MAN page from RedHat EL 6 toolchain-armiwmmx-2012-2012.12.1-oselas.2.2.x86_64.rpm


Section: GNU (1)
Updated: 2011-10-26


gcov - coverage testing tool 


gcov [-v|--version] [-h|--help]
     [-o|--object-directory directory|filesourcefiles


gcov is a test coverage program. Use it in concert with GCCto analyze your programs to help create more efficient, faster runningcode and to discover untested parts of your program. You can usegcov as a profiling tool to help discover where youroptimization efforts will best affect your code. You can also usegcov along with the other profiling tool, gprof, toassess which parts of your code use the greatest amount of computingtime.

Profiling tools help you analyze your code's performance. Using aprofiler such as gcov or gprof, you can find out somebasic performance statistics, such as:

how often each line of code executes
what lines of code are actually executed
how much computing time each section of code uses

Once you know these things about how your code works when compiled, youcan look at each module to see which modules should be optimized.gcov helps you determine where to work on optimization.

Software developers also use coverage testing in concert withtestsuites, to make sure software is actually good enough for a release.Testsuites can verify that a program works as expected; a coverageprogram tests to see how much of the program is exercised by thetestsuite. Developers can then determine what kinds of test cases needto be added to the testsuites to create both better testing and a betterfinal product.

You should compile your code without optimization if you plan to usegcov because the optimization, by combining some lines of codeinto one function, may not give you as much information as you need tolook for `hot spots' where the code is using a great deal of computertime. Likewise, because gcov accumulates statistics by line (atthe lowest resolution), it works best with a programming style thatplaces only one statement on each line. If you use complicated macrosthat expand to loops or to other control structures, the statistics areless helpful---they only report on the line where the macro callappears. If your complex macros behave like functions, you can replacethem with inline functions to solve this problem.

gcov creates a logfile called sourcefile.gcov whichindicates how many times each line of a source file sourcefile.chas executed. You can use these logfiles along with gprof to aidin fine-tuning the performance of your programs. gprof givestiming information you can use along with the information you get fromgcov.

gcov works only on code compiled with GCC. It is notcompatible with any other profiling or test coverage mechanism. 


Display help about using gcov (on the standard output), andexit without doing any further processing.
Display the gcov version number (on the standard output),and exit without doing any further processing.
Write individual execution counts for every basic block. Normally gcovoutputs execution counts only for the main blocks of a line. With thisoption you can determine if blocks within a single line are not beingexecuted.
Write branch frequencies to the output file, and write branch summaryinfo to the standard output. This option allows you to see how ofteneach branch in your program was taken. Unconditional branches will notbe shown, unless the -u option is given.
Write branch frequencies as the number of branches taken, rather thanthe percentage of branches taken.
Do not create the gcov output file.
Create long file names for included source files. For example, if theheader file x.h contains code, and was included in the filea.c, then running gcov on the file a.c will producean output file called a.c##x.h.gcov instead of x.h.gcov.This can be useful if x.h is included in multiple sourcefiles. If you use the -p option, both the including andincluded file names will be complete path names.
Preserve complete path information in the names of generated.gcov files. Without this option, just the filename component isused. With this option, all directories are used, with / characterstranslated to # characters, . directory componentsremoved and ..components renamed to ^. This is useful if sourcefiles are in severaldifferent directories. It also affects the -l option.
Output summaries for each function in addition to the file level summary.
-o directory|file
--object-directory directory
--object-file file
Specify either the directory containing the gcov data files, or theobject path name. The .gcno, and.gcda data files are searched for using this option. If a directoryis specified, the data files are in that directory and named after thesource file name, without its extension. If a file is specified here,the data files are named after that file, without its extension. If thisoption is not supplied, it defaults to the current directory.
When branch probabilities are given, include those of unconditional branches.Unconditional branches are normally not interesting.
Display the progress on the standard output.

gcov should be run with the current directory the same as thatwhen you invoked the compiler. Otherwise it will not be able to locatethe source files. gcov produces files calledmangledname.gcov in the current directory. These containthe coverage information of the source file they correspond to.One .gcov file is produced for each source file containing code,which was compiled to produce the data files. The mangledname partof the output file name is usually simply the source file name, but canbe something more complicated if the -l or -p options aregiven. Refer to those options for details.

The .gcov files contain the : separated fields along withprogram source code. The format is

        <execution_count>:<line_number>:<source line text>

Additional block information may succeed each line, when requested bycommand line option. The execution_count is - for linescontaining no code and ##### for lines which were never executed.Some lines of information at the start have line_number of zero.

The preamble lines are of the form


The ordering and number of these preamble lines will be augmented asgcov development progresses --- do not rely on them remainingunchanged. Use tag to locate a particular preamble line.

The additional block information is of the form

        <tag> <information>

The information is human readable, but designed to be simpleenough for machine parsing too.

When printing percentages, 0% and 100% are only printed when the valuesare exactly 0% and 100% respectively. Other values which wouldconventionally be rounded to 0% or 100% are instead printed as thenearest non-boundary value.

When using gcov, you must first compile your program with twospecial GCC options: -fprofile-arcs -ftest-coverage.This tells the compiler to generate additional information needed bygcov (basically a flow graph of the program) and also includesadditional code in the object files for generating the extra profilinginformation needed by gcov. These additional files are placed in thedirectory where the object file is located.

Running the program will cause profile output to be generated. For eachsource file compiled with -fprofile-arcs, an accompanying.gcda file will be placed in the object file directory.

Running gcov with your program's source file names as argumentswill now produce a listing of the code along with frequency of executionfor each line. For example, if your program is called tmp.c, thisis what you see when you use the basic gcov facility:

        $ gcc -fprofile-arcs -ftest-coverage tmp.c        $ a.out        $ gcov tmp.c        90.00% of 10 source lines executed in file tmp.c        Creating tmp.c.gcov.

The file tmp.c.gcov contains output from gcov.Here is a sample:

                -:    0:Source:tmp.c                -:    0:Graph:tmp.gcno                -:    0:Data:tmp.gcda                -:    0:Runs:1                -:    0:Programs:1                -:    1:#include <stdio.h>                -:    2:                -:    3:int main (void)                1:    4:{                1:    5:  int i, total;                -:    6:                1:    7:  total = 0;                -:    8:               11:    9:  for (i = 0; i < 10; i++)               10:   10:    total += i;                -:   11:                1:   12:  if (total != 45)            #####:   13:    printf ("Failure\n");                -:   14:  else                1:   15:    printf ("Success\n");                1:   16:  return 0;                -:   17:}

When you use the -a option, you will get individual blockcounts, and the output looks like this:

                -:    0:Source:tmp.c                -:    0:Graph:tmp.gcno                -:    0:Data:tmp.gcda                -:    0:Runs:1                -:    0:Programs:1                -:    1:#include <stdio.h>                -:    2:                -:    3:int main (void)                1:    4:{                1:    4-block  0                1:    5:  int i, total;                -:    6:                1:    7:  total = 0;                -:    8:               11:    9:  for (i = 0; i < 10; i++)               11:    9-block  0               10:   10:    total += i;               10:   10-block  0                -:   11:                1:   12:  if (total != 45)                1:   12-block  0            #####:   13:    printf ("Failure\n");            $$$$$:   13-block  0                -:   14:  else                1:   15:    printf ("Success\n");                1:   15-block  0                1:   16:  return 0;                1:   16-block  0                -:   17:}

In this mode, each basic block is only shown on one line --- the lastline of the block. A multi-line block will only contribute to theexecution count of that last line, and other lines will not be shownto contain code, unless previous blocks end on those lines.The total execution count of a line is shown and subsequent lines showthe execution counts for individual blocks that end on that line. After eachblock, the branch and call counts of the block will be shown, if the-b option is given.

Because of the way GCC instruments calls, a call count can be shownafter a line with no individual blocks.As you can see, line 13 contains a basic block that was not executed.

When you use the -b option, your output looks like this:

        $ gcov -b tmp.c        90.00% of 10 source lines executed in file tmp.c        80.00% of 5 branches executed in file tmp.c        80.00% of 5 branches taken at least once in file tmp.c        50.00% of 2 calls executed in file tmp.c        Creating tmp.c.gcov.

Here is a sample of a resulting tmp.c.gcov file:

                -:    0:Source:tmp.c                -:    0:Graph:tmp.gcno                -:    0:Data:tmp.gcda                -:    0:Runs:1                -:    0:Programs:1                -:    1:#include <stdio.h>                -:    2:                -:    3:int main (void)        function main called 1 returned 1 blocks executed 75%                1:    4:{                1:    5:  int i, total;                -:    6:                1:    7:  total = 0;                -:    8:               11:    9:  for (i = 0; i < 10; i++)        branch  0 taken 91% (fallthrough)        branch  1 taken 9%               10:   10:    total += i;                -:   11:                1:   12:  if (total != 45)        branch  0 taken 0% (fallthrough)        branch  1 taken 100%            #####:   13:    printf ("Failure\n");        call    0 never executed                -:   14:  else                1:   15:    printf ("Success\n");        call    0 called 1 returned 100%                1:   16:  return 0;                -:   17:}

For each function, a line is printed showing how many times the functionis called, how many times it returns and what percentage of thefunction's blocks were executed.

For each basic block, a line is printed after the last line of the basicblock describing the branch or call that ends the basic block. There canbe multiple branches and calls listed for a single source line if thereare multiple basic blocks that end on that line. In this case, thebranches and calls are each given a number. There is no simple way to mapthese branches and calls back to source constructs. In general, though,the lowest numbered branch or call will correspond to the leftmost constructon the source line.

For a branch, if it was executed at least once, then a percentageindicating the number of times the branch was taken divided by thenumber of times the branch was executed will be printed. Otherwise, themessage ``never executed'' is printed.

For a call, if it was executed at least once, then a percentageindicating the number of times the call returned divided by the numberof times the call was executed will be printed. This will usually be100%, but may be less for functions that call "exit" or "longjmp",and thus may not return every time they are called.

The execution counts are cumulative. If the example program wereexecuted again without removing the .gcda file, the count for thenumber of times each line in the source was executed would be added tothe results of the previous run(s). This is potentially useful inseveral ways. For example, it could be used to accumulate data over anumber of program runs as part of a test verification suite, or toprovide more accurate long-term information over a large number ofprogram runs.

The data in the .gcda files is saved immediately before the programexits. For each source file compiled with -fprofile-arcs, theprofiling code first attempts to read in an existing .gcda file; ifthe file doesn't match the executable (differing number of basic blockcounts) it will ignore the contents of the file. It then adds in thenew execution counts and finally writes the data to the file. 

Using gcov with GCC Optimization

If you plan to use gcov to help optimize your code, you mustfirst compile your program with two special GCC options:-fprofile-arcs -ftest-coverage. Aside from that, you can use anyother GCC options; but if you want to prove that every single linein your program was executed, you should not compile with optimizationat the same time. On some machines the optimizer can eliminate somesimple code lines by combining them with other lines. For example, codelike this:

        if (a != b)          c = 1;        else          c = 0;

can be compiled into one instruction on some machines. In this case,there is no way for gcov to calculate separate execution countsfor each line because there isn't separate code for each line. Hencethe gcov output looks like this if you compiled the program withoptimization:

              100:   12:if (a != b)              100:   13:  c = 1;              100:   14:else              100:   15:  c = 0;

The output shows that this block of code, combined by optimization,executed 100 times. In one sense this result is correct, because therewas only one instruction representing all four of these lines. However,the output does not indicate how many times the result was 0 and howmany times the result was 1.

Inlineable functions can create unexpected line counts. Line counts areshown for the source code of the inlineable function, but what is showndepends on where the function is inlined, or if it is not inlined at all.

If the function is not inlined, the compiler must emit an out of linecopy of the function, in any object file that needs it. IffileA.o and fileB.o both contain out of line bodies of aparticular inlineable function, they will also both contain coveragecounts for that function. When fileA.o and fileB.o arelinked together, the linker will, on many systems, select one of thoseout of line bodies for all calls to that function, and remove or ignorethe other. Unfortunately, it will not remove the coverage counters forthe unused function body. Hence when instrumented, all but one use ofthat function will show zero counts.

If the function is inlined in several places, the block structure ineach location might not be the same. For instance, a condition mightnow be calculable at compile time in some instances. Because thecoverage of all the uses of the inline function will be shown for thesame source lines, the line counts themselves might seem inconsistent. 


gpl(7), gfdl(7), fsf-funding(7), gcc(1) and the Info entry for gcc. 


Copyright (c) 1996, 1997, 1999, 2000, 2001, 2002, 2003, 2004,2005, 2008, 2010 Free Software Foundation, Inc.

Permission is granted to copy, distribute and/or modify this documentunder the terms of the GNU Free Documentation License, Version 1.3 orany later version published by the Free Software Foundation; with theInvariant Sections being ``GNU General Public License'' and ``FundingFree Software'', the Front-Cover texts being (a) (see below), and withthe Back-Cover Texts being (b) (see below). A copy of the license isincluded in the gfdl(7) man page.

(a) The FSF's Front-Cover Text is:

     A GNU Manual

(b) The FSF's Back-Cover Text is:

     You have freedom to copy and modify this GNU Manual, like GNU     software.  Copies published by the Free Software Foundation raise     funds for GNU development.



Using gcov with GCC Optimization

This document was created byman2html,using the manual pages.