PROFGEN: a procedure for generating machine independent high-level language profilers

  • Authors:
  • Paul A. Fishwick;Stefan Feyock

  • Affiliations:
  • University of Pennsylvania Moore Bldg./D2, Philadelphia, PA;College of William and Mary Williamsburg, VA

  • Venue:
  • ACM SIGMETRICS Performance Evaluation Review
  • Year:
  • 1984

Quantified Score

Hi-index 0.00

Visualization

Abstract

Many of the tools used in software metrics for evaluating the execution characteristics of a program are predicated on specific counting rules for operators and operands [1, 2]. The analyst may use these counting techniques to determine such program attributes as estimation of object code size prior to actual compilation and the relative efficiencies of various language compilers. Operator/operand measures provide useful results for certain analyses, but a deficiency exists in that the data derived from this technique does not directly reflect the program structure afforded by a high-level language such as FORTRAN, Pascal, or Ada. There are many instances where it is desirable to measure the program at the source level where the execution data may be directly associated with specific high level program units such as source statements and blocks.