System for monitoring compute system performance

• US 5,732,273 A
• Filed: 01/03/1997
• Issued: 03/24/1998
• Est. Priority Date: 08/11/1995
• Status: Expired due to Term

• Current Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Original Assignee: Digital Equipment Corporation (HP Inc.)
• Inventors: Srivastava, Amitabh, Eustace, Robert Alan
• Primary Examiner(s): Kriess, Kevin A.
• Assistant Examiner(s): Chavis, John Q.
• Application Number:
  US08/778,648
• Time in Patent Office: 445 Days
• Field of Search: 395/704, 395/183.14
• US Class Current: 717/128
• CPC Class Codes: G06F 11/3466 Performance evaluation by tracing or monitoring
• CPC Class Codes: G06F 11/3612 by runtime analysis performance monitoring G06F11/3466
• CPC Class Codes: G06F 2201/865 Monitoring of software

Abstract

A program for monitoring computer system performance includes a collection of source code modules in the form of a high level language. Each of the source code modules is compiled into a corresponding object code module. The object code modules are assembled into machine dependent code. The machine code is translated into a program module in the form of a machine independent register translation language. The program module is partitioned into basic program components. The basic program components include procedures, basic blocks within procedures, and instructions within basic blocks. Fundamental instrumentation routines identify, locate, and modify specific program components to be monitored. The modified basic program components are converted to an instrumented machine executable code to be executed in the computer system so that performance dam can be collected while the program is executing in the computer.

Claims

1. A computer implemented method for monitoring the performance of a computer system while the computer system is executing a program, the program in the form of a machine dependent executable instructions, comprising the steps of:

translating the machine dependent executable instructions into a program module in the form of an intermediate language, the intermediate language having instructions independent of a particular computer system hardware architecture;

partitioning the program module into a plurality of basic program components;

tracing an execution flow through the plurality of basic program components using navigational routines;

while tracing the execution flow through the plurality of basic program components, inserting a call instruction in a specific program component, the call instruction to transfer execution control to a user analysis routine when the specific program component is executed in the computer system;

generating an instrumented machine dependent executable code from the modified plurality of basic program components after inserting the call instruction;

executing the instrumented machine dependent executable code in the computer system; and

collecting performance data of the computer system by the user analysis routine.

2. The method as in claim 1 further comprising:

partitioning the plurality of basic program components module into procedures;

partitioning the procedures into basic blocks, wherein the basic blocks include instructions.

3. The method as in claim 2 further comprising:

constructing, in a memory of the computer system, a procedure flow graph for each procedure of the plurality of basic program components using an organizer;

constructing, in the memory, a program call graph for the module, the procedure flow graphs indicating the execution flow within each of the procedures, the program call graph to indicate the execution flow among the procedures; and

tracing the execution flow through the plurality of basic program components using the procedure flow graphs and the program control graph and instrumentation routines.

4. The method as in claim 3 further comprising:

parsing the instructions of the specific program component of the linked code module using parsing routines, the parsing routines parsing the instructions into operator and operand fields, the operator and operand fields to locate the specific program component.

5. The method as in claim 1 further comprising:

combining the user analysis routine with the plurality of basic program components.

6. The method of claim 1 a first hardware architecture is associated with the machine dependent executable instructions, and a second hardware architecture is associated with the instrumented machine executable code, the first and second hardware architectures being different.

7. A system for monitoring the performance of a computer system while the computer system is executing a program, the program including a plurality of source code modules in the form of a high level language, comprising the means for:

a compiler for compiling the plurality of source code module into a plurality of object code modules;

a linker for producing machine dependent executable code from the plurality of object code modules;

means for translating the plurality of object code modules into a program module in the form of an intermediate language, the intermediate language having instructions independent of a particular computer system hardware architectures;

means for partitioning the program module into a plurality of basic program components;

navigational routines for tracing the execution flow through the plurality of basic program components prior to execution to locate a specific program component;

an instrumentor to insert a call instruction in the specific program component, the call instruction to transfer execution;

control to a user analysis routine when the specific program component is executed in the computer;

a code generator for generating instrumented machine executable code from the modified plurality of basic program components;

a processor for executing the instrumented machine executable code; and

the user analysis routine collecting performance data while the processor is executing the instrumented machine executable code.

8. A computer implemented method for collecting performance data of a computer system, the performance data including data states and execution states of the computer system, comprising:

translating a machine dependent executable code produced by a code generator into machine independent code;

instrumenting the machine independent code to insert a call instruction in the machine independent code, the call instruction to reference a user analysis routine;

generating an instrumented machine dependent executable code from the machine independent code including the call instruction;

executing the instrumented machine dependent executable code in a processor of the computer system; and

collecting the performance data of the computer system by the user analysis routine while executing the instrumented machine dependent executable code.

9. The method of claim 8 further comprising:

constructing a procedure flow graph of procedures of the machine independent code in a memory of the computer system;

constructing a program call graph of the machine independent code in the memory; and

tracing an execution flow of the machine independent code using the procedure flow graph and the program call graph and navigational routines operating on the machine independent code.

10. The method of claim 9 further comprising:

while tracing the execution flow of the machine independent code, identifying a program component of the machine independent code to be monitored; and

modifying the program component to insert the call instruction to the user analysis routine.

11. The method of claim 8 wherein the machine dependent executable machine code and the instrumented machine dependent executable code are for different hardware architectures.