset counter overflows, the terminal device needs to judge the failure type of the memory corresponding to the physical address stored in the memory, so as to determine the failure type of the memory. The terminal device can repeatedly read the data corresponding to the physical address stored in the memory to determine whether the data corresponding to the physical address stored in the memory has been repaired. If the data is repaired, the failure type of the memory corresponding to the physical address is soft Failure; if the data has not been repaired, the failure type of the memory corresponding to the physical address is a hard failure. For example, when the data corresponding to a physical address is read multiple times, and it is known through detection that the data corresponding to the physical address still has errors, the failure type of the memory corresponding to the physical address is a hard failure.

Preferably, judging whether the data corresponding to the physical address stored in the memory has been repaired can be known through detection by the ECC register.

When it is determined that the failure type corresponding to the physical address stored in the memory is a soft failure, then write back the correct data to the physical address corresponding to the soft failure; when it is determined that the failure type corresponding to the physical address stored in the memory is a hard failure, then The correct data cannot be written back to the physical address corresponding to the hard failure. Correspondingly, the user is prompted that the memory error is a hard failure, and the memory corresponding to the physical address needs to be replaced manually to prevent the accumulation of multiple bit errors from causing the system to hang. death and other issues.

In the method for repairing data errors provided by the embodiments of the present invention, by judging whether the preset counter in the memory overflows, the preset counter is used to count data errors in the memory; if the preset counter overflows, according to the memory The physical address where the data error occurs is stored in the memory, and the failure type of the memory is determined, so that subsequent corresponding repairs can be performed. By adopting the technical solution provided by the embodiment of the present invention, the failure type of the memory can be effectively distinguished and repaired according to the failure type, avoiding the system hanging or failing to start caused by the accumulation of data errors, and ensuring the normal operation of the business.

Fig. 2 is a flow chart of a method for repairing data errors provided in the embodiment of the present invention. The execution subject of the embodiment of the present invention is a terminal device, and the data error is a single-bit data error as an example for illustration. Referring to Figure 2, the method includes:

201: When a data error occurs in the memory, obtain the physical address where the data error occurs;

When a single-bit data error occurs in the memory, the terminal device obtains a physical address corresponding to the single-bit data error according to the single-bit data error detected by the ECC.

202: Store the physical address where the data error occurs in the memory, and write back data to the physical address where the data error occurs;

Specifically, after obtaining the physical address where the data error occurred, the terminal device stores the physical address where the data error occurred in the memory, and at the same time starts the Demand Scrubbing function to write back the physical address where the single-bit data error occurred. correct data to enable recovery of single-bit data errors.

When the failure type corresponding to the single-bit data error is soft failure, ideally, the demand clearing function can write back the correct data. When the failure type corresponding to the single-bit data error is hard failure, The demand clear function cannot write back correct data. Therefore, the terminal device needs to further determine the failure type of the memory with a single-bit data error. If it is a soft failure, the single-bit data error has been repaired; if it is a hard failure, the single-bit data error has not been repaired, and further steps are required deal with.

203: Determine whether the preset counter in the memory overflows, the preset counter is used to count data errors in the memory, if yes, perform step 204, if not, continue to perform step 203;

When the value of the preset counter is greater than the overflow threshold, the preset counter overflows.

It should be noted that, when the preset counter does not overflow, no subsequent steps are performed, and the terminal device continues to determine whether the preset counter overflows.

Steps 201-202 are optional steps. Storing the physical address where the data error occurs to the memory can be used as a trigger condition for step 203. When the physical address where the data error occurs is stored, step 203 is executed, and in the present invention In another embodiment of the present invention, step 203 may also be performed at preset time intervals, without taking the occurrence of a single bit error and storing the physical address where a data error occurs as a trigger condition.

204: Judging whether the same physical address exists in the physical address stored in the memory where the data error occurs, if yes, perform step 205, and if not, perform step 206;

If the preset counter overflows, the terminal device reads the physical addresses where data errors occur stored in the memory, and judges whether there is the same physical address among the physical addresses where data errors occur.

If it is determined that there are two or more identical physical addresses in the judgment process, it can be determined that there are identical physical addresses in the memory; if there are no two or more identical physical addresses, it can be determined that there are no identical physical addresses in the memory. physical address.

205: If the same physical address exists in the memory, determine that the failure type of the single-bit data error corresponding to the same physical address is a hard failure, and perform step 211;

According to step 202, for soft failure, after obtaining the physical address where the single-bit data error occurred, the physical address where the single-bit data error occurred is stored in the memory, and at the same time, the demand scrubbing (DemandScrubbing) function is started. The clear function writes the correct data at the physical address where a single-bit data error occurred. Therefore, if the failure type of the single-bit data error is a soft failure, correct data will be written in the physical address of the single-bit data error, and when the physical address is detected again, the If the data is correct, the physical address will not be written into the memory, that is, the physical address of the soft failure will only be stored in the memory once; and if the failure type of the single-bit data error is a hard failure, due to data write-back The correct data cannot be written into the physical address where the single-bit data error occurs, resulting in that the wrong data of the physical address cannot be repaired. When the physical address is detected again, the physical address will be written into the memory again. Therefore, when the type of failure where the single-bit data error occurs is a hard failure, the physical address where the single-bit data error occurs may be stored in the memory multiple times.

Since the physical address where a single-bit data error occurs may be stored in the memory multiple times, if there are two or more identical physical addresses in the memory, it is determined that the failure type of the single-bit data error corresponding to the same physical address is hard invalidated.

206: If the same physical address does not exist in the memory, convert the physical address in the memory where a data error occurs into an inspection address;

When the terminal device starts the PatrolScrubbing function, it performs patrol inspection on the memory data corresponding to the patrol address according to the converted patrol address. The terminal device converts the physical address in the memory where a data error occurs into an inspection address, so as to facilitate inspection based on the inspection address.

Specifically, converting the physical address where a data error occurs in the memory into a patrol address includes: the terminal device judges whether the physical address where a data error occurs in the memory is a memory address, and if it is determined that the physical address where a data error occurs in the memory If the address is a memory address, read the DRAM_RULE register to determine the socket where the memory is located; query the TAD0-TAD11 registers to determine the ChannelID; according to the RIRWAYNESS register and ririlvXoffset, you can determine the faulty DIMM, RankID and Rank internal address, according to the obtained socketID, ChannelID, DIMM , RankID, and Rank address to obtain the inspection address. The process for the terminal device to obtain the inspection address according to the physical address is well known to those skilled in the art, and will not be described in detail in this embodiment of the present invention.

If the same physical address does not exist in the memory, the type of single-bit data error corresponding to the physical address stored in the memory may be a soft failure or a hard failure.

207: Perform inspection according to the data in the memory corresponding to the inspection address;

Specifically, the terminal device stops the automatic inspection of the system to clear PatrolScrubbing, writes the converted inspection address into the SCRUBADDRESSLO register and the SCRUBADDRESSI register, enables inspection, and uses the SCRUBADDRESSLO register and the SCRUBADDRESSI register to match the memory address corresponding to the converted inspection address. Inspection of data in .

During the inspection process, if there is a data error in the memory data corresponding to the inspection address, write back the data in the memory corresponding to the inspection address; deal with.

Steps 206-207 are the process of performing memory inspection if the same physical address does not exist in the memory.

208: After the inspection is finished, determine whether the single-bit data error in the memory has been repaired, if yes, perform step 209, if not, perform step 210;

After the terminal device inspects the data corresponding to the inspection address in the memory, it reads the flag in the ECC register indicating whether there is a single-bit data error. If the flag in the ECC register indicates that the inspection address corresponds to the data in the memory If there is an error in the data in the memory, it means that the single-bit data error in the memory has not been repaired; if the flag bit in the ECC register indicates that the data in the memory corresponding to the inspection address has no error, it means that the single-bit data error in the memory has been repaired.

209: If the single-bit data error has been repaired, determine that the failure type of the single-bit data error is a soft failure, and end;

If it is determined that the single-bit data error in the memory has been repaired according to the identification bit in the ECC register, it means that the single-bit data error has been corrected during the start-up of the demand scrubbing (DemandScrubbing) function of the terminal device, and the invalidation of the single-bit data error is determined. Type is soft failure.

210: If the single-bit data error has not been repaired, determine that the failure type of the single-bit data error is a hard failure;

If it is determined that the single-bit data error in the memory has not been repaired according to the identification bit in the ECC register, it means that the single-bit data error has not been corrected during the terminal device startup demand scrubbing (DemandScrubbing) function process, and it is determined that the single-bit data error The failure type is hard failure.

211: When it is determined that the failure type of the memory is a hard failure, acquire a physical address of a single-bit data error whose failure type is a hard failure;

When it is determined that the failure type of the memory is a hard failure, the process for the terminal device to obtain the physical address of the single-bit data error whose failure type is a hard failure may be any of the following:

(1) The terminal device detects the same physical address in the memory. When the same physical address exists and determines that the failure type of the memory is a hard failure, the terminal device directly reads the same physical address;

(2) The terminal device performs inspection on the data corresponding to the inspection address. After the inspection, according to the flag in the ECC register, when the failure type of the memory is determined to be a hard failure, the terminal device retrieves the data from the mcelog file of the operating system OS. Obtain the physical address of the data corresponding to the hard failure.

212: An alarm is triggered, so as to prompt the user to replace the memory corresponding to the physical address of the single-bit data error whose failure type is a hard failure.

Preferably, after the terminal device obtains the physical address of the single-bit data error whose failure type is hard failure, it displays the physical address of the single-bit data error whose failure type is hard failure on the display screen, and triggers an alarm, so that the user After learning this information, replace the memory corresponding to the physical address of the single-bit data error whose failure type is hard failure, so as to avoid the accumulation of single-bit data error of hard failure, causing the system to hang up, and preventing the memory problem from being reset or upgraded on the single board exploded in large numbers.

Steps 204-212 are a process of determining the failure type of the memory according to the physical address of the memory where the data error occurs if the preset counter overflows, so as to perform subsequent corresponding repairs.

It should be noted that the execution subject of this embodiment of the present invention may also be a memory controller in the terminal device.

In the method for repairing data errors provided by the embodiments of the present invention, by judging whether the preset counter in the memory overflows, the preset counter is used to count data errors in the memory; if the preset counter overflows, according to the memory The physical address where the data error occurs is stored in the memory, and the failure type of the memory is determined, so that subsequent repairs can be carried out accordingly. By adopting the technical solution provided by the embodiment of the present invention, the failure type of the memory can be effectively distinguished, and the failure type can be repaired according to the failure type to avoid It prevents the system from hanging or failing to start caused by the accumulation of data errors, and ensures the normal operation of the business.

Fig. 3 is a schematic structural diagram of a data error repairing device provided in an embodiment of the present invention. Referring to Fig. 3, the device includes:

A judging module 301, configured to judge whether a preset counter in the memory overflows, and the preset counter is used to count data errors in the memory;

The determining module 302 is configured to, if the preset counter overflows, determine the failure type of the memory according to the physical address stored in the memory where a data error occurs, so as to perform corresponding repairs subsequently.

The determination module 302 is configured to determine that the failure type of the data error corresponding to the same physical address is a hard failure if the same physical address exists in the physical addresses stored in the memory where the data error occurs.

The determining module 302 includes:

The inspection unit is used to perform memory inspection if the same physical address does not exist in the physical address where the data error occurs stored in the memory;

a judging unit, configured to judge whether the data error in the memory has been repaired after the inspection is finished;

a determining unit, configured to determine that the failure type of the data error is a hard failure if the data error has not been repaired;

The determining unit is configured to determine that the failure type of the data error is a soft failure if the data error has been repaired.

The inspection unit includes:

The patrol address conversion subunit is used to convert the physical address in the preset counter into a patrol address if the same physical address does not exist in the physical address stored in the memory where a data error occurs;

The inspection subunit is configured to perform inspection according to the data in the internal memory corresponding to the inspection address.

The device also includes:

A hard failure physical address acquisition module 303, configured to obtain a physical address of a data error whose failure type is a hard failure when the failure type of the memory is determined to be a hard failure;

The trigger module 304 is configured to trigger an alarm, so as to prompt the user to replace the memory corresponding to the physical address of the data error whose failure type is hard failure.

The device also includes:

The data error physical address acquisition module 305 is used to obtain the physical address where the data error occurs when a data error occurs in the memory;

A storage module 306, configured to store the physical address where the data error occurs in the memory;

A write-back module 307, configured to write back data to the physical address where a data error occurs.

The data error recovery device provided by the embodiment of the present invention judges whether the preset counter in the memory overflows, and the preset counter is used to count data errors in the memory; if the preset counter overflows, according to the memory The physical address where the data error occurs is stored in the memory, and the failure type of the memory is determined, so that subsequent corresponding repairs can be performed. By adopting the technical solution provided by the embodiment of the present invention, the failure type of the memory can be effectively distinguished and repaired according to the failure type, avoiding the system hanging or failing to start caused by the accumulation of data errors, and ensuring the normal operation of the business.

It should be noted that: the data error repairing device provided by the above embodiment only uses the division of the above-mentioned functional modules as an example for the data error repairing. In practical applications, the above-mentioned function allocation can be completed by different functional modules according to needs. , which divides the internal structure of the device into different functional modules to complete all or part of the functions described above. In addition, the device for repairing data errors provided by the above embodiments belongs to the same idea as the embodiments of the method for repairing data errors, and its specific implementation process is detailed in the method embodiments, and will not be repeated here.

Fig. 4 is a schematic structural diagram of a data error repairing device provided in an embodiment of the present invention. Referring to FIG. 4, the data error repairing device includes: a processor 401 and a memory 402,

The processor 401 is configured to determine whether a preset counter in the memory overflows, and the preset counter is used to count data errors occurring in the memory 402;

The processor 401 is configured to determine the failure type of the memory 402 according to the physical address stored in the memory where a data error occurs if the preset counter overflows, so as to perform corresponding repairs subsequently.

The memory 402 is used for storing data and physical addresses where data errors occur.

The processor 401 is configured to determine that the failure type of the data error corresponding to the same physical address is a hard failure if the same physical address exists in the physical addresses stored in the memory 402 where the data error occurs.

The processor 401 is configured to perform memory inspection if the same physical address does not exist in the physical address where the data error occurs stored in the memory 402;

The processor 401 is configured to determine whether the data error in the memory 402 has been repaired after the inspection is completed;

Processor 401, configured to determine that the failure type of the data error is a hard failure if the data error has not been repaired;

The processor 401 is configured to determine that the failure type of the data error is a soft failure if the data error has been repaired.

The processor 401 is configured to convert the physical address in the preset counter into a patrol address if the same physical address does not exist in the physical address in which the data error occurs stored in the memory 402;

The processor 401 is configured to perform inspection according to the data in the memory 402 corresponding to the inspection address.

The processor 401 is configured to, when determining that the failure type of the memory 402 is a hard failure, obtain a physical address of a data error whose failure type is a hard failure;

The processor 401 is configured to trigger an alarm, so as to prompt the user to replace the memory 402 corresponding to the physical address of the data error whose failure type is hard failure.

The processor 401 is configured to obtain a physical address where a data error occurs when a data error occurs in the memory 402;

The processor 401 is configured to store the physical address where the data error occurs in the memory 402, and perform data write-back on the physical address where the data error occurs.

The data error recovery device provided by the embodiment of the present invention judges whether the preset counter in the memory overflows, and the preset counter is used to count data errors in the memory; if the preset counter overflows, according to the memory The physical address where the data error occurs is stored in the memory, and the failure type of the memory is determined, so that subsequent corresponding repairs can be performed. By adopting the technical solution provided by the embodiment of the present invention, the failure type of the memory can be effectively distinguished and repaired according to the failure type, avoiding the system hanging or failing to start caused by the accumulation of data errors, and ensuring the normal operation of the business.

Those of ordinary skill in the art can understand that all or part of the steps for implementing the above embodiments can be completed by hardware, and can also be completed by instructing related hardware through a program. The program can be stored in a computer-readable storage medium. The above-mentioned The storage medium mentioned may be a read-only memory, a magnetic disk or an optical disk, and the like.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection of the present invention. within range.

Claims (9)

1. A data error recovery method, characterized in that said method comprises: judging whether a preset counter in the memory overflows, and the preset counter is used to count data errors in the memory; If the preset counter overflows, determine the failure type of the memory according to the physical address stored in the memory where a data error occurs, so as to perform corresponding repairs subsequently; If the preset counter overflows, the failure type of the memory is determined according to the physical address of the data error stored in the memory, so as to be repaired accordingly, including: If the same physical address does not exist in the physical address where the data error occurs stored in the memory, converting the physical address in the preset counter into a patrol address; performing inspection according to the data in the internal memory corresponding to the inspection address; After the inspection is finished, it is judged whether the data error in the memory has been repaired; If the data error has not been repaired, determining that the failure type of the data error is a hard failure; If the data error has been repaired, it is determined that the failure type of the data error is a soft failure. 2. The method according to claim 1, wherein if the preset counter overflows, the failure type of the memory is determined according to the physical address of the data error stored in the memory, so that subsequent corresponding Fixes, including: If the same physical address exists in the physical addresses stored in the internal memory where the data error occurs, it is determined that the failure type of the data error corresponding to the same physical address is a hard failure. 3. The method according to claim 1, wherein if the preset counter overflows, the failure type of the memory is determined according to the physical address of the data error stored in the memory, so that subsequent corresponding After repairing, the method also includes: When it is determined that the failure type of the memory is a hard failure, acquiring the physical address of the data error whose failure type is a hard failure; An alarm is triggered so as to prompt the user to replace the memory corresponding to the physical address of the data error whose failure type is hard failure. 4. The method according to claim 1, wherein before judging whether the preset counter in the internal memory overflows, the method also includes: When a data error occurs in the memory, obtain the physical address where the data error occurs; The physical address where the data error occurs is stored in the memory, and the data is written back to the physical address where the data error occurs. 5. A data error recovery device, characterized in that the device comprises: A judging module, configured to judge whether a preset counter in the memory overflows, and the preset counter is used to count data errors in the memory; A determining module, configured to determine the type of failure of the memory according to the physical address of the memory where the data error occurs if the preset counter overflows, so as to perform subsequent corresponding repairs; The determination module includes: The inspection unit is used to perform memory inspection if the same physical address does not exist in the physical address where the data error occurs stored in the memory; a judging unit, configured to judge whether the data error in the memory has been repaired after the inspection is finished; a determining unit, configured to determine that the failure type of the data error is a hard failure if the data error has not been repaired; The determining unit is configured to determine that the failure type of the data error is a soft failure if the data error has been repaired; The inspection unit includes: The patrol address conversion subunit is used to convert the physical address in the preset counter into a patrol address if the same physical address does not exist in the physical address stored in the memory where a data error occurs; The inspection subunit is configured to perform inspection according to the data in the internal memory corresponding to the inspection address. 6. The device according to claim 5, wherein the determining module is configured to determine the corresponding physical address of the same physical address if there is the same physical address in the physical address stored in the memory where a data error occurs. The failure type of a data error is a hard failure. 7. The device according to claim 5, further comprising: A hard failure physical address acquisition module, configured to obtain a physical address of a data error whose failure type is a hard failure when the failure type of the memory is determined to be a hard failure; The triggering module is configured to trigger an alarm, so as to prompt the user to replace the memory corresponding to the physical address of the data error whose failure type is hard failure. 8. The device according to claim 5, further comprising: The data error physical address acquisition module is used to obtain the physical address where the data error occurs when a data error occurs in the memory; A storage module, configured to store the physical address where the data error occurs in the memory; A write-back module, configured to write back data to the physical address where a data error occurs. 9. A data error recovery device, characterized in that the device comprises: Memory, used to store data and the physical address where data errors occur; a processor, configured to determine whether a preset counter in the memory overflows, and the preset counter is used to count data errors occurring in the memory; The processor is further configured to, if the preset counter overflows, determine the failure type of the memory according to the physical address stored in the memory where a data error occurs, so as to perform subsequent corresponding repairs; The processor is further configured to convert the physical address in the preset counter into a patrol address if the same physical address does not exist in the physical address stored in the memory where a data error occurs; according to the patrol The data in the memory corresponding to the address is inspected; after the inspection is completed, it is judged whether the data error in the memory has been repaired; if the data error has not been repaired, it is determined that the failure type of the data error is Hard failure; if the data error has been repaired, determine that the failure type of the data error is a soft failure.