Tracking converage results in a batch simulation farm network

• US 7,359,847 B2
• Filed: 11/30/2001
• Issued: 04/15/2008
• Est. Priority Date: 11/30/2001
• Status: Active Grant

• Current Assignee: Meta Platforms, Inc. (f/k/a Facebook, Inc.)
• Original Assignee: International Business Machines Corporation
• Inventors: Roesner, Wolfgang, Williams, Derek Edward, Gabele, Carol Ivash
• Primary Examiner(s): Shah; Kamini
• Assistant Examiner(s): Silver; David
• Application Number:
  US09/997,460
• Publication Number:
  US 20030135354A1
• Time in Patent Office: 2,328 Days
• Field of Search: 703/21, 703/16, 703/17, 716/4, 716/5
• US Class Current: 703/17
• CPC Class Codes: G06F 30/33 Design verification, e.g. functional simulation or model checking

Abstract

A method and system for providing centralized access to count event information from testing of a hardware simulation model within a batch simulation farm which includes simulation clients and an instrumentation server. Count event data for said hardware simulation model is received by the instrumentation server from one or more simulation clients. A first and a second counter report are generated for the hardware simulation model, in which the first and second counter reports are derived from the count event data received by the instrumentation server. The first counter report is compared to the second counter report, and responsive to this comparison, a counter difference report is generated within the instrumentation server that conveys count event trends associated with the simulation model under different simulation testcases.

Claims

1. In a hardware description language (HDL) batch simulation farm having multiple simulation clients coupled to an instrumentation server, a method for providing centralized access to trends in count event data, wherein the count event data represents sequences of signal values that indicate the occurrence of events triggered during simulation testing of HDL models by the simulation clients, said method comprising:

utilizing said instrumentation server to;

receive a first set of count event data for a first simulation test of an HDL model;

generate a first counter report from the first set of count event data, wherein the first counter report specifies a number of occurrences of one or more count events for the first simulation test and further specifies a number of simulation cycles over which the first simulation test was processed;

receive a second set of count event data for a second simulation test of the HDL model;

generate a second counter report from the second set of count event data, wherein the second counter report specifies a number of occurrences of one or more count events for the second simulation test and further specifies a number of simulation cycles over which the second simulation test was processed;

compare said first counter report to said second counter report to detect variations in rates of occurrences of count events recorded in the first and second counter reports, said comparing including;

utilizing the specified number of simulation cycles specified by said first counter report and the specified number of simulation cycles specified by the second counter report to normalize the number of count event occurrences specified by said first counter report with respect to the number of count event occurrences specified by said second counter report; and

determining the difference in the normalized numbers of occurrences of corresponding count events specified by said first counter report and said second counter report; and

generate a counter difference report that specifies one or more count events for which the determined difference in the normalized numbers of occurrences of corresponding count events exceeds a pre-specified difference threshold.

2. The method of claim 1, further comprising:

executing a testcase using said HDL simulation model within said one or more simulation clients;

receiving an aggregate count event packet from said one or more simulation clients, wherein said aggregate count event packet includes count event data recorded during said testcase; and

within said instrumentation server, storing said count event data within count data storage files.

3. The method of claim 2, wherein said first and second counter reports are generated as output from count event queries processed with respect to said count data storage files.

4. The method of claim 2, wherein said first and second counter reports are generated directly from said counter data storage files.

5. The method of claim 1, wherein said first and second counter reports each include a simulator cycle count value that specifies the number of simulation cycles over which simulation testing was processed, said normalizing the number of count event occurrences specified by said first counter report with respect to the number of count event occurrences specified by said second counter report further comprising, computing a count normalization factor that is a ratio of the values of the simulator cycle count values contained in said first and second counter reports.