Abstract: An automated method for determining problematic binary files. A group of source code files is compiled using two different compiling environments. One of the compiling environments yields a failing set of binary files (i.e., the resultant executable program exhibits problematic behavior). The other compiling environment yields a passing set of binary files (i.e., the resultant executable program does not exhibit problematic behavior). Complementary sets of files, containing some passing files and some failing files, are used to create test executable programs. The test executable programs are evaluated against pass/fail criteria. The results of these evaluations are used to isolate the problematic binary file or set of files. The process is continued iteratively until the problematic file is determined. According to one embodiment the two compiling environments may be different modes of a given compiler.

Abstract: An automated method for determining problematic binary files. A group of source code files is compiled using two different compiling environments. One of the compiling environments yields a failing set of binary files (i.e., the resultant executable program exhibits problematic behavior). The other compiling environment yields a passing set of binary files (i.e., the resultant executable program does not exhibit problematic behavior). Complementary sets of files, containing some passing files and some failing files, are used to create test executable programs. The test executable programs are evaluated against pass/fail criteria. The results of these evaluations are used to isolate the problematic binary file or set of files. The process is continued iteratively until the problematic file is determined. According to one embodiment the two compiling environments may be different modes of a given compiler.