Computer graphic application compatibility conflict automatic positioning method

Abstract

The invention discloses an automatic positioning method for compatibility conflicts of computer graphic applications, and relates to the technical field of computer graphic processing, and the method comprises the following steps: obtaining conflict association multi-source data of the computer graphic applications; the conflict association multi-source data comprises code compiling adaptation data, graphic rendering output data, system resource scheduling data and cross-platform interaction log data; performing anomaly detection and feature extraction on the conflict association multi-source data to obtain a conflict feature vector, matching a conflict type label according to the conflict feature vector, and generating a conflict influence evaluation parameter according to the conflict type label; the conflict influence evaluation parameters comprise comprehensive quantitative indexes of conflicts to application operation performance, function implementation and user experience, and the effect is to avoid resource waste of full-quantity verification; the positioning accuracy and the verification efficiency are effectively balanced, and the conflict checking cost in a cross-platform scene is remarkably reduced.

Description

Technical Field

[0001] This invention relates to the field of computer graphics processing technology, and more specifically, to an automated method for locating compatibility conflicts in computer graphics applications. Background Technology

[0002] With the deep penetration of computer graphics technology into game development, industrial simulation, digital twins, and other fields, the demand for cross-platform deployment of graphics applications has exploded. These applications face increasing compatibility issues when running on multiple operating systems such as Windows, Linux, and mobile devices, as well as on hardware environments with different GPU architectures and driver versions. The causes of these conflicts are often distributed throughout the entire chain, from code compilation to cross-platform interaction. For example, incompatible dependency library versions during compilation, differences in pipeline adaptation during rendering, imbalances in computing power allocation during resource scheduling, or heterogeneous protocols during interaction can ultimately manifest as image distortion, sudden drops in frame rates, functional abnormalities, or even application crashes, severely hindering the efficiency of graphics application deployment and user experience.

[0003] Traditional compatibility conflict localization relies primarily on developers' experience-based troubleshooting, using full log analysis and manual breakpoint debugging to pinpoint issues. This approach is not only labor-intensive and time-consuming, but also struggles to understand the dependencies and conflict propagation paths between components when dealing with complex graphics applications containing dozens or even hundreds of components. Furthermore, existing technologies often focus on single-dimensional anomaly detection, such as isolated analysis of rendering output or resource usage data, lacking end-to-end modeling of code compilation, graphics rendering, resource scheduling, and cross-platform interactions, making it difficult to trace the underlying root cause of conflicts at the system level. Summary of the Invention

[0004] To address the shortcomings of existing technologies, the present invention aims to provide an automated method for locating compatibility conflicts in computer graphics applications.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] An automated method for locating compatibility conflicts in computer graphics applications, comprising the following steps:

[0007] Acquire conflict-related multi-source data for computer graphics applications; the conflict-related multi-source data includes code compilation and adaptation data, graphics rendering output data, system resource scheduling data, and cross-platform interaction log data;

[0008] Anomaly detection and feature extraction are performed on conflict-related multi-source data to obtain conflict feature vectors. Conflict type labels are matched based on the conflict feature vectors, and conflict impact assessment parameters are generated based on the conflict type labels. The conflict impact assessment parameters include comprehensive quantitative indicators of the impact of the conflict on application performance, functional implementation, and user experience.

[0009] Based on the conflict impact assessment parameters and the preset positioning requirements, the key verification dimensions for conflict positioning are determined, and the resource consumption of the key verification dimensions is judged to obtain the conflict positioning confidence coefficient.

[0010] Based on the component dependencies of computer graphics applications, conflict-related components are prioritized to obtain core conflict components and secondary conflict components. The component localization efficiency coefficient is obtained based on the verification complexity and association strength of the core conflict components and secondary conflict components.

[0011] Automatic collision localization is performed in computer graphics applications based on the collision localization confidence coefficient and the component localization efficiency coefficient.

[0012] Preferably, the conflict feature vector is obtained by performing anomaly detection and feature extraction on conflict-related multi-source data, specifically including the following steps:

[0013] The code compilation adaptation data is subjected to syntax judgment and compatibility rule verification, and compilation layer features are extracted, including compilation error codes, incompatible syntax structures, and dependency library version conflicts.

[0014] Perform pixel-level comparison and rendering quality evaluation on the graphics rendering output data, and extract rendering layer features, including image distortion parameters, frame rate fluctuation range, and rendering stutter duration.

[0015] Real-time monitoring and statistical analysis of system resource scheduling data are performed to extract resource layer features, including peak CPU usage, memory leaks, and GPU computing power allocation ratio.

[0016] Log parsing and semantic mining are performed on cross-platform interaction log data to extract interaction layer features, including interaction protocol incompatibility identifiers, data transmission failure records, and platform response anomaly information.

[0017] The extracted features from the compilation layer, rendering layer, resource layer, and interaction layer are combined to form a conflict feature vector.

[0018] Preferably, the conflict type label is matched based on the conflict feature vector, and conflict impact assessment parameters are generated based on the conflict type label, specifically including the following steps:

[0019] A pre-defined conflict type tag library is provided, which contains feature vector templates and influence weight configurations corresponding to various compatibility conflicts. Each conflict type tag is associated with an influence evaluation dimension.

[0020] Calculate the similarity between the conflict feature vector and each feature vector template in the tag library, and mark the conflict type label corresponding to the feature vector template with the highest similarity as the target label;

[0021] The conflict impact assessment parameters are calculated based on the impact assessment dimensions corresponding to the target labels and the quantified values ​​of each feature in the conflict feature vector.

[0022] Preferably, based on the conflict impact assessment parameters and preset positioning requirements, the key verification dimensions for conflict positioning are determined, specifically including the following steps:

[0023] The impact values ​​of operational performance, functional implementation, and user experience in the conflict impact assessment parameters are used as impact factors, and the comprehensive impact value is obtained by calculating the impact factors.

[0024] If the overall impact value is greater than or equal to the preset high impact threshold, the key verification dimensions include full-dimensional verification of the code layer, rendering layer, resource layer, and interaction layer.

[0025] If the overall impact score is less than the preset high impact threshold, but greater than or equal to the preset medium impact threshold, then the key verification dimensions include the verification of the rendering layer and interaction layer corresponding to the core function.

[0026] If the overall impact score is less than the preset impact threshold, then only the resource layer verification related to runtime performance will be retained as the key verification dimension.

[0027] Based on the positioning requirements, candidate verification dimensions that meet the reliability threshold are selected, and these candidate verification dimensions are combined to form key verification dimensions.

[0028] Preferably, the conflict location confidence coefficient is obtained by judging resource consumption on key verification dimensions, specifically including the following steps:

[0029] The resource consumption parameters corresponding to the key verification dimensions are calculated. The resource consumption parameters include the computing power resources, time costs, and storage resource usage required for the verification process.

[0030] Based on the resource loss parameters and the preset maximum resource loss threshold, calculate the resource loss adaptation rate.

[0031] The conflict location confidence coefficient is calculated by combining the comprehensive impact value of the conflict impact assessment parameters. The conflict location confidence coefficient is positively correlated with the comprehensive impact value and negatively correlated with the resource loss adaptation rate.

[0032] Preferably, based on the component dependencies of a computer graphics application, conflict-related components are prioritized to obtain core conflict components and secondary conflict components, specifically including the following steps:

[0033] By using component dependency identification tools, identify all functional components of the application and the dependencies between them, including direct and indirect dependencies;

[0034] Calculate the conflict propagation weight of each component, whereby the conflict propagation weight is a quantified value of the scope and degree of influence of the component on other components when a conflict occurs.

[0035] Set a weight threshold, mark components with conflict propagation weights greater than or equal to the threshold as core conflict components, and mark components with conflict propagation weights less than the threshold as secondary conflict components.

[0036] Preferably, the component localization efficiency coefficient is obtained based on the verification complexity and correlation strength of the core conflict component and the secondary conflict component, specifically including the following steps:

[0037] The verification complexity coefficient of each conflicting component is determined by assessing its code size, logical complexity, and number of interfaces.

[0038] Calculate the correlation strength value between each conflicting component and the compatibility conflict feature vector;

[0039] The component localization efficiency coefficient is calculated by using the conflict propagation weights of the core conflict component and the secondary conflict component as the base weights, combined with the verification complexity coefficient and the association strength value.

[0040] Preferably, the automatic location of collisions in computer graphics applications is performed based on the collision location confidence coefficient and the component location efficiency coefficient, specifically including the following steps:

[0041] If the conflict location confidence coefficient is greater than or equal to the component location efficiency coefficient, then the core conflict component and the secondary conflict component are simultaneously verified and investigated according to the key verification dimensions until the root cause of the conflict is located.

[0042] If the conflict location confidence coefficient is less than the component location efficiency coefficient, the core conflict component will be verified and investigated first according to the key verification dimensions. After that, the secondary conflict components will be verified step by step. If the root cause of the conflict is located in the middle, the subsequent verification process will be terminated.

[0043] Compared with existing technologies, this invention has the following advantages: By integrating multi-source data such as code compilation and adaptation, graphics rendering output, system resource scheduling, and cross-platform interaction logs, a full-link feature vector from the underlying code to the user experience is constructed, breaking through the limitations of single-dimensional technology detection; by matching a preset conflict type tag library and generating quantitative impact assessment parameters, the underlying root cause of the conflict can be accurately traced, avoiding local errors caused by empirical investigation and significantly improving the accuracy of conflict location in complex scenarios; by using dual-dimensional quantitative indicators of conflict location confidence coefficient and component location efficiency coefficient, combined with preset thresholds to dynamically adjust the verification strategy, when the conflict location confidence is high, core and secondary components are verified simultaneously to quickly stop the loss; when the confidence is insufficient, priority is given to focusing on the core components for layered investigation, avoiding the waste of resources in full verification; effectively balancing location accuracy and verification efficiency, significantly reducing the cost of conflict investigation in cross-platform scenarios. Attached Figure Description

[0044] Figure 1 This invention provides a schematic diagram illustrating the steps of an automated method for locating compatibility conflicts in computer graphics applications.

[0045] Figure 2 This is a schematic diagram illustrating the steps of obtaining the conflict feature vector in an automated method for locating compatibility conflicts in computer graphics applications, as provided in an embodiment of the present invention. Detailed Implementation

[0046] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0047] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0048] Secondly, the term "an embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places throughout this specification does not necessarily refer to the same embodiment, nor is it a single embodiment or an embodiment selectively excluded from other embodiments.

[0049] Reference Figures 1-2 As shown.

[0050] Example 1 further illustrates the automatic location method for compatibility conflicts in computer graphics applications proposed in this invention.

[0051] An automated method for locating compatibility conflicts in computer graphics applications, comprising the following steps:

[0052] Acquire conflict-related multi-source data for computer graphics applications; the conflict-related multi-source data includes code compilation and adaptation data, graphics rendering output data, system resource scheduling data, and cross-platform interaction log data;

[0053] Anomaly detection and feature extraction are performed on conflict-related multi-source data to obtain conflict feature vectors. Conflict type labels are matched based on the conflict feature vectors, and conflict impact assessment parameters are generated based on the conflict type labels. The conflict impact assessment parameters include comprehensive quantitative indicators of the impact of the conflict on application performance, functional implementation, and user experience.

[0054] Based on the conflict impact assessment parameters and the preset positioning requirements, the key verification dimensions for conflict positioning are determined, and the resource consumption of the key verification dimensions is judged to obtain the conflict positioning confidence coefficient.

[0055] Based on the component dependencies of computer graphics applications, conflict-related components are prioritized to obtain core conflict components and secondary conflict components. The component localization efficiency coefficient is obtained based on the verification complexity and association strength of the core conflict components and secondary conflict components.

[0056] Automatic collision localization is performed in computer graphics applications based on the collision localization confidence coefficient and the component localization efficiency coefficient.

[0057] The process of anomaly detection and feature extraction from conflict-related multi-source data to obtain conflict feature vectors includes the following steps:

[0058] The code compilation adaptation data is subjected to syntax judgment and compatibility rule verification, and compilation layer features are extracted, including compilation error codes, incompatible syntax structures, and dependency library version conflicts.

[0059] Perform pixel-level comparison and rendering quality evaluation on the graphics rendering output data, and extract rendering layer features, including image distortion parameters, frame rate fluctuation range, and rendering stutter duration.

[0060] Real-time monitoring and statistical analysis of system resource scheduling data are performed to extract resource layer features, including peak CPU usage, memory leaks, and GPU computing power allocation ratio.

[0061] Log parsing and semantic mining are performed on cross-platform interaction log data to extract interaction layer features, including interaction protocol incompatibility identifiers, data transmission failure records, and platform response anomaly information.

[0062] The extracted features from the compilation layer, rendering layer, resource layer, and interaction layer are combined to form a conflict feature vector.

[0063] First, syntax checking and compatibility rule verification are performed on the code compilation adaptation data. When computer graphics applications are compiled under different operating systems or compiler environments, compilation errors may occur, or certain syntax structures supported only on specific platforms may be encountered, or conflicts may arise due to incompatible dependency library versions. These issues are extracted as compilation-level features and used as the underlying basis for identifying compatibility problems.

[0064] Next, pixel-level comparison and rendering quality evaluation are performed on the graphics rendering output data. For example, when the same 3D model is rendered under different graphics card drivers, there may be image texture distortion, excessive fluctuations in frame rate in different scenes, or rendering stuttering in complex scenes. These image distortion parameters, frame rate fluctuation range, and rendering stuttering duration will be accurately captured to form rendering layer features, which directly reflect the compatibility issues in the graphics output stage.

[0065] Then, the system resource scheduling data is monitored and statistically analyzed in real time. During the operation of graphics applications, high-load scenarios may result in excessively high CPU usage peaks, continuous memory leaks, or unreasonable GPU computing power allocation leading to the underutilization of some hardware performance. These CPU usage peaks, memory leaks, and GPU computing power allocation ratios are extracted as resource-layer features to reveal potential conflicts from the perspective of hardware resource scheduling.

[0066] Next, log parsing and semantic mining are performed on the cross-platform interaction log data. When applications interact with data across different platforms, communication failures may occur due to incompatible interaction protocols, or packet loss or platform response timeouts may occur during data transmission. These incompatible interaction protocol identifiers, data transmission failure records, and platform response anomaly information are extracted as interaction layer features to reflect compatibility issues in the cross-platform communication process.

[0067] Finally, the system organically combines compilation layer features, rendering layer features, resource layer features, and interaction layer features to form a compatibility conflict feature vector for the application. This feature vector contains anomaly information across the entire chain from code compilation to cross-platform interaction, providing comprehensive and accurate data support for subsequent conflict type matching and impact assessment. This allows the system to automatically locate compatibility conflicts based on complete feature dimensions.

[0068] Matching conflict type labels to conflict feature vectors and generating conflict impact assessment parameters based on these labels involves the following steps:

[0069] A pre-defined conflict type tag library is provided, which contains feature vector templates and influence weight configurations corresponding to various compatibility conflicts. Each conflict type tag is associated with an influence evaluation dimension.

[0070] Calculate the similarity between the conflict feature vector and each feature vector template in the tag library, and mark the conflict type label corresponding to the feature vector template with the highest similarity as the target label;

[0071] The conflict impact assessment parameters are calculated based on the impact assessment dimensions corresponding to the target labels and the quantified values ​​of each feature in the conflict feature vector.

[0072] First, a conflict type tag library is constructed, which forms the foundation for all subsequent matching and calculations. This library contains feature vector templates corresponding to various compatibility conflicts. Each type of conflict has its unique feature combination pattern; for example, a dependency library version conflict at the compilation layer would have a feature vector template containing specific compilation error codes, dependency library version mismatch indicators, and other features. Simultaneously, the tag library configures an impact weight value for each conflict type tag. These weight values ​​are based on numerous historical cases and industry experience, reflecting the degree of impact of different conflicts in real-world scenarios. Each conflict type tag is also associated with a corresponding impact assessment dimension. For instance, a dependency library version conflict might be associated with dimensions such as application startup success rate and functional module completeness, while a rendering stuttering conflict might be associated with dimensions such as frame rate stability and screen smoothness.

[0073] Next, the system will calculate the similarity between the current conflicting feature vector and all feature vector templates in the tag library. A commonly used similarity calculation method is cosine similarity, and its formula is: Here, A represents the current conflict feature vector, and B represents the feature vector template in the tag library. The system iterates through all templates in the tag library, calculates a set of similarity values, and finally marks the conflict type tag corresponding to the template with the highest similarity as the target tag. For example, when the current conflict feature vector contains features such as compilation error code C1001 and a mismatch between dependency library versions v2.0 and v3.0, the system will compare it with the templates in the tag library that have dependency library version conflicts. If the similarity reaches 0.9 or higher, the dependency library version conflict will be marked as the target tag.

[0074] Finally, the system calculates the conflict impact assessment parameters based on the impact assessment dimensions corresponding to the target label and the quantified values ​​of each feature in the conflict feature vector. Assume that the impact assessment dimensions associated with the target label include three dimensions: application performance, functionality implementation, and user experience. Each dimension corresponds to different feature quantified values ​​and weights. Taking application performance as an example, its corresponding feature quantified values ​​might include peak CPU usage x1 and memory leak amount x2, with corresponding weights w1 and w2 respectively; the feature quantified values ​​corresponding to the functionality implementation dimension might include the number of compilation errors x3 and the number of rendering failures x4, with weights w3 and w4; the feature quantified values ​​corresponding to the user experience dimension might include frame rate fluctuation range x5 and rendering stutter duration x6, with weights w5 and w6. Therefore, the calculation formula for the conflict impact assessment parameters is: This formula allows the system to integrate scattered feature quantification values ​​into a comprehensive evaluation parameter, thereby quantifying the overall impact of the conflict on the application. For example, when a dependency library version conflict is marked as a target label, the system extracts quantification values ​​from the conflict feature vector, such as a peak CPU usage of 85%, a memory leak of 120MB, and 3 compilation errors. These values ​​are then combined with preset weights to calculate a specific conflict impact evaluation parameter, providing a clear quantitative basis for subsequent conflict localization and handling.

[0075] Based on the conflict impact assessment parameters and the preset positioning requirements, the key verification dimensions for conflict positioning are determined, specifically including the following steps:

[0076] The impact values ​​of operational performance, functional implementation, and user experience in the conflict impact assessment parameters are used as impact factors, and the comprehensive impact value is obtained by calculating the impact factors.

[0077] If the overall impact value is greater than or equal to the preset high impact threshold, the key verification dimensions include full-dimensional verification of the code layer, rendering layer, resource layer, and interaction layer.

[0078] If the overall impact score is less than the preset high impact threshold, but greater than or equal to the preset medium impact threshold, then the key verification dimensions include the verification of the rendering layer and interaction layer corresponding to the core function.

[0079] If the overall impact score is less than the preset impact threshold, then only the resource layer verification related to runtime performance will be retained as the key verification dimension.

[0080] Based on the positioning requirements, candidate verification dimensions that meet the reliability threshold are selected, and these candidate verification dimensions are combined to form key verification dimensions.

[0081] First, core impact factors—operational performance impact value, function implementation impact value, and user experience impact value—are extracted from the conflict impact assessment parameters, and the comprehensive impact value is calculated. The formula for calculating the comprehensive impact value is: Where P represents the performance impact value, reflecting the degree of interference of the conflict on the scheduling of resources such as CPU and memory; F represents the function implementation impact value, reflecting the degree of obstruction of the conflict on the normal operation of the application's core functions; U represents the user experience impact value, measuring the degree of impact of the conflict on screen smoothness and interaction response speed; α, β, and γ are preset weight coefficients, which are adjusted according to the business needs of different scenarios. For example, in industrial graphics applications that pursue stability, the weight of α will be set higher, while in entertainment graphics applications aimed at ordinary consumers, the weight of γ will be relatively larger. After calculating the comprehensive impact value, the system will compare it with the preset high impact threshold and medium impact threshold to dynamically determine the initial range of verification dimensions. For example, when a graphics application crashes frequently while running across platforms, the calculated comprehensive impact value S = 0.92, while the preset high impact threshold is 0.8. In this case, the system determines that the comprehensive impact value is greater than the high impact threshold, and the key verification dimensions will cover the full-dimensional verification of the code layer, rendering layer, resource layer, and interaction layer. This means that the system will simultaneously check whether there are syntax errors in code compilation and adaptation, whether there is image distortion in graphics rendering output, whether there is memory leak in system resource scheduling, and whether there are protocol incompatibility issues in cross-platform interaction logs, to ensure a comprehensive investigation of all links that may cause serious conflicts.

[0082] If the overall impact value is less than the high impact threshold but greater than or equal to the medium impact threshold—for example, if the application only experiences slight rendering stuttering in complex scenes, the calculated overall impact value S = 0.65, while the preset medium impact threshold is 0.5—then the key verification dimensions will focus on the rendering and interaction layers corresponding to the core functions. The system will compare the pixel-level data of the rendering output, analyze the frame rate fluctuation range and rendering stutter duration, and simultaneously parse cross-platform interaction logs to check for issues such as data transmission latency or response timeouts, avoiding resource waste caused by full-dimensional verification.

[0083] If the overall impact value is less than the medium impact threshold, for example, if the application only experiences a slight increase in CPU usage while running in the background, the calculated overall impact value S = 0.32. In this case, the key verification dimensions only retain resource layer verification related to runtime performance. The system will monitor peak CPU usage, memory leaks, and GPU computing power allocation in real time, and locate potential problems at the resource scheduling level through statistical analysis. No additional verification of the rendering or interaction layers is required, thereby improving the efficiency of conflict localization.

[0084] Finally, the system combines preset positioning requirements to perform reliability checks on the initially selected verification dimensions, filtering out candidate verification dimensions that meet the reliability threshold requirements. These candidate dimensions are then combined to form the key verification dimensions. For example, in scenarios where the positioning requirement is rapid investigation, the reliability threshold will be set relatively leniently, and the system will prioritize dimensions with lower data collection and analysis costs. Conversely, in scenarios where the positioning requirement is precise location, the reliability threshold will be set more strictly, and the system will retain more dimensions to ensure the accuracy of the positioning results.

[0085] The resource consumption assessment of key verification dimensions yields the conflict location confidence coefficient, specifically including the following steps:

[0086] The resource consumption parameters corresponding to the key verification dimensions are calculated. The resource consumption parameters include the computing power resources, time costs, and storage resource usage required for the verification process.

[0087] Based on the resource loss parameters and the preset maximum resource loss threshold, calculate the resource loss adaptation rate.

[0088] The conflict location confidence coefficient is calculated by combining the comprehensive impact value of the conflict impact assessment parameters. The conflict location confidence coefficient is positively correlated with the comprehensive impact value and negatively correlated with the resource loss adaptation rate.

[0089] First, the system comprehensively calculates the resource consumption parameters corresponding to the key verification dimensions. These parameters mainly include: the computing resources required for the verification process, such as the GPU computing power consumption when performing pixel-level rendering comparison; time costs, such as the processing time required to parse cross-platform interaction logs; and storage resource consumption, such as the disk space required to store multi-source data feature vectors. For example, when the key verification dimension is full-dimensional verification, the system needs to simultaneously calculate the computing power consumption of code layer syntax validation, the time cost of pixel comparison in the rendering layer, the storage consumption of real-time monitoring in the resource layer, and the computing power and time expenditure of parsing interaction layer logs, to ensure that there is a complete quantitative basis for the resource consumption of the verification process.

[0090] Next, the system will calculate the resource loss adaptation rate based on these resource loss parameters and the preset maximum resource loss threshold. The formula for calculating the resource loss adaptation rate is: , where C c T represents the actual consumption of computing resources. c S represents the actual time cost. c Represents the actual storage resource usage; C max T max S maxThese are the preset maximum computing power resource threshold, maximum time cost threshold, and maximum storage resource threshold, respectively. This formula reflects the degree of adaptation between the current verification process's resource consumption and the system's acceptable upper limit. The value ranges from 0 to 1; the closer the value is to 1, the more controllable the resource consumption. For example, if the actual computing power consumption for full-dimensional verification is 80 units, the time cost is 30 units, and the storage usage is 20 units, while the preset maximum thresholds are 100, 50, and 40 units respectively, then the calculated resource consumption adaptation rate is... This indicates that the current verification resource consumption is close to the upper limit of the threshold, and the adaptation rate is low.

[0091] Finally, the system combines the comprehensive impact values ​​of the conflict impact assessment parameters to calculate the conflict location confidence coefficient. The formula for this coefficient is: Here, S represents the overall impact value, which is positively correlated with the conflict location confidence coefficient, meaning that the greater the overall impact of the conflict, the higher the confidence level of the location result. R represents the resource consumption adaptation rate, which is negatively correlated with the conflict location confidence coefficient, indicating that the closer the resource consumption during the verification process is to the upper limit of the threshold, the lower the adaptation rate, and the lower the confidence level of the location result will be. For example, when the overall impact value of a conflict is S=0.9 and the resource consumption adaptation rate is R=0.316, the conflict location confidence coefficient C=0.9×0.316≈0.284; while if the overall impact value of another conflict is S=0.7, but the resource consumption adaptation rate is R=0.8, then the conflict location confidence coefficient C=0.7×0.8=0.56. This result shows that even if the former conflict has a greater impact, its location confidence is lower than the latter due to excessive verification resource consumption. The system will use this coefficient to weigh the reliability of the location result, providing a more accurate quantitative reference for subsequent conflict handling.

[0092] Based on the component dependencies of computer graphics applications, conflict-related components are prioritized to obtain core conflict components and secondary conflict components. This process includes the following steps:

[0093] By using component dependency identification tools, identify all functional components of the application and the dependencies between them, including direct and indirect dependencies;

[0094] Calculate the conflict propagation weight of each component, whereby the conflict propagation weight is a quantified value of the scope and degree of influence of the component on other components when a conflict occurs.

[0095] Set a weight threshold, mark components with conflict propagation weights greater than or equal to the threshold as core conflict components, and mark components with conflict propagation weights less than the threshold as secondary conflict components.

[0096] First, the system uses a component dependency assessment tool to comprehensively identify all functional components of the computer graphics application and their inter-component dependencies, including both direct and indirect dependencies. Direct dependencies refer to other components that a component directly calls at runtime; for example, the rendering component directly depends on the graphics driver adapter component. Indirect dependencies are dependencies passed from other components; for example, the UI / UX component depends on the rendering component, and the rendering component in turn depends on the graphics driver adapter component, thus the UI / UX component indirectly depends on the graphics driver adapter component. By analyzing the complete dependency network, the system can clearly understand the scope of each component's associations within the application, laying the foundation for subsequent calculations of conflict propagation weights.

[0097] Calculate the conflict propagation weight for each component. This weight is a quantified value representing the scope and extent of the impact of a conflict on other components. The formula for calculating the conflict propagation weight is: Where D represents the number of directly dependent components of the component, reflecting the direct impact of the conflict; I represents the number of indirectly dependent components of the component, reflecting the indirect impact of the conflict; and N represents the total number of components in the application, used to normalize the impact range and ensure that the weight value is within a reasonable range. For example, in a graphics application, the graphics driver adapter component has 2 directly dependent components (rendering and effects components) and 3 indirectly dependent components (interface interaction and scene loading components). The total number of components in the application is 10. Then, the conflict propagation weight of this component is W=2.6. This value indicates that when the graphics driver adapter component has a conflict, it will directly affect 2 components and indirectly affect 3 components, resulting in a high propagation range and degree on the overall application.

[0098] Finally, the system sets a weight threshold, marking components with a conflict propagation weight greater than or equal to the threshold as core conflict components, and those with a weight less than the threshold as secondary conflict components. For example, if the preset weight threshold is 2.0, then the weight of the aforementioned graphics-driven adaptation component, 2.6, is greater than the threshold and will be marked as a core conflict component; while if a UI component has only one direct dependency component (interface layout component) and zero indirect dependencies, its calculated weight is W=1.0, which is less than the threshold, and it will be marked as a secondary conflict component. Through this priority division, the system can focus on core conflict components for in-depth verification, reducing invalid checks on secondary components, thereby significantly improving the efficiency and accuracy of conflict localization.

[0099] The component localization efficiency coefficient is obtained based on the verification complexity and correlation strength of the core conflict components and secondary conflict components. The specific steps include:

[0100] The verification complexity coefficient of each conflicting component is determined by assessing its code size, logical complexity, and number of interfaces.

[0101] Calculate the correlation strength value between each conflicting component and the compatibility conflict feature vector;

[0102] The component localization efficiency coefficient is calculated by using the conflict propagation weights of the core conflict component and the secondary conflict component as the base weights, combined with the verification complexity coefficient and the association strength value.

[0103] First, the system determines the verification complexity coefficient of each conflicting component by analyzing its code size, logical complexity, and number of interfaces. Code size can be measured by lines of code, logical complexity can be evaluated using cyclomatic complexity metrics, and the number of interfaces is the total number of API calls the component provides. These three factors together determine the time and resource costs required to verify the component. For example, graphics rendering components typically have large code sizes, many logical branches, and a large number of interfaces, resulting in a higher verification complexity coefficient; while simple log output components have small code sizes, simple logic, and few interfaces, resulting in a lower verification complexity coefficient. The verification complexity coefficient can be expressed as: Where L is the code size, M is the logical complexity, I is the number of interfaces, and a, b, c are preset weight coefficients that can be adjusted according to the application scenario.

[0104] Next, the association strength value between each conflicting component and the compatibility conflict feature vector is calculated. The component's own features are matched with the extracted conflict feature vector to quantify the degree of association. For example, if the conflict feature vector contains features such as rendering stuttering or frame rate fluctuations, the association strength value between the graphics rendering component and this vector will be significantly higher than that of other unrelated components. The formula for calculating the association strength value can use cosine similarity, i.e.: , where Vc is the component feature vector, Vf is the compatibility conflict feature vector, and ρ ranges from 0 to 1. The larger the value, the closer the association between the component and the conflict.

[0105] Finally, the system uses the conflict propagation weight W of the core conflict component and the secondary conflict component as the basic weight, combined with the verification complexity coefficient C and the association strength value ρ, to calculate the component localization efficiency coefficient. The formula for calculating this coefficient is: This reflects the core logic of efficiency: the higher the conflict propagation weight, the stronger the correlation with conflict features, and the lower the verification complexity, the higher the component's localization efficiency coefficient. For example, assuming a graphics-driven adaptation component has a conflict propagation weight W=2.6, a correlation strength value ρ=0.85 with the conflict feature vector, and a verification complexity coefficient C=1.2, its localization efficiency coefficient is E≈1.84. Another UI component has a conflict propagation weight W=1.0, a correlation strength value ρ=0.3, and a verification complexity coefficient C=0.5, resulting in a localization efficiency coefficient E=1.0×0.5×0.3=0.6. By comparing the efficiency coefficients of the two, the graphics-driven adaptation component is prioritized for deep verification. Because the graphics-driven adaptation component can locate core conflicts with a wider impact range at a relatively controllable verification cost, it significantly improves the overall efficiency and accuracy of conflict localization.

[0106] Based on the collision location confidence coefficient and component location efficiency coefficient, automatic collision location is performed in computer graphics applications, specifically including the following steps:

[0107] If the conflict location confidence coefficient is greater than or equal to the component location efficiency coefficient, then the core conflict component and the secondary conflict component are simultaneously verified and investigated according to the key verification dimensions until the root cause of the conflict is located.

[0108] If the conflict location confidence coefficient is less than the component location efficiency coefficient, the core conflict component will be verified and investigated first according to the key verification dimensions. After that, the secondary conflict components will be verified step by step. If the root cause of the conflict is located in the middle, the subsequent verification process will be terminated.

[0109] First, the calculated conflict location confidence coefficient and component location efficiency coefficient are compared numerically. The relative magnitude of these two coefficients directly determines the subsequent verification and troubleshooting strategy. The conflict location confidence coefficient reflects the balance between the reliability of the location result and resource consumption under the current verification dimension; the component location efficiency coefficient reflects the cost-effectiveness of a single conflicting component in the location process, that is, the comprehensive performance of propagation weight, correlation strength and verification complexity.

[0110] When the conflict location confidence coefficient is greater than or equal to the component location efficiency coefficient, it indicates that the reliability of the current verification dimension is high and the resource consumption is within a controllable range. The system will then simultaneously conduct verification and investigation on the core conflict components and secondary conflict components according to the key verification dimensions. For example, if a graphics application experiences widespread functional failures while running across platforms, the calculated conflict location confidence coefficient is 0.75, while the component location efficiency coefficient is 0.68. In this case, the system determines that the location confidence is sufficient to support full-scale verification. It will simultaneously investigate the core components of the graphics driver adaptation component and rendering component, as well as the secondary components of the UI component and logging component, according to the full-dimensional verification requirements of the code layer, rendering layer, resource layer, and interaction layer, until the root cause of the conflict is located. For example, it may be found that the entire chain of functional abnormalities is caused by the incompatibility of the dependency library version of the graphics driver adaptation component. The synchronous verification method can quickly cover all potential conflict points with high confidence, making it suitable for scenarios with a wide impact and requiring rapid damage mitigation.

[0111] When the conflict localization confidence coefficient is less than the component localization efficiency coefficient, it indicates that the reliability of the current verification dimension is insufficient, or resource consumption is approaching the threshold. The system will adopt a layered verification strategy, first focusing on verifying and investigating the core conflict components according to key verification dimensions, and then gradually verifying the secondary conflict components. If the root cause of the conflict is located midway, the subsequent verification process will be terminated. For example, a graphics application may only experience slight frame rate fluctuations in complex scenes, with a calculated conflict localization confidence coefficient of 0.42 and a component localization efficiency coefficient of 0.55. In this case, the system will prioritize focusing on the core conflict components of the graphics rendering component and the special effects component, focusing on key verification dimensions of the rendering layer and the interaction layer, such as analyzing the pixel-level data of the rendering output and statistically analyzing the frame rate fluctuation range. If, during the verification of the graphics rendering component, it is found that the frame rate fluctuation is caused by an optimization defect in its shader code under a specific GPU architecture, the system will immediately terminate the subsequent verification of the secondary components to avoid unnecessary resource consumption. Layered verification can efficiently locate core conflicts under low confidence, making it suitable for scenarios with a small impact range and where verification costs need to be controlled. By dynamically adjusting the verification strategy, it is possible to balance positioning accuracy and verification efficiency in different conflict scenarios, ultimately achieving automated and intelligent positioning of compatibility conflicts.

[0112] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for automatically locating compatibility conflicts in computer graphics applications, characterized in that, The method includes the following steps: Acquire conflict-related multi-source data for computer graphics applications; the conflict-related multi-source data includes code compilation and adaptation data, graphics rendering output data, system resource scheduling data, and cross-platform interaction log data; Anomaly detection and feature extraction are performed on conflict-related multi-source data to obtain conflict feature vectors. Conflict type labels are matched based on the conflict feature vectors, and conflict impact assessment parameters are generated based on the conflict type labels. The conflict impact assessment parameters include comprehensive quantitative indicators of the impact of the conflict on application performance, functional implementation, and user experience. Based on the conflict impact assessment parameters and the preset positioning requirements, the key verification dimensions for conflict positioning are determined, and the resource consumption of the key verification dimensions is judged to obtain the conflict positioning confidence coefficient. Based on the component dependencies of computer graphics applications, conflict-related components are prioritized to obtain core conflict components and secondary conflict components. The component localization efficiency coefficient is obtained based on the verification complexity and association strength of the core conflict components and secondary conflict components. Automatic collision localization is performed in computer graphics applications based on the collision localization confidence coefficient and the component localization efficiency coefficient.

2. The method for automatically locating compatibility conflicts in computer graphics applications according to claim 1, characterized in that, The process of anomaly detection and feature extraction from conflict-related multi-source data to obtain conflict feature vectors includes the following steps: The code compilation adaptation data is subjected to syntax judgment and compatibility rule verification, and compilation layer features are extracted, including compilation error codes, incompatible syntax structures, and dependency library version conflicts. Perform pixel-level comparison and rendering quality evaluation on the graphics rendering output data, and extract rendering layer features, including image distortion parameters, frame rate fluctuation range, and rendering stutter duration. Real-time monitoring and statistical analysis of system resource scheduling data are performed to extract resource layer features, including peak CPU usage, memory leaks, and GPU computing power allocation ratio. Log parsing and semantic mining are performed on cross-platform interaction log data to extract interaction layer features, including interaction protocol incompatibility identifiers, data transmission failure records, and platform response anomaly information. The extracted features from the compilation layer, rendering layer, resource layer, and interaction layer are combined to form a conflict feature vector.

3. The method for automatically locating compatibility conflicts in computer graphics applications according to claim 2, characterized in that, Matching conflict type labels to conflict feature vectors and generating conflict impact assessment parameters based on these labels involves the following steps: A pre-defined conflict type tag library is provided, which contains feature vector templates and influence weight configurations corresponding to various compatibility conflicts. Each conflict type tag is associated with an influence evaluation dimension. Calculate the similarity between the conflict feature vector and each feature vector template in the tag library, and mark the conflict type label corresponding to the feature vector template with the highest similarity as the target label; The conflict impact assessment parameters are calculated based on the impact assessment dimensions corresponding to the target labels and the quantified values ​​of each feature in the conflict feature vector.

4. The method for automatically locating compatibility conflicts in computer graphics applications according to claim 3, characterized in that, Based on the conflict impact assessment parameters and the preset positioning requirements, the key verification dimensions for conflict positioning are determined, specifically including the following steps: The impact values ​​of operational performance, functional implementation, and user experience in the conflict impact assessment parameters are used as impact factors, and the comprehensive impact value is obtained by calculating the impact factors. If the overall impact value is greater than or equal to the preset high impact threshold, the key verification dimensions include full-dimensional verification of the code layer, rendering layer, resource layer, and interaction layer. If the overall impact score is less than the preset high impact threshold, but greater than or equal to the preset medium impact threshold, then the key verification dimensions include the verification of the rendering layer and interaction layer corresponding to the core function. If the overall impact score is less than the preset impact threshold, then only the resource layer verification related to runtime performance will be retained as the key verification dimension. Based on the positioning requirements, candidate verification dimensions that meet the reliability threshold are selected, and these candidate verification dimensions are combined to form key verification dimensions.

5. The method for automatically locating compatibility conflicts in computer graphics applications according to claim 4, characterized in that, The resource consumption assessment of key verification dimensions yields the conflict location confidence coefficient, specifically including the following steps: The resource consumption parameters corresponding to the key verification dimensions are calculated. The resource consumption parameters include the computing power resources, time costs, and storage resource usage required for the verification process. Based on the resource loss parameters and the preset maximum resource loss threshold, calculate the resource loss adaptation rate. The conflict location confidence coefficient is calculated by combining the comprehensive impact value of the conflict impact assessment parameters. The conflict location confidence coefficient is positively correlated with the comprehensive impact value and negatively correlated with the resource loss adaptation rate.

6. The method for automatically locating compatibility conflicts in computer graphics applications according to claim 5, characterized in that, Based on the component dependencies of computer graphics applications, conflict-related components are prioritized to obtain core conflict components and secondary conflict components. This process includes the following steps: By using component dependency identification tools, identify all functional components of the application and the dependencies between them, including direct and indirect dependencies; Calculate the conflict propagation weight of each component, whereby the conflict propagation weight is a quantified value of the scope and degree of influence of the component on other components when a conflict occurs. Set a weight threshold, mark components with conflict propagation weights greater than or equal to the threshold as core conflict components, and mark components with conflict propagation weights less than the threshold as secondary conflict components.

7. The method for automatically locating compatibility conflicts in computer graphics applications according to claim 6, characterized in that, The component localization efficiency coefficient is obtained based on the verification complexity and correlation strength of the core conflict components and secondary conflict components. The specific steps include: The verification complexity coefficient of each conflicting component is determined by assessing its code size, logical complexity, and number of interfaces. Calculate the correlation strength value between each conflicting component and the compatibility conflict feature vector; The component localization efficiency coefficient is calculated by using the conflict propagation weights of the core conflict component and the secondary conflict component as the base weights, combined with the verification complexity coefficient and the association strength value.

8. The method for automatically locating compatibility conflicts in computer graphics applications according to claim 7, characterized in that, Based on the collision location confidence coefficient and component location efficiency coefficient, automatic collision location is performed in computer graphics applications, specifically including the following steps: If the conflict location confidence coefficient is greater than or equal to the component location efficiency coefficient, then the core conflict component and the secondary conflict component are simultaneously verified and investigated according to the key verification dimensions until the root cause of the conflict is located. If the conflict location confidence coefficient is less than the component location efficiency coefficient, the core conflict component will be verified and investigated first according to the key verification dimensions. After that, the secondary conflict components will be verified step by step. If the root cause of the conflict is located in the middle, the subsequent verification process will be terminated.

Images