Methods, Systems, and Media for Generating a Regression Suite Database

• US 20080065687A1
• Filed: 11/10/2007
• Published: 03/13/2008
• Est. Priority Date: 06/09/2004
• Status: Active Grant

• Current Assignee: Meta Platforms, Inc. (f/k/a Facebook, Inc.)
• Original Assignee: Amir Hekmatpour, James Coulter
• Inventors: Coulter, James, Hekmatpour, Amir
• Granted Patent: US 7,512,839 B2
• Time in Patent Office: Days
• US Class Current: 707/102
• CPC Class Codes: G06F 11/261 by simulating additional hardware, e.g. fault simulation
• CPC Class Codes: G06F 11/263 Generation of test inputs, e.g. test vectors, patterns or sequences ; with adaptation of the tested hardware for testability with external testers
• CPC Class Codes: G06F 11/3688 for test execution, e.g. scheduling of test suites

Abstract

Methods, systems and media for generating an improved regression suite by applying harvesting models and/or regression algorithms to tests utilized in verification of a system are disclosed. In one embodiment, a regression manager responsive to user input may be coupled to a harvester module, an analysis module, and a management module. In one embodiment, the harvester module is responsive to harvesting models defined in a modeling language, where the harvester module is coupled to a regression suite database. In another embodiment, a regression methodology may be defined from a collection of regression strategies and each regression strategy may be defined from a combination of harvesting models and/or regression algorithms. A regression generator to receive tests, to apply one or more regression strategies to the tests, to provide reports, and to allow user control may also be provided.

Claims

1-31. -31. (canceled)

32. A regression suite database generation system comprising:

a regression manager responsive to user input;

a regression suite database coupled to the regression manager to store a plurality of tests;

a harvest module to update the regression suite database by applying harvesting models to the plurality of tests before at least one regression strategy is applied, wherein the harvest module utilizes a modeling language for defining harvesting models;

an extractor module coupled to the harvest module to receive simulation logs and extract information from the simulation logs for transmittal to the harvest module;

a regression generator, coupled to the regression manager and the regression suite database, to receive a plurality of tests and apply a plurality of regression strategies to the received tests, the plurality of regression strategies each including one or more harvesting models and one or more regression algorithms to be applied to tests;

wherein the regression generator performs regression optimization by determining whether new test results should be saved in the regression suite database and by analyzing received tests and eliminating earlier tests from the regression suite database if they are supplanted by later received tests; and

wherein further the regression generator implements the plurality of regression strategies of scheduled harvesting models and regression algorithms based on predefined regression strategy settings and utilized regression resources.

33. The system of claim 32, wherein the predefined regression strategy settings include a predefined schedule.

34. The system of claim 32, wherein the predefined regression strategy settings include verification environment conditions and attributes

35. The system of claim 32, wherein the regression manager is a Web-based interface, the regression manager being adapted to allow a user to define, monitor, and control regression strategies, and wherein further the regression manager is adapted to provide feedback on the regression to the user.

36. A method for applying a regression strategy to tests, the method comprising:

receiving a plurality of tests;

deleting any of the received tests that already exist in a regression suite database;

determining implementation of one or more harvesting models and one or more regression algorithms based on a plurality of regression strategies that are each based oil predefined regression strategy settings and utilized regression resources;

applying the determined one or more harvesting models to the received tests to remove any tests that are eliminated by an applied harvesting model;

applying the determined one or more regression algorithms to tests remaining after application of the one or more harvesting models to selectively choose one or more selected texts based on application of the one or more regression algorithms; and

saving the selected tests.

37. The method of claim 36, wherein the predefined regression strategy settings include a predefined schedule.

38. The method of claim 36, wherein the predefined regression strategy settings include verification environment conditions and attributes.

39. The method of claim 36, further comprising defining a schedule of applying regression strategies, and further comprising defining conditions under which to start, stop or choose each regression strategy.

40. The method of claim 36, further comprising selecting the regression strategies from a group of regression strategies.

41. The method of claim 36, wherein different strategies of the plurality of r egression strategies are applied to a design addressing different units of the design.

42. The method of claim 36, wherein different strategies of the plurality of regression strategies are applied to a design based oil a tool or method used to generate the tests.

43. The method of claim 36, further comprising selecting one of multiple versions of a regression algorithm for a regression strategy based on characteristics of the plurality of tests.

44. The method of claim 36, further comprising selecting one of multiple versions of a regression algorithm for a regression strategy based on coverage density.

45. The method of claim 36, further comprising deleting any of the received tests that are not valid tests.

46. A computer readable medium containing a program which, when executed, performs an operation, comprising:

receiving a plurality of tests;

deleting any of the received tests that already exist in a regression suite database;

determining implementation of one or more harvesting models sand one or more regression algorithms based on a plurality of regression strategies that are each based on predefined regression strategy settings and utilized regression resources;

applying the determined one or more harvesting models to the received tests to remove any tests that are eliminated by an applied harvesting model;

applying the determined one or more regression algorithms to tests remaining after application of the one or more harvesting models to selectively choose one or more selected texts based on application of the one or more regression algorithms; and

saving the selected tests.

47. The computer readable medium of claim 46, further comprising defining a schedule of applying regression strategies, and further comprising defining conditions tinder which to start, stop or choose each regression strategy.

48. The computer readable medium of claim 46, further comprising selecting the regression strategies from a group of regression strategies.