Embodiments provide tools and techniques for clustering failing runs in a design verification environment to aid in determining causes of the failing runs. Embodiments may include determining multiple failing runs of the design verification environment. Multiple partitions of the multiple failing ru
Embodiments provide tools and techniques for clustering failing runs in a design verification environment to aid in determining causes of the failing runs. Embodiments may include determining multiple failing runs of the design verification environment. Multiple partitions of the multiple failing runs may be generated. Each respective partition may partition one or more subsets of the multiple failing runs into one or more non-overlapping clusters of failing runs. The multiple partitions of the subsets of multiple failing runs may be merged into a hierarchical structure that includes at least one of the clusters. One or more clusters of failing runs from the merged hierarchical structure may be selected; these may be referred to as core clusters. Core clusters may be presented to a user based on the size and distance between the clusters.
대표청구항▼
1. A method of clustering failing runs in a design verification environment to aid in determining causes of the failing runs, the method comprising: determining a plurality of failing runs of the design verification environment;generating a plurality of partitions of the plurality of failing runs, e
1. A method of clustering failing runs in a design verification environment to aid in determining causes of the failing runs, the method comprising: determining a plurality of failing runs of the design verification environment;generating a plurality of partitions of the plurality of failing runs, each respective partition partitions one or more subsets of the plurality of failing runs into one or more non-overlapping clusters of failing runs, wherein each core cluster of the one or more non-overlapping clusters represents a different bug;merging the plurality of partitions of the plurality of failing runs into a hierarchical structure that includes at least one of the clusters; andselecting one or more clusters of failing runs from the merged hierarchical structure. 2. The method of clustering failing runs in a design verification environment to aid in determining causes of the failing runs of claim 1, wherein the one or more clusters is each represented by one or more attributes. 3. The method of clustering failing runs in a design verification environment to aid in determining causes of the failing runs of claim 2 further comprising: determining one or more grades for each respective cluster based on their respective attributes. 4. The method of clustering failing runs in a design verification environment to aid in determining causes of the failing runs of claim 3, wherein the one or more grades includes at least an external grade, an internal grade, or a total grade, where an external grade represents a measures of a union of all attributes that correlate with a failure, the internal grade represents a measure of whether the cluster is consistent, and the total grade represents a measure of the internal and external grade. 5. The method of clustering failing runs in a design verification environment to aid in determining causes of the failing runs of claim 1, further comprising: separating the plurality of failing runs into a plurality of failure classes and wherein generating the plurality of partitions of failure runs includes partitioning into the separate classes. 6. The method of clustering failing runs in a design verification environment to aid in determining causes of the failing runs of claim 1 further comprising: receiving user input to determine one or more of the plurality of partitions. 7. The method of clustering failing runs in a design verification environment to aid in determining causes of the failing runs of claim 6, wherein the receiving user input includes receiving coverage definitions. 8. The method of clustering failing runs in a design verification environment to aid in determining causes of the failing runs of claim 1 further comprising: presenting the selected one or more clusters to a user. 9. The method of clustering failing runs in a design verification environment to aid in determining causes of the failing runs of claim 8, wherein presenting the selected one or more clusters to users includes presenting the identified clusters in specific order. 10. The method of clustering failing runs in a design verification environment to aid in determining causes of the failing runs of claim 9, wherein the specific order comprises an order based on a distance metric between clusters. 11. The method of clustering failing runs in a design verification environment to aid in determining causes of the failing runs of claim 10, wherein a first cluster in the specific order is a largest cluster from the merged hierarchy. 12. A system for clustering failing runs in a design verification environment to aid in determining causes of the failing runs, the system comprising: a storage medium;a processor coupled with the storage medium, the processor configured for clustering failing runs of the design verification environment by a configuration to: determine a plurality of failing runs of the design verification environment;generate a plurality of partitions of the plurality of failing runs, each respective partition partitions one or more subsets of the plurality of failing runs into one or more non-overlapping clusters of failing runs, wherein each core cluster of the one or more non-overlapping clusters represents a different bug;merge the plurality of partitions of the plurality of failing runs into a hierarchical structure that includes at least one of the clusters; andselect one or more clusters of failing runs from the merged hierarchical structure. 13. The system for clustering failing runs in a design verification environment to aid in determining causes of the failing runs of claim 12, wherein the processor configured for clustering failing runs of the design verification environment by a further configuration to: separate the plurality of failing runs into a plurality of failure classes and wherein generating the plurality of partitions of failure runs includes partitioning into the separate classes. 14. The system for clustering failing runs in a design verification environment to aid in determining causes of the failing runs of claim 12, wherein the processor configured for clustering failing runs of the design verification environment by a further configuration to: receive user input to determine one or more of the plurality of partitions. 15. The system for clustering failing runs in a design verification environment to aid in determining causes of the failing runs of claim 14, wherein the receiving user input includes receiving coverage definitions. 16. The system for clustering failing runs in a design verification environment to aid in determining causes of the failing runs of claim 12, wherein the processor configured for clustering failing runs of the design verification environment by a further configuration to: presenting the selected one or more clusters to a user. 17. The system for clustering failing runs in a design verification environment to aid in determining causes of the failing runs of claim 16, wherein presenting the selected one or more clusters to users includes presenting the identified clusters in specific order. 18. The system for clustering failing runs in a design verification environment to aid in determining causes of the failing runs of claim 17, wherein the specific order comprises an order based on a distance metric between clusters. 19. The system for clustering failing runs in a design verification environment to aid in determining causes of the failing runs of claim 18, wherein a first cluster in the specific order is a largest cluster from the merged hierarchy. 20. A non-transitory machine-readable storage medium containing program instructions that, when executed by a controller within a computer, cause the controller to execute a method of clustering failing runs in a design verification environment to aid in determining causes of the failing runs, the method comprising: determining a plurality of failing runs of the design verification environment;generating a plurality of partitions of the plurality of failing runs, each respective partition partitions one or more subsets of the plurality of failing runs into one or more non-overlapping clusters of failing runs, wherein each core cluster of the one or more non-overlapping clusters represents a different bug;merging the plurality of partitions of the plurality of failing runs into a hierarchical structure that includes at least one of the clusters; andselecting one or more clusters of failing runs from the merged hierarchical structure. 21. The non-transitory machine-readable storage medium of claim 20, wherein the one or more clusters is each represented by one or more attributes. 22. The non-transitory machine-readable storage medium of claim of claim 21 further comprising: determining one or more grades for each respective cluster based on their respective attributes. 23. The non-transitory machine-readable storage medium of claim of claim 22, wherein the one or more grades includes at least an external grade, an internal grade, or a total grade, where an external grade represents a measures of a union of all attributes that correlate with a failure, the internal grade represents a measure of whether the cluster is consistent, and the total grade represents a measure of the internal and external grade. 24. The non-transitory machine-readable storage medium of claim of claim 20 further comprising: receiving user input to determine one or more of the plurality of partitions. 25. The non-transitory machine-readable storage medium of claim 20 further comprising: presenting the selected one or more clusters to a user. 26. The non-transitory machine-readable storage medium of claim of claim 25, wherein presenting the selected one or more clusters to users includes presenting the identified clusters in specific order. 27. The non-transitory machine-readable storage medium of claim of claim 26, wherein the specific order comprises an order based on a distance metric between clusters.
연구과제 타임라인
LOADING...
LOADING...
LOADING...
LOADING...
LOADING...
이 특허에 인용된 특허 (6)
Schmidt Dennis E. ; Hanus Michael D., APU troubleshooting system.
Park,Yong Cheol, Recording medium with overlapping segment information thereon and apparatus and methods for forming, recording, and reproducing the recording medium.
Segal, Nili; Kinderman, Yael; Gupta, Hemant; Oren, Oded, Method and system for automated refined exclusion of entities from a metric driven verification analysis score.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.