« PreviousContinue »
PROBLEMATIC BINARY ISOLATOR
FIELD OF THE INVENTION
The present invention relates generally to computer program compiling and debugging, and more specifically to an automated system for determining the file or set of files that cause a program run-time failure and/or performance problem when complied with a particular compiler, or compiled in a particular manner.
The need for computer programs with greater functionality is increasing dramatically and with added functionality of computer programs comes added complexity. It is not uncommon today for a computer program to contain hundreds or even thousands of source code files. Computer program compilers are used to transform these source code files to machine-readable files (e.g., executable files). There are many varieties of compilers available for different languages and from different manufacturers. Often times a computer programmer may wish to take a set of source code files already compiled on one type of compiler and compile the same source code files on another interoperable compiler. This may be desirable to realize the benefit of some advantage which one compiler has over another, for example greater speed or additional optimization features.
A computer programmer may find that the set of source code files that compiles and runs successfully using one compiler may not execute properly when compiled with another one. That is, the executable program produced using one compiler runs fine and the executable program produced using another compiler crashes for the same set of source code files. Alternatively, the programmer may find that the set of source files compiles using one set of compilation options and does not compile with a different set of options. Typically the problem is caused by one source code file or the interaction of two or more source code files. The vast majority of the source code files will usually present no difficulty using either of the two interoperable compilers. The problem may be due to the source code itself or some unknown deficiency in the compiler. It may be possible and desirable to compile all of the source code files except the problematic file(s) on a preferred compiler and compile the problematic file(s) using a compiling environment that does not cause a problem.
To do this a computer programmer may need to determine which source code file or set of files is problematic. This process typically involves creating test executable programs made up of some source code files compiled using the functioning compiling environment and some source code files compiled using the problematic compiling environment and determining if the produced executable file runs properly. Through iterations of this process the computer programmer may be able to isolate the problematic source code file(s). This process requires a great deal of effort on the part of the computer programmer. The computer programmer must decide which set of files will be used to construct the test executable program and must determine if the test executable program experiences a failure or degradation in performance. The computer programmer, typically, may monitor the evaluation process and make decisions for each iteration of the debugging process.
BRIEF DESCRIPTION OF THE DRAWINGS
The present invention is illustrated by way of example and not limited by the figures of the accompanying drawings in which like references indicate similar elements and in which:
FIG. 1 is a diagram illustrating an exemplary digital processing system 100 for implementing the present invention;
FIG. 2 is a process flow diagram according to one 5 embodiment of the present invention;
FIGS. 3A and 3B depict the selection of files and the creation of test executable programs in accordance with an embodiment of the present invention; and 1Q FIG. 4 depicts an example of the file selecting process of one embodiment of the present invention.
According to one aspect of the present invention, a
15 method and system are provided to automate the process of isolating problematic source code files. The methodology that the problematic binary isolator uses to find a run time failure is similar to the way an expert debugger would attack the same problem. Because the tool systematically isolates
20 the problem, it excludes the possibility of human error; thus it is more accurate. According to one embodiment of the present invention a method is provided to ascertain runtime failures caused by the interaction of two or more files. FIG. 1 is a diagram illustrating an exemplary digital
25 processing system 100 for implementing the present invention. The compilation, segregation, and evaluation of source code files described herein can be implemented and utilized within digital processing system 100, which can represent a general-purpose computer, portable computer, or other like
30 device. The components of digital processing system 100 are exemplary in which one or more components can be omitted or added. For example, one or more memory devices can be utilized for digital processing system 100.
35 Referring to FIG. 1, digital processing system 100 includes a central processing unit 102 and a signal processor 103 coupled to a display circuit 105, main memory 104, static memory 106, and mass storage device 107 via bus 101. Digital processing system 100 can also be coupled to a display 121, keypad input 122, cursor control 123, hard copy device 124, input/output (I/O) devices 125, and audio/ speech device 126 via bus 101.
Bus 101 is a standard system bus for communicating information and signals. CPU 102 and signal processor 103
45 are processing units for digital processing system 100. CPU 102 or signal processor 103 or both can be used to process information and/or signals for digital processing system 100. CPU 102 includes a control unit 131, an arithmetic logic unit (ALU) 132, and several registers 133, which are used to
5Q process information and signals. Signal processor 103 can also include similar components as CPU 102.
Main memory 104 can be, e.g., a random access memory (RAM) or some other dynamic storage device, for storing information or instructions (program code), which are used
55 by CPU 102 or signal processor 103. Main memory 104 may store temporary variables or other intermediate information during execution of instructions by CPU 102 or signal processor 103. Static memory 106, can be, e.g., a read only memory (ROM) and/or other static storage devices, for
go storing information or instructions, which can also be used by CPU 102 or signal processor 103. Mass storage device 107 can be, e.g., a hard or floppy disk drive or optical disk drive, for storing information or instructions for digital processing system 100.
65 Display 121 can be, e.g., a cathode ray tube (CRT) or liquid crystal display (LCD). Display device 121 displays information or graphics to a user. Digital processing system
100 can interface with display 121 via display circuit 105. Keypad input 122 is a alphanumeric input device with an analog to digital converter. Cursor control 123 can be, e.g., a mouse, a trackball, or cursor direction keys, for controlling movement of an object on display 121. Hard copy device 5 124 can be, e.g., a laser printer, for printing information on paper, film, or some other like medium. A number of input/output devices 125 can be coupled to digital processing system 100. The automated process of isolating problematic source code files described herein can be imple- 10 mented by hardware and/or software contained within digital processing system 100. For example, CPU 102 or signal processor 103 can execute code or instructions stored in a machine-readable medium, e.g., main memory 104.
The machine-readable medium may include a mechanism ^ that provides (i.e., stores and/or transmits) information in a form readable by a machine such as computer or digital processing device. For example, a machine-readable medium may include a read only memory (ROM), random access memory (RAM), magnetic disk storage media, opti- 2o cal storage media, flash memory devices. The code or instructions can be represented by carrier wave signals, infrared signals, digital signals, and by other like signals.
FIG. 2 is a process flow diagram according to one embodiment of the present invention. Process 200, shown in 25 FIG. 2 begins at operation 205 in which the system determines that one of two interoperable compiling environments produces a passing executable program and one produces a failing executable program. That is, the system determines that a set of source code files can be compiled using a given 30 compiler to produce an executable program that has no runtime failures or performance problems (i.e., the executable program passes). The system also determines that the same source code files, compiled using another compiler or the same compiler with a different compilation mode, pro- 35 duce an executable program that experiences runtime failures or performance problems (i.e., the executable program fails).
At operation 210 the two sets of binary files, obtained through compiling the same source code files using different 40 compiling environments are placed in different directories. The set of binary files that produce an executable program that does not exhibit run-time failure or performance degradation are placed in a passing directory (PD). The set of binary files that produce an executable program that exhibits 45 run-time failure or performance degradation are placed in a failing directory (FD). At operation 215 some files are automatically selected from the PD and the complementary files are automatically selected from the FD. The selected files from each directory are copied to a separate directory 50 called the test directory (TD). The files of the TD are then linked to create a test executable program at operation 220. Depending upon the type of files in the TD, linking may not be necessary. For example, if the files in the TD were dynamic link library (.dll) files, linking may not be neces- 55 sary. At operation 225 the test executable program is run against pass/fail criteria. The output of the test executable program may be compared to model output. For example a text file from the test executable program may be compared to a model text file. A determination is made as to whether 60 the test executable program passes or fails. If the test executable program passes then the system has determined that the files from the FD that make up the test executable may not be problematic. If the test executable program fails, the system has determined that the problematic file(s) is one 65 of the files from the FD that make up the test executable program. The process is then automatically repeated from
operation 215 using different, complementary combinations of files from the passing and failing directories. Through this process it is possible to determine which file or files is causing the run-time failure or performance problems. At operation 230 the process will continue if the state of each file has not been determined. The process ends at operation 235 when the state of each file is known. That is, after repeated iterations the system determines which files are problematic and which are not.
FIGS. 3A and 3B depict the selection of files and the creation of test executable programs in accordance with an embodiment of the present invention. FIG. 3A shows an exemplary set of source code files 302 labeled A through G. Typically, the number of source code files used is hundreds or thousands. In one embodiment source code files 302 are compiled using a first compiling environment. The compiling environment may be an Intel compiler available from Intel Corporation of Sunnyvale, Calif. The compiling of source code files 302 produces a set of object files 304 labeled Aj.obj through Gj.obj. Object files 304 are linked to create a passing executable program 306. Executable program 306 runs without runtime failures or performance problems. The same set of source code files (i.e., source code files 302) when compiled using a second compiling environment, interoperable with the first, produce a set of object files 308 labeled A2.obj through G2.obj. The second compiling environment may be a different compiler, for example, a Microsoft compiler available from Microsoft Corporation of Redmond, Wash., or alternatively could be a different mode of the first compiling environment. Object files 308, when linked, create a failing executable program 310 that exhibits runtime failure or performance degradation. A single source code file, due to a particular compiling environment, may cause such runtime failures or performance problems. Such problems may also be caused by the interaction of two or more source code files. For example, object file 312, labeled E2.obj is the problematic file that has caused executable program 310 to fail.
FIG. 3B shows an object file test set 314 created in accordance with one embodiment of the present invention. In one embodiment the present invention creates an object file test set comprised of object files created through the failing compiler environment and the passing compiler environment as discussed above in reference to FIG. 2. Object file test set 314 contains object files A2.obj, B2.obj, and C2.obj selected from the set of object files 308 that produced failing executable program 310 as well as object files Dj.obj, Ej.obj, Fj.obj, and Gj.obj selected from the set of object files 304 that produced passing executable program 306. Because object file test set 314 does not contain the problematic file E2.obj, the executable program 316 created by linking the files of object file test set 314 does not exhibit the problematic behavior. Because an executable program 316, created by linking a set of object files 314 that included files A2.obj, B2.obj, and C2.obj, did not exhibit problematic behavior, the system has now determined that the set of files A2.obj, B2.obj, and C2.obj does not contain the problematic file. One of these files interacting with other files may be the source of the runtime failure or performance problem, as discussed below, but for purposes of this example, the system has eliminated these files as suspect problematic files.
The system then creates another object file test set 318 that contains additional files from the set of object files 308 (e.g., object files test set 318 contains A2.obj, B2.obj, C2.obj D2.obj, E2.obj, Fj.obj, and Gj.obj). Because object file test set 318 contains the problematic file E2.obj, the executable
program 320 created by linking the files of object file test set 318 exhibits the problematic behavior.
Through an iterative process the system of the present invention may determine the state of all files and thus determine the problematic file(s). 5
In a more complex example, as referred to above, the interaction of two or more files is the source of the runtime failure or performance problem. In a preferred embodiment the present invention employs a search algorithm as described in pseudo code in Appendix A. The algorithm of 10 Appendix A can be employed not only to ascertain a problematic file, but also to ascertain the problematic interaction of two or more files. The caller of the search algorithm enumerates the suspect files, which may be all of the files. The caller creates two Boolean arrays of size N, where 15 N is the number of suspect files. One array is the current state array which is used internally by the search algorithm to represent the current configuration of binary files that is being tested. The other array is the answer array which is initialized to passing. As each problematic file is found by 20 the search algorithm, the corresponding element of the answer array is changed to failing. Upon completion of the search algorithm, each failing element of the answer array represents a problematic binary file.
FIG. 4 depicts an example of the process of selecting files employing the algorithm of Appendix A. A file set 402, created by compiling source code files using a "passing" compiling environment exhibits no problematic behavior (i.e., it passes). File set 402 contains files Ft through P8. A 3Q file set 404 created from the same source code using a "failing" compiling environment exhibits problematic behavior such as runtime failure or performance problems (i.e., it fails). File set 404 contains files F1 through F8. For purposes of example, the problematic behavior of file set 3J 404 is caused by the interaction of files F6 and F7 when compiled using the "failing" compiling environment. In other words, the resulting executable program exhibits problematic behavior, if, and only if, both files F6 and F7 are present in the file set. In a first iteration, the system creates 4Q a test file set 408 consisting of the first half of the files from the failing file set 404 (i.e., files Fj-F^ and the second half of the files from the passing set 402 (i.e., P5-P8). The interaction of files F6 and F7 is not present in file set 408 so file set 408 does not exhibit problematic behavior. The 4J system then creates a test file set 410 consisting of the first half of the files from the passing file set 402 (i.e., files Pj-P^ and the second half of the failing set 402 (i.e., F5-F8). The interaction of files F6 and F7 is present in file set 410 so file set 410 exhibits problematic behavior. The system has now 5Q determined that the problematic file(s) exist in files F5-F8. Therefore, files F1-P4 will be used as the first half of subsequent test file sets.
Test file set 412 consists of Pj-P^ F5, F6, P7, and P8. The first half of test file set 412 consists of the first four passing 55 files Pt-P4. Of the files still to be evaluated (i.e. the 5th through 8th files), the first half are failing files F5 and F6 and the second half are passing files P7 and P8. Because test file set 412 does not contain the interaction of files F6 and F7 it passes. Test file set 414 reverses the order of the passing and 60 failing files (i.e., P5 and P6 are included with F7 and F8). Test file set 414 likewise passes, as it does not contain the interaction of F6 and F7. This fact alerts the system that the problem is not due to an individual file, but due to the interaction of two or more of the files F5 through F8. 65
Because the problematic behavior is caused by the interaction of two or more files, the first half of the files still to
be evaluated are failing files, F5 and F6 and the first half of the remaining files are failing files, F7. Test file set 416 is created using F5, F6, F7 and P8. Test file set 416 checks the interaction between F5 and F7 and the interaction between F6 and F7, as well as the interaction of F5, F6 and F7. Because this includes the interaction between F6 and F7, test file set 416 fails. The failure of test file set 416 shows that the interaction of F5 and F7 or the interaction of F6 and F7, or the interaction of F5, F6, and F7 is problematic. If test file set 416 had passed, this would eliminate the interaction the interaction of F5 and F7, the interaction of F6 and F7, and the interaction of F5, F6, and F7 as possible problematic interactions.
Test file set 418 is created as described above and contains F5, P6, F7, and F8. Test file set 418 passes because it does not contain the interaction of F6 and F7. Test file set 418 eliminates the interaction of F5 and F7 as a source of the problematic behavior. The process is continued with the creation of test file set 420 containing P5, F6, F7, and F8. Test file set 420 contains the interaction of F6 and F7 and therefore fails. This iteration shows that the interaction of F6 and F7 is sufficient to cause failure.
The system has now ascertained that files F6 and F7 when compiled together cause a runtime failure or performance problem for a given compiling environment (i.e., the "failing" compiling environment). A computer programmer may now compile one or both of files F6 and F7 using a "passing" compiling environment. The remainder of the files may be compiled using the "failing" compiling environment in order to take advantage, to the extent possible, of such a compiling environment. Alternatively, the information generated by the system may be used to debug the problematic file(s) and eliminate the source of the problematic behavior.
In the foregoing specification the invention has been described with reference to specific exemplary embodiments thereof. It will, however, be evident that various modifications and changes may be made thereto without departing from broader spirit and scope of the invention as set forth in the appended claims. The specification and drawings are, accordingly, to be regarded in an illustrative rather a restrictive sense.
procedure search (start, end, count, state, answer)
integer start, end, count
boolean answerf] begin
integer middle, i
// If the search has been narrowed to a single file,
// then the faulty file has been found
if start = end then
answer[start] := FAILING
// Divide the search area into 2 halves
middle := (start + end) / 2
// First try with the 1st half failing and the 2nd half passing
for i := start to end do
if i ^ middle then
state[i] := FAILING
state[i] := PASSING
// If it failed, call search again with the 2nd half passing
if query(state, count) = FAIL then
search(begin, middle, count, state, answer)