Fault Detection Method, Device, Equipment, Storage Medium and Program Product for Graphics Processing Unit

Abstract

The present application relates to a method, device, equipment, storage medium and program product for fault detection of a graphics processor. The method includes, when it is detected that a target image output by an image output unit in the graphics processor is abnormal, obtaining a rasterization file output by a rasterization unit in the graphics processor, then generating a detection image according to the rasterization file, and performing fault detection on the rasterization unit according to the detection image to obtain a first detection result, and then detecting other units in the graphics processor according to the first detection result to obtain a target detection result. In the above method, since the rasterization file output by the rasterization unit already has a preliminary form of an image, a detection image can be generated from the rasterization file, and the detection image can be used to perform fault detection on the rasterization unit, so as to quickly determine whether the rasterization unit fails, thereby improving the efficiency of fault detection.

Description

Technical Field

[0001] The present application relates to the technical field of graphics processing units, and particularly to a method, device, equipment, storage medium, and program product for fault detection of a graphics processing unit. Background Art

[0002] With the rapid development of computer technology, rendering technologies based on Graphics Processing Units (GPUs) have been widely applied in multiple fields, including 3D animation production, video game development, and autonomous driving systems. As the core process for realizing the rendering of 3D models, the Graphic Pipeline has a complex structure and consists of multiple closely connected processing stages. Each stage undertakes specific graphic operations, such as vertex processing, rasterization, texture mapping, and pixel processing, and finally presents the 3D model as a 2D image that we can perceive. To ensure the accuracy and reliability of the graphics processor rendering technology, when the image rendered by the Graphic Pipeline is abnormal, it is very necessary to accurately perform fault detection.

[0003] Currently, when the image rendered by the Graphic Pipeline is abnormal, the signal waveforms corresponding to each processing stage of the Graphic Pipeline are usually checked one by one to perform fault detection. However, the above-mentioned fault detection method has the problem of low efficiency. Summary of the Invention

[0004] Based on this, it is necessary to provide a method, device, equipment, storage medium, and program product for fault detection of a graphics processing unit that can improve the fault detection efficiency of the graphics processing unit in view of the above technical problems.

[0005] In a first aspect, the present application provides a method for fault detection of a graphics processing unit, and the method includes:

[0006] When it is detected that the target image output by the image output unit in the graphics processing unit is abnormal, obtain the rasterization file output by the rasterization unit in the graphics processing unit;

[0007] Generate a detection image according to the rasterization file, and perform fault detection on the rasterization unit according to the detection image to obtain a first detection result;

[0008] Detect other units in the graphics processing unit according to the first detection result to obtain a target detection result.

[0009] In one of the embodiments, detecting other units in the graphics processing unit according to the first detection result to obtain a target detection result includes:

[0010] If the first detection result indicates that the rasterization unit is in a normal working state, it is determined that the target detection result indicates a fault in the image output unit in the graphics processing unit;

[0011] If the first detection result indicates that the rasterization unit is in a potentially abnormal state, fault detection is performed on the rasterization unit according to the signal waveform output by the rasterization unit to obtain a second detection result, and other units are detected according to the second detection result to obtain a target detection result.

[0012] In one embodiment, detecting other units according to the second detection result to obtain a target detection result includes:

[0013] If the second detection result indicates that the rasterization unit is in a normal working state, other units are detected according to the signal waveform output by other units to obtain a target detection result;

[0014] If the second detection result indicates that the rasterization unit is in an abnormal working state, it is determined that the target detection result indicates a fault in the rasterization unit in the graphics processing unit.

[0015] In one embodiment, other units include a tessellation processing unit, an input loading unit, and an image output unit. Detecting other units according to the signal waveform output by other units to obtain a target detection result includes:

[0016] Detecting the tessellation processing unit according to the signal waveform output by the tessellation processing unit to obtain a third detection result;

[0017] When the third detection result indicates that the tessellation processing unit is in a normal working state, the input loading unit is detected according to the signal waveform output by the input loading unit to obtain a fourth detection result;

[0018] If the fourth detection result indicates that the input loading unit is in a normal working state, it is determined that the target detection result indicates a fault in the image output unit.

[0019] In one embodiment, obtaining the rasterization file output by the rasterization unit in the graphics processing unit includes:

[0020] Obtaining the coordinate information and attribute information output by each rasterization subunit in the rasterization unit;

[0021] Generating a rasterization file according to the coordinate information and attribute information output by each rasterization subunit.

[0022] In one embodiment, performing fault detection on the rasterization unit according to the detection image to obtain a first detection result includes:

[0023] Determining whether the detection image is consistent with the standard image;

[0024] If the detected image is the same as the standard image, it is determined that the first detection result indicates that the rasterization unit is in a normal working state;

[0025] If the detected image is different from the standard image, it is determined that the first detection result indicates that the rasterization unit is in a potentially abnormal state.

[0026] In a second aspect, the present application further provides a fault detection device for a graphics processor, and the device includes:

[0027] An acquisition module, configured to acquire a rasterization file output by a rasterization unit in the graphics processor when it is detected that a target image output by an image output unit in the graphics processor is abnormal;

[0028] A first detection module, configured to generate a detection image according to the rasterization file, and perform fault detection on the rasterization unit according to the detection image to obtain a first detection result;

[0029] A second detection module, configured to detect other units in the graphics processor according to the first detection result to obtain a target detection result.

[0030] In a third aspect, the present application further provides a computer device, and the computer device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0031] When it is detected that a target image output by an image output unit in the graphics processor is abnormal, acquire a rasterization file output by a rasterization unit in the graphics processor;

[0032] Generate a detection image according to the rasterization file, and perform fault detection on the rasterization unit according to the detection image to obtain a first detection result;

[0033] Detect other units in the graphics processor according to the first detection result to obtain a target detection result.

[0034] In a fourth aspect, the present application further provides a computer-readable storage medium, and the computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the following steps are implemented:

[0035] When it is detected that a target image output by an image output unit in the graphics processor is abnormal, acquire a rasterization file output by a rasterization unit in the graphics processor;

[0036] Generate a detection image according to the rasterization file, and perform fault detection on the rasterization unit according to the detection image to obtain a first detection result;

[0037] Detect other units in the graphics processor according to the first detection result to obtain a target detection result.

[0038] In a fifth aspect, the present application also provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the following steps are implemented:

[0039] When it is detected that the target image output by the image output unit in the graphics processor is abnormal, obtain the rasterization file output by the rasterization unit in the graphics processor;

[0040] Generate a detection image according to the rasterization file, and perform a fault detection on the rasterization unit according to the detection image to obtain a first detection result;

[0041] Detect other units in the graphics processor according to the first detection result to obtain a target detection result.

[0042] In the above-mentioned fault detection method, device, equipment, storage medium and program product of the graphics processor, when it is detected that the target image output by the image output unit in the graphics processor is abnormal, the method obtains the rasterization file output by the rasterization unit in the graphics processor, then generates a detection image according to the rasterization file, and performs a fault detection on the rasterization unit according to the detection image to obtain a first detection result, and then detects other units in the graphics processor according to the first detection result to obtain a target detection result. In the above method, since the rasterization file output by the rasterization unit already has the preliminary form of an image, a detection image can be generated from the rasterization file, and the detection image can be used to perform a fault detection on the rasterization unit, so as to quickly determine whether the rasterization unit has a fault, thereby improving the efficiency of fault detection and solving the problem of low efficiency caused by relying on a large number of signal waveforms for fault detection in the traditional technology. In addition, this method uses the rasterization unit as a demarcation point, and by judging the fault situation of the rasterization unit, it can quickly detect the fault of the image output unit. Compared with the method of performing fault detection step by step in the traditional technology, this method can significantly improve the efficiency of fault detection. Description of the Drawings

[0043] Figure 1 It is a schematic diagram of the graphics pipeline of the graphics processor in the prior art;

[0044] Figure 2 It is the internal structure diagram of a computer device in an embodiment;

[0045] Figure 3 It is a schematic flowchart of the fault detection method of the graphics processor in another embodiment;

[0046] Figure 4Schematic flowchart of a fault detection method for a graphics processor in another embodiment;

[0047] Figure 5 Schematic flowchart of a fault detection method for a graphics processor in another embodiment;

[0048] Figure 6 Schematic flowchart of a fault detection method for a graphics processor in another embodiment;

[0049] Figure 7 Schematic flowchart of a fault detection method for a graphics processor in another embodiment;

[0050] Figure 8 Block diagram of the rasterization unit in a graphics processor in one embodiment;

[0051] Figure 9 Schematic flowchart of a fault detection method for a graphics processor in another embodiment;

[0052] Figure 10 Schematic flowchart of a fault detection method for a graphics processor in another embodiment;

[0053] Figure 11 Block diagram of a fault detection device for a graphics processor in one embodiment. Detailed implementation manners

[0054] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application, and are not used to limit the present application.

[0055] With the rapid development of computer technology, rendering technologies based on Graphics Processing Unit (GPU) have been widely used in many fields, including 3D animation production, video game development, and autonomous driving systems. As the core process for realizing the rendering of 3D models, the Graphic Pipeline has a complex structure and consists of multiple closely connected processing stages. Each stage undertakes specific graphic operations, such as vertex processing, rasterization, texture mapping, and pixel processing, and finally presents the 3D model as a 2D image that we can perceive. As Figure 1As shown in the figure, the processing stages of the graphics pipeline include the IA (Input Assembler), VS (Vertex Shader), HS (Hull Shader), TS (Tessellation Shader), DS (Domain Shader), GS (Geometry Shader), RS (Rasterizer), PS (Pixel Shader), and OM (Output Merger) stages. Among them, the IA is used to assemble the input source data (Resources, such as vertex data) into basic graphic elements; the VS is used to transform and process each vertex, such as the model-view-projection transformation; the HS is used to define the outer contour of the graphic and generate more complex geometric shapes; the TS is used to refine the details of the geometric shape and generate more vertices; the DS is used to process the refined vertex data for further processing and calculation; the GS is used to process the basic graphic elements to generate more vertices or discard some vertices; the RS is used to convert vertices into fragments (pixels) and perform clipping; the PS is used to process the color and texture of each fragment; and the OM is used to merge all the processed data to generate the final frame buffer content. To ensure the accuracy and reliability of the graphics processor rendering technology, it is very necessary to accurately perform fault detection when an abnormal image is rendered by the graphics pipeline. Currently, when an abnormal image is rendered by the graphics pipeline, the signal waveforms corresponding to each processing stage of the graphics pipeline are usually checked step by step for fault detection. However, the above fault detection method has the problem of low efficiency.

[0056] This application provides a fault detection method for a graphics processor, aiming to solve the above technical problems. The following embodiments will specifically illustrate the fault detection method for the graphics processor described in this application.

[0057] The fault detection method for the graphics processor provided in the embodiments of this application can be applied to a computer device as shown in Figure 2 the figure. This computer device can be a terminal or a server, and its internal structure diagram can be as shown in Figure 2As shown in the figure, the computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit, and an input device. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface, the display unit, and the input device are connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with external terminals in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a mobile cellular network, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it realizes a method for detecting faults of a graphics processor. The display unit of the computer device is used to form a visually visible picture, which can be a display screen, a projection device, or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer covering the display screen, or a button, a trackball, or a touchpad provided on the housing of the computer device, or an external keyboard, touchpad, or mouse, etc.

[0058] Those skilled in the art can understand that Figure 2 the structure shown in the figure is only a block diagram of some structures related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements.

[0059] In one embodiment, as Figure 3 shown, a method for detecting faults of a graphics processor is provided. Taking the method applied to the Figure 1 computer device in the figure as an example for description, the method includes the following steps:

[0060] S201, when it is detected that the target image output by the image output unit in the graphics processor is abnormal, obtain the rasterization file output by the rasterization unit in the graphics processor.

[0061] Among them, the graphics pipeline includes IA, VS, HS, TS, DS, GS, RS, PS, and OM. Among them, IA, TS, RS, and OM belong to hardware functional units, and VS, HS, DS, GS, and PS belong to software functional units. Usually, when the image output by the graphics processor is abnormal, it is generally caused by a failure of the hardware functional unit. Therefore, in the embodiments of the present application, it is defaulted that the software functional units are normal, and a method for fault detection of the hardware functional unit is proposed. The working process of the image output unit corresponds to Figure 1 the OM stage in Figure 1 Among them, the rasterization unit corresponds to the RS stage in

[0062] In the embodiments of the present application, the graphics processor is pre-set in a computer device, and a variety of functional units are designed in the graphics processor. The working condition of the graphics processor can be detected by the computer device. The computer device can pre-obtain a desired image. The specific obtaining method can be generated through a C language model, or can be generated through other algorithm models, or receive a ready-made image input by the user.

[0063] After the computer device receives the target image output by the image output unit in the graphics processor, it can compare the target image with the desired image to determine whether the target image and the desired image are consistent. If the target image and the desired image are consistent, it is determined that the target image is normal. If the target image and the desired image are inconsistent, it is determined that the target image is abnormal. Optionally, if the computer device does not obtain the target image within a preset time period, it is determined that the target image is abnormal. When the computer device detects that the target image is abnormal, it can first determine whether the previous-level hardware functional unit connected to the image output unit has failed, and then perform fault detection on the graphics processor according to the fault condition of the previous-level hardware functional unit. Specifically, after the computer device determines that the previous-level hardware functional unit is the rasterization unit, it can further obtain the rasterization file generated by the rasterization unit in the graphics pipeline stage.

[0064] Optionally, a user interface is provided on the computer device. The user interface includes a result display window, a fault detection button, a fault detection window, etc. Among them, the result display window is used to display the target image and display an abnormality prompt message when the target image is abnormal; the fault detection window includes all functional unit components, which can be specifically a functional unit list or a functional unit box. The processing records of all functional units are associated with the corresponding functional unit list or functional unit box. Among them, the processing records of IA and TS include a large number of primitive signals, the processing records of RS include a large number of pixel signals and rasterization files, and OM includes a large number of pixel signals and images; when the computer device detects that the target image is abnormal, it can display an abnormality prompt message on the user interface. After the user receives the abnormality alarm prompt, the user can click the fault detection button. After the computer device receives the fault detection button triggered by the user, all functional unit components can be popped up, and the user can select the functional unit component to be detected. After the computer device receives the user's selection instruction, it can determine the target functional unit component according to the selection instruction, and then obtain the processing record of the target functional unit. Optionally, the computer can pre-determine the target functional unit component. When the computer device detects that the target image is abnormal, it can display an abnormality prompt message on the user interface. After the user receives the abnormality alarm prompt, the user can click the fault detection button. After the computer device receives the fault detection button triggered by the user, it can directly obtain the processing record of the target functional unit component. For example, if the target functional unit component is a rasterization unit, after the computer device receives the fault detection button triggered by the user, it can directly obtain the rasterization file.

[0065] S202, generate a detection image according to the rasterization file, and perform a fault detection on the rasterization unit according to the detection image to obtain a first detection result.

[0066] Among them, the detection image is an image generated by a method different from the graphics pipeline, and is used for fault detection of the rasterization unit, and can be specifically a bmp picture. The first detection result includes that the rasterization unit is in a normal working state or the rasterization unit is in a potentially abnormal state.

[0067] In the embodiment of the present application, an image generation script can be pre-designed, and the image generation logic of the preset script is different from the image generation logic in the graphics pipeline. After the computer device obtains the rasterization file based on the above steps, it can call the preset script, and then execute the preset script to generate a detection image according to the rasterization file. After the computer device generates the detection image, it can determine whether the detection image meets the preset requirements. If the detection image meets the preset requirements, the first detection result is determined to be that the rasterization unit is in a normal working state. If the detection image does not meet the preset requirements, the first detection result is determined to be that the rasterization unit is in a potentially abnormal state.

[0068] S203. Detect other units in the graphics processor according to the first detection result to obtain a target detection result.

[0069] Among them, the above other units include IA, TS, and OM. The target detection result includes one of the states that IA is in a normal or abnormal state, TS is in a normal or abnormal state, RS is in a normal or abnormal state, and OM is in a normal or abnormal state.

[0070] In the embodiment of the present application, after the computer device obtains the first detection result based on the above steps, it can first determine the working state of the rasterization unit according to the first detection result, then determine the working state of the image output unit, and further determine the working states of other functional units. Optionally, the detection of other units in the graphics processor can also be implemented through a human-computer interaction method. Specifically, the computer can preset the detection order of the functional unit components in advance. After the computer device receives the user-triggered fault detection button, it directly performs fault detection according to the detection order using the above detection method and displays the target detection result. Optionally, after the computer device receives the user-triggered fault detection button, it can pop up all the functional unit components, and the user can select the functional unit components to be detected, then perform the detection of the functional unit, and display the detection result, and the user can repeat the above steps to detect all the functional units. Optionally, when the computer device detects that there is an abnormality in the target image, it can display an abnormality prompt message on the user interface. After the user receives the abnormal alarm prompt, the user can click the fault detection button, and the computer device can execute the steps of S201-S203 above with one key to obtain and display the target detection result on the user interface.

[0071] The fault detection method for a graphics processor provided by an embodiment of the present application, in the case of detecting that a target image output by an image output unit in the graphics processor is abnormal, obtains a rasterization file output by a rasterization unit in the graphics processor, then generates a detection image according to the rasterization file, and performs fault detection on the rasterization unit according to the detection image to obtain a first detection result, and then detects other units in the graphics processor according to the first detection result to obtain a target detection result. In the above method, since the rasterization file output by the rasterization unit already has a preliminary form of an image (converted from primitive signals to pixel signals), a detection image can be generated from the rasterization file, and the detection image can be used to perform fault detection on the rasterization unit, so as to quickly determine whether the rasterization unit has a fault, thereby improving the efficiency of fault detection, and solving the problem of low efficiency caused by relying on a large number of signal waveforms for fault detection in the traditional technology. In addition, this method uses the rasterization unit as a demarcation point, and by judging the fault situation of the rasterization unit, it can quickly detect the fault of the image output unit. Compared with the method of performing fault detection step by step in the traditional technology, this method can significantly improve the efficiency of fault detection.

[0072] In one embodiment, a specific implementation manner of detecting other units according to the second detection result is further provided, as Figure 4 shown, the "detecting other units according to the second detection result to obtain a target detection result" in the above step S203 includes:

[0073] S301, if the first detection result indicates that the rasterization unit is in a normal working state, it is determined that the target detection result indicates that there is a fault in the image output unit in the graphics processor.

[0074] In the embodiment of the present application, if the first detection result indicates that the rasterization unit is in a normal working state, this situation indicates that the rasterization unit and the functional units before it are both normal, that is, IA and TS are not faulty, then it can be determined that there is a fault in the image output unit after the rasterization unit, that is, OM is faulty.

[0075] S302, if the first detection result indicates that the rasterization unit is in a potentially abnormal state, perform fault detection on the rasterization unit according to the signal waveform output by the rasterization unit to obtain a second detection result, and detect other units according to the second detection result to obtain a target detection result.

[0076] Among them, the second detection result includes that the rasterization unit is in a normal working state or the rasterization unit is in an abnormal working state. The signal waveform is used to reflect the working state of its own unit, and the signal waveform output by the rasterization unit is generated by a large number of pixel signals in the RS processing record.

[0077] In the embodiments of the present application, if the first detection result indicates that the rasterization unit is in a potentially abnormal state, this situation indicates that all functional units in the graphics pipeline may be abnormal, that is, IA, TS, RS, or OM may malfunction. Therefore, the rasterization unit may be normal or abnormal. Therefore, it is necessary to further perform fault detection based on the signal waveform. Specifically, after the computer device determines that the first detection result indicates that the rasterization unit is in a potentially abnormal state, it can obtain the processing record of the rasterization unit. The obtaining method can be executed by the computer in the background as described above, or can be implemented by means of user interaction through the user interface. After the computer device obtains the processing record of the rasterization unit, it can directly extract the signal waveform in the processing record, and then compare the signal waveform with the expected waveform to determine the detection result. If the signal waveform is consistent with the expected waveform, it is determined that the rasterization unit is in a normal working state. If the signal waveform is inconsistent with the expected waveform, or the signal waveform cannot be generated, it is determined that the rasterization unit is in an abnormal working state. After the computer device obtains the second detection result, it can detect other units according to the second detection result to obtain the target detection result. Optionally, after the computer device obtains the processing record of the rasterization unit, it can extract the input signal and output signal of the rasterization unit, and then generate a signal waveform according to the input signal and output signal, and determine the second detection result according to the signal waveform.

[0078] Specifically, as Figure 5 shown, the "detecting other units according to the second detection result to obtain the target detection result" in step S302 above includes:

[0079] S401, if the second detection result indicates that the rasterization unit is in a normal working state, then detect other units according to the signal waveforms output by other units to obtain the target detection result.

[0080] Among them, the signal waveforms output by other units include the signal waveform output by IA, the signal waveform output by TS, and the signal waveform output by OM. The signal waveform output by IA is generated from a large number of primitive signals in the IA processing record, the signal waveform output by TS is generated from a large number of primitive signals in the TS processing record, and the signal waveform output by OM is generated from a large number of pixel signals in the OM processing record.

[0081] In the embodiment of the present application, if the second detection result indicates that the rasterization unit is in a normal working state, this situation indicates that the rasterization unit has not failed. Therefore, the unit that has failed may be other units, that is, IA, TS, or OM may be faulty. Therefore, it is necessary to detect other units according to the signal waveforms output by other units to obtain the target detection result. When detecting other units, the order of the three, IA, TS, or OM, is not limited. This order can be TS-IA-OM, or OM-TS-IA, or other orders.

[0082] Taking the detection order of TS-IA-OM as an example, the fault detection is carried out as follows Figure 6 As shown, the "detecting other units according to the signal waveforms output by other units to obtain the target detection result" in the above step S301 includes:

[0083] S4011, detecting the subdivision processing unit according to the signal waveform output by the subdivision processing unit to obtain the third detection result.

[0084] Among them, the subdivision processing unit is TS.

[0085] In the embodiment of the present application, after the computer device determines that the second detection result indicates that the rasterization unit is in a normal working state, it can obtain the processing record of the subdivision processing unit. The obtaining method can be executed by the computer in the background as described above, or can be implemented by means of user interaction through the user interface. After the computer device obtains the processing record of the subdivision processing unit, it can directly extract the signal waveform in the processing record, and then compare the signal waveform with the expected waveform to determine the detection result. If the signal waveform and the expected waveform are consistent, it is determined that the subdivision processing unit is in a normal working state. If the signal waveform and the expected waveform are inconsistent, or the signal waveform cannot be generated, it is determined that the subdivision processing unit is in an abnormal working state. After the computer device obtains the third detection result, it can detect other units according to the third detection result to obtain the target detection result. Optionally, after the computer device obtains the processing record of the subdivision processing unit, it can extract the input signal and output signal of the subdivision processing unit, and then generate a signal waveform according to the input signal and output signal, and determine the third detection result according to the signal waveform. If the third detection result indicates that the subdivision processing unit is in an abnormal working state, it is determined that the subdivision processing unit has failed.

[0086] S4012, in the case where the third detection result indicates that the subdivision processing unit is in a normal working state, detecting the input loading unit according to the signal waveform output by the input loading unit to obtain the fourth detection result.

[0087] Among them, the input loading unit is IA.

[0088] In the embodiments of the present application, if the third detection result indicates that the segmentation processing unit is in a normal working state, this situation indicates that the TS is not faulty. Then, the faulty unit may be the IA or the OM. Therefore, it is necessary to detect the input loading unit according to the signal waveform output by the input loading unit. Specifically, the computer device can obtain the processing record of the input loading unit. The obtaining method can be implemented by the computer executing in the background as described above, or by means of user interaction through the user interface. After the computer device obtains the processing record of the input loading unit, it can directly extract the signal waveform in the processing record, and then compare the signal waveform with the expected waveform to determine the detection result. If the signal waveform is consistent with the expected waveform, it is determined that the input loading unit is in a normal working state. If the signal waveform is inconsistent with the expected waveform, or the signal waveform cannot be generated, it is determined that the input loading unit is in an abnormal working state. After the computer device obtains the fourth detection result, it can detect other units according to the fourth detection result to obtain the target detection result. Optionally, after the computer device obtains the processing record of the input loading unit, it can extract the input signal and output signal of the input loading unit, and then generate a signal waveform according to the input signal and output signal, and determine the fourth detection result according to the signal waveform. If the fourth detection result indicates that the input loading unit is in an abnormal working state, it is determined that the input loading unit has a fault.

[0089] S4013, if the fourth detection result indicates that the input loading unit is in a normal working state, it is determined that the target detection result indicates that the image output unit has a fault.

[0090] Wherein, the image output unit is the OM.

[0091] In the embodiments of the present application, if the fourth detection result indicates that the input loading unit is in a normal working state, this situation indicates that the input loading unit is not faulty. Then, it can be determined that the image output unit has a fault.

[0092] S402, if the second detection result indicates that the rasterization unit is in an abnormal working state, it is determined that the target detection result indicates that the rasterization unit in the graphics processor has a fault.

[0093] In the embodiments of the present application, if the second detection result indicates that the rasterization unit is in an abnormal working state, it is determined that the target detection result indicates that the rasterization unit in the graphics processor has a fault.

[0094] In the method described in the embodiments of the present application, when an abnormal target image is detected, the rasterization unit is first checked. If the rasterization unit is normal, the detection scope is then expanded to other units. This hierarchical detection method improves the efficiency of fault location and reduces the time for troubleshooting unnecessary units.

[0095] In one embodiment, a specific implementation manner for obtaining a rasterization file output by a rasterization unit in a graphics processing unit is further provided. For example, Figure 7 as shown, the "obtaining a rasterization file output by a rasterization unit in a graphics processing unit" in step S201 above includes:

[0096] S501, obtaining coordinate information and attribute information output by each rasterization subunit in the rasterization unit.

[0097] Among them, as Figure 8 shown, the graphics processing unit adopts an xcore architecture. The rasterization unit in the graphics processing unit includes multiple rasterization subunits. Each rasterization subunit needs to instantiate a position monitor (Position uvm agent), an attribute monitor (Attribute uvm agent), and a processing unit (Quad processor). The position monitor is used to obtain the coordinate information output by the rasterization subunit, the attribute monitor is used to obtain the attribute information output by the rasterization subunit, such as color information, etc., and the processing unit is used to align the information output by the position monitor and the attribute monitor and generate a raster subunit file of the rasterization subunit.

[0098] In the embodiment of the present application, when the computer device detects that the target image output by the image output unit in the graphics processing unit is abnormal, it can obtain the coordinate information and attribute information output by each rasterization subunit from the processing record of the rasterization unit.

[0099] S502, generating a raster file according to the coordinate information and attribute information output by each rasterization subunit.

[0100] In the embodiment of the present application, after the computer device obtains the coordinate information and attribute information output by each rasterization subunit based on the above steps, it can first generate raster subunit files corresponding to each rasterization subunit, and then merge all the raster subunit files to obtain a rasterization file.

[0101] In one embodiment, a specific implementation manner for performing a fault detection on a rasterization unit according to a detected image is further provided. For example, Figure 9 as shown, the "performing a fault detection on a rasterization unit according to a detected image to obtain a first detection result" in step S202 above includes:

[0102] S601, determining whether the detected image is consistent with a standard image. If they are consistent, execute S602; if not, execute S603.

[0103] Among them, the standard image can be an expected image.

[0104] In the embodiments of the present application, after obtaining the detection image, the computer device may compare the detection image with the standard image. Specifically, it can be compared manually, or a preset algorithm or a preset model can be used for comparison, and then it is determined whether the detection image is consistent with the standard image according to the comparison result.

[0105] S602. Determine that the first detection result indicates that the rasterization unit is in a normal working state.

[0106] In the embodiments of the present application, if the detection image is consistent with the standard image, it is determined that the first detection result indicates that the rasterization unit is in a normal working state.

[0107] S603. Determine that the first detection result indicates that the rasterization unit is in a potentially abnormal state.

[0108] In the embodiments of the present application, if the detection image is inconsistent with the standard image, it is determined that the first detection result indicates that the rasterization unit is in a potentially abnormal state.

[0109] In combination with all the above embodiments, a method for detecting faults of a graphics processor is further provided. As Figure 10 shown, the method includes:

[0110] S701. When it is detected that the target image output by the image output unit in the graphics processor is abnormal, obtain the coordinate information and attribute information output by each rasterization subunit in the rasterization unit.

[0111] S702. Generate a rasterization file according to the coordinate information and attribute information output by each rasterization subunit.

[0112] S703. Generate a detection image according to the rasterization file and determine whether the detection image is consistent with the standard image.

[0113] S704. If they are consistent, determine that the first detection result indicates that the rasterization unit is in a normal working state, and determine that the target detection result indicates that the image output unit has a fault.

[0114] S705. If they are inconsistent, determine that the first detection result indicates that the rasterization unit is in a potentially abnormal state, and perform a fault detection on the rasterization unit according to the signal waveform output by the rasterization unit to obtain a second detection result.

[0115] S706. If the second detection result indicates that the rasterization unit is in an abnormal working state, determine that the target detection result indicates that the rasterization unit has a fault.

[0116] S707. If the second detection result indicates that the rasterization unit is in a normal working state, perform a detection on the subdivision processing unit according to the signal waveform output by the subdivision processing unit to obtain a third detection result.

[0117] S708, if the third detection result indicates that the segmentation processing unit is in an abnormal working state, then determine that the target detection result indicates that the segmentation processing unit has a fault.

[0118] S709, if the third detection result indicates that the segmentation processing unit is in a normal working state, then detect the input loading unit according to the signal waveform output by the input loading unit to obtain a fourth detection result.

[0119] S710, if the fourth detection result indicates that the input loading unit is in an abnormal working state, then determine that the target detection result indicates that the input loading unit has a fault.

[0120] S711, if the fourth detection result indicates that the input loading unit is in a normal working state, then determine that the target detection result indicates that the image output unit has a fault.

[0121] The methods described in the above steps have all been described in the foregoing embodiments. For detailed content, please refer to the foregoing description and will not be elaborated here.

[0122] It should be understood that although the steps in the flowcharts involved in the above-described embodiments are displayed in sequence according to the arrows, these steps do not necessarily need to be executed in the order indicated by the arrows. Unless there is a clear description in this article, the execution of these steps has no strict order limit, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-described embodiments may include multiple steps or multiple stages. These steps or stages do not necessarily need to be executed at the same moment, but can be executed at different moments. The execution order of these steps or stages is not necessarily sequential, but can be executed alternately or alternately with at least a part of other steps or steps or stages in other steps.

[0123] Based on the same inventive concept, an embodiment of the present application further provides a fault detection device for a graphics processor for implementing the above-mentioned fault detection method for a graphics processor. The implementation solution provided by this device to solve problems is similar to the implementation solution described in the above method. Therefore, the specific limitations in one or more embodiments of the following fault detection devices for a graphics processor can refer to the limitations on the fault detection method for a graphics processor in the above text and will not be elaborated here.

[0124] In one embodiment, as Figure 11 shown, a fault detection device for a graphics processor is provided, including:

[0125] An acquisition module 11, configured to acquire a rasterization file output by a rasterization unit in a graphics processor when it is detected that a target image output by an image output unit in the graphics processor is abnormal.

[0126] A first detection module 12, configured to generate a detection image according to the rasterization file, and perform a fault detection on the rasterization unit according to the detection image to obtain a first detection result.

[0127] A second detection module 13, configured to detect other units in the graphics processor according to the first detection result to obtain a target detection result.

[0128] In one embodiment, the above-mentioned second detection module 13 includes:

[0129] A first determination unit, configured to determine that the target detection result indicates that there is a fault in the image output unit in the graphics processor if the first detection result indicates that the rasterization unit is in a normal working state.

[0130] A second determination unit, configured to perform a fault detection on the rasterization unit according to the signal waveform output by the rasterization unit to obtain a second detection result if the first detection result indicates that the rasterization unit is in a potentially abnormal state, and detect other units according to the second detection result to obtain a target detection result.

[0131] In one embodiment, the above-mentioned second determination unit includes:

[0132] A detection subunit, configured to detect other units according to the signal waveform output by other units to obtain a target detection result if the second detection result indicates that the rasterization unit is in a normal working state.

[0133] A determination subunit, configured to determine that the target detection result indicates that there is a fault in the rasterization unit in the graphics processor if the second detection result indicates that the rasterization unit is in an abnormal working state.

[0134] In one embodiment, the above-mentioned detection subunit is specifically configured to detect a subdivision processing unit according to the signal waveform output by the subdivision processing unit to obtain a third detection result; when the third detection result indicates that the subdivision processing unit is in a normal working state, detect an input loading unit according to the signal waveform output by the input loading unit to obtain a fourth detection result; if the fourth detection result indicates that the input loading unit is in a normal working state, determine that the target detection result indicates that there is a fault in the image output unit.

[0135] In one embodiment, the above-mentioned acquisition module 11 includes:

[0136] An acquisition unit, configured to acquire the coordinate information and attribute information output by each rasterization subunit in the rasterization unit.

[0137] A generation unit, configured to generate a rasterization file according to the coordinate information and attribute information output by each rasterization subunit.

[0138] In one embodiment, the above-mentioned first detection module 12 includes:

[0139] A third determination unit, configured to determine whether the detected image is consistent with the standard image.

[0140] A fourth determination unit, configured to determine that the first detection result indicates that the rasterization unit is in a normal operating state if the detected image is consistent with the standard image.

[0141] A fifth determination unit, configured to determine that the first detection result indicates that the rasterization unit is in a potentially abnormal state if the detected image is inconsistent with the standard image.

[0142] Each module in the above-mentioned fault detection device of the graphics processor can be implemented in whole or in part by software, hardware, and their combination. The above-mentioned modules can be embedded in the processor in the computer device in hardware form or be independent of it, or can be stored in the memory in the computer device in software form, so as to facilitate the processor to call and execute the operations corresponding to the above-mentioned modules.

[0143] In one embodiment, a computer device is provided, including a memory and a processor. A computer program is stored in the memory. When the processor executes the computer program, the following steps are implemented:

[0144] When it is detected that there is an abnormality in the target image output by the image output unit in the graphics processor, obtain the rasterization file output by the rasterization unit in the graphics processor;

[0145] Generate a detected image according to the rasterization file, and perform fault detection on the rasterization unit according to the detected image to obtain a first detection result;

[0146] Detect other units in the graphics processor according to the first detection result to obtain a target detection result.

[0147] In one embodiment, when the processor executes the computer program, the following steps are further implemented:

[0148] If the first detection result indicates that the rasterization unit is in a normal operating state, determine that the target detection result indicates that there is a fault in the image output unit in the graphics processor;

[0149] If the first detection result indicates that the rasterization unit is in a potentially abnormal state, perform fault detection on the rasterization unit according to the signal waveform output by the rasterization unit to obtain a second detection result, and detect other units according to the second detection result to obtain a target detection result.

[0150] In one embodiment, when the processor executes the computer program, the following steps are further implemented:

[0151] If the second detection result indicates that the rasterization unit is in a normal operating state, detect other units based on the signal waveforms output by the other units to obtain a target detection result;

[0152] If the second detection result indicates that the rasterization unit is in an abnormal operating state, determine that the target detection result indicates that there is a fault in the rasterization unit of the graphics processor.

[0153] In one embodiment, when the processor executes the computer program, the following steps are further implemented:

[0154] Detect the tessellation processing unit based on the signal waveform output by the tessellation processing unit to obtain a third detection result;

[0155] When the third detection result indicates that the tessellation processing unit is in a normal operating state, detect the input loading unit based on the signal waveform output by the input loading unit to obtain a fourth detection result;

[0156] If the fourth detection result indicates that the input loading unit is in a normal operating state, determine that the target detection result indicates that there is a fault in the image output unit.

[0157] In one embodiment, when the processor executes the computer program, the following steps are further implemented:

[0158] Obtain the coordinate information and attribute information output by each rasterization subunit in the rasterization unit;

[0159] Generate a rasterization file based on the coordinate information and attribute information output by each rasterization subunit.

[0160] In one embodiment, when the processor executes the computer program, the following steps are further implemented:

[0161] Determine whether the detected image is consistent with the standard image;

[0162] If the detected image is consistent with the standard image, determine that the first detection result indicates that the rasterization unit is in a normal operating state;

[0163] If the detected image is inconsistent with the standard image, determine that the first detection result indicates that the rasterization unit is in a potentially abnormal state.

[0164] For the computer device provided in the above embodiment, its implementation principle and technical effects are similar to those of the above method embodiment, and will not be elaborated here.

[0165] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented:

[0166] When it is detected that the target image output by the image output unit in the graphics processor is abnormal, obtain the rasterization file output by the rasterization unit in the graphics processor;

[0167] Generate a detection image according to the rasterization file, and perform a fault detection on the rasterization unit according to the detection image to obtain a first detection result;

[0168] Detect other units in the graphics processor according to the first detection result to obtain a target detection result.

[0169] In one embodiment, when the computer program is executed by the processor, the following steps are further implemented:

[0170] If the first detection result indicates that the rasterization unit is in a normal working state, it is determined that the target detection result indicates that there is a fault in the image output unit in the graphics processor;

[0171] If the first detection result indicates that the rasterization unit is in a potentially abnormal state, perform a fault detection on the rasterization unit according to the signal waveform output by the rasterization unit to obtain a second detection result, and detect other units according to the second detection result to obtain a target detection result.

[0172] In one embodiment, when the computer program is executed by the processor, the following steps are further implemented:

[0173] If the second detection result indicates that the rasterization unit is in a normal working state, detect other units according to the signal waveform output by other units to obtain a target detection result;

[0174] If the second detection result indicates that the rasterization unit is in an abnormal working state, it is determined that the target detection result indicates that there is a fault in the rasterization unit in the graphics processor.

[0175] In one embodiment, when the computer program is executed by the processor, the following steps are further implemented:

[0176] Detect the tessellation processing unit according to the signal waveform output by the tessellation processing unit to obtain a third detection result;

[0177] When the third detection result indicates that the tessellation processing unit is in a normal working state, detect the input loading unit according to the signal waveform output by the input loading unit to obtain a fourth detection result;

[0178] If the fourth detection result indicates that the input loading unit is in a normal working state, it is determined that the target detection result indicates that there is a fault in the image output unit.

[0179] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0180] Obtain the coordinate information and attribute information output by each rasterization subunit in the rasterization unit;

[0181] Generate a rasterization file according to the coordinate information and attribute information output by each rasterization subunit.

[0182] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0183] Determine whether the detected image is consistent with the standard image;

[0184] If the detected image is consistent with the standard image, determine that the first detection result indicates that the rasterization unit is in a normal working state;

[0185] If the detected image is inconsistent with the standard image, determine that the first detection result indicates that the rasterization unit is in a potentially abnormal state.

[0186] For the computer-readable storage medium provided in the above embodiment, its implementation principle and technical effects are similar to those of the above method embodiment, and will not be elaborated here.

[0187] In one embodiment, a computer program product is provided, including a computer program, which when executed by a processor, implements the following steps:

[0188] In the case where an abnormality is detected in the target image output by the image output unit in the graphics processor, obtain the rasterization file output by the rasterization unit in the graphics processor;

[0189] Generate a detected image according to the rasterization file, and perform a fault detection on the rasterization unit according to the detected image to obtain a first detection result;

[0190] Detect other units in the graphics processor according to the first detection result to obtain a target detection result.

[0191] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0192] If the first detection result indicates that the rasterization unit is in a normal working state, determine that the target detection result indicates that there is a fault in the image output unit in the graphics processor;

[0193] If the first detection result indicates that the rasterization unit is in a potentially abnormal state, perform a fault detection on the rasterization unit according to the signal waveform output by the rasterization unit to obtain a second detection result, and detect other units according to the second detection result to obtain a target detection result.

[0194] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0195] If the second detection result indicates that the rasterization unit is in a normal operating state, then detect other units based on the signal waveforms output by the other units to obtain a target detection result;

[0196] If the second detection result indicates that the rasterization unit is in an abnormal operating state, then determine that the target detection result indicates that there is a fault in the rasterization unit of the graphics processor.

[0197] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0198] Detect the tessellation processing unit based on the signal waveform output by the tessellation processing unit to obtain a third detection result;

[0199] When the third detection result indicates that the tessellation processing unit is in a normal operating state, detect the input loading unit based on the signal waveform output by the input loading unit to obtain a fourth detection result;

[0200] If the fourth detection result indicates that the input loading unit is in a normal operating state, then determine that the target detection result indicates that there is a fault in the image output unit.

[0201] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0202] Obtain the coordinate information and attribute information output by each rasterization subunit in the rasterization unit;

[0203] Generate a rasterization file according to the coordinate information and attribute information output by each rasterization subunit.

[0204] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented:

[0205] Determine whether the detected image is consistent with the standard image;

[0206] If the detected image is consistent with the standard image, then determine that the first detection result indicates that the rasterization unit is in a normal operating state;

[0207] If the detected image is inconsistent with the standard image, then determine that the first detection result indicates that the rasterization unit is in a potentially abnormal state.

[0208] For the computer program product provided in the above embodiment, its implementation principle and technical effects are similar to those of the above method embodiment, and will not be elaborated here.

[0209] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, database, or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memories. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The databases involved in the embodiments provided in the present application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the embodiments provided in the present application can be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, data processing logics based on quantum computing, etc., without limitation.

[0210] The technical features of the above embodiments can be combined arbitrarily. For the sake of concise description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope recorded in this specification.

[0211] The above-described embodiments only represent several implementation manners of the present application. The description is relatively specific and detailed, but it should not be construed as a limitation on the patent scope of the present application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the appended claims.

Claims

1. A method for detecting faults in a graphics processor, characterized in that The method includes: When it is detected that the target image output by the image output unit in the graphics processor is abnormal, obtaining the rasterization file output by the rasterization unit in the graphics processor; Generating a detection image according to the rasterization file, and performing a fault detection on the rasterization unit according to the detection image to obtain a first detection result; If the first detection result indicates that the rasterization unit is in a normal working state, it is determined that the target detection result indicates that the image output unit in the graphics processor is faulty; If the first detection result indicates that the rasterization unit is in a potentially abnormal state, performing a fault detection on the rasterization unit according to the signal waveform output by the rasterization unit to obtain a second detection result, and performing a detection on other units in the graphics processor according to the second detection result to obtain a target detection result.

2. The method according to claim 1, wherein The performing a detection on other units according to the second detection result to obtain a target detection result includes: If the second detection result indicates that the rasterization unit is in a normal working state, performing a detection on other units according to the signal waveform output by the other units to obtain the target detection result; If the second detection result indicates that the rasterization unit is in an abnormal working state, it is determined that the target detection result indicates that the rasterization unit in the graphics processor is faulty.

3. The method according to claim 2, wherein The other units include a tessellation processing unit, an input loading unit, and the image output unit. The performing a detection on other units according to the signal waveform output by the other units to obtain the target detection result includes: Performing a detection on the tessellation processing unit according to the signal waveform output by the tessellation processing unit to obtain a third detection result; When the third detection result indicates that the tessellation processing unit is in a normal working state, performing a detection on the input loading unit according to the signal waveform output by the input loading unit to obtain a fourth detection result; If the fourth detection result indicates that the input loading unit is in a normal working state, it is determined that the target detection result indicates that the image output unit is faulty.

4. The method according to claim 3, characterized in that The performing a detection on the tessellation processing unit according to the signal waveform output by the tessellation processing unit to obtain a third detection result includes: Comparing the signal waveform with an expected waveform to determine whether the signal waveform is consistent with the expected waveform; If the signal waveform is consistent with the expected waveform, it is determined that the third detection result indicates that the tessellation processing unit is in a normal working state; If the signal waveform is inconsistent with the expected waveform, it is determined that the third detection result indicates that the tessellation processing unit is in an abnormal working state.

5. The method according to claim 1, characterized in that, The obtaining the rasterization file output by the rasterization unit in the graphics processor includes: Obtaining the coordinate information and attribute information output by each rasterization subunit in the rasterization unit; Generating a rasterization file according to the coordinate information and attribute information output by each rasterization subunit.

6. The method according to any one of claims 1-5, characterized in that, The performing a fault detection on the rasterization unit according to the detection image to obtain a first detection result includes: Determine whether the detected image is consistent with the standard image; If they are consistent, determine that the first detection result indicates that the rasterization unit is in a normal working state; If they are inconsistent, determine that the first detection result indicates that the rasterization unit is in a potentially abnormal state.

7. A fault detection device for a graphics processor, characterized in that, The device includes: An acquisition module, configured to acquire a rasterization file output by a rasterization unit in the graphics processor when it is detected that a target image output by an image output unit in the graphics processor is abnormal; A first detection module, configured to generate a detection image according to the rasterization file, and perform a fault detection on the rasterization unit according to the detection image to obtain a first detection result; A second detection module, configured to, if the first detection result indicates that the rasterization unit is in a normal working state, determine that the target detection result indicates that there is a fault in the image output unit in the graphics processor; if the first detection result indicates that the rasterization unit is in a potentially abnormal state, perform a fault detection on the rasterization unit according to a signal waveform output by the rasterization unit to obtain a second detection result, and perform a detection on other units in the graphics processor according to the second detection result to obtain a target detection result.

8. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that, When this computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images