NAME

gprof - Displays call graph profile data

SYNOPSIS

gprof [-absz] [-e routine | -f routine] ...  [-E routine | -F routine] ...  [object_file] [profile_file ...]

DESCRIPTION

The gprof command produces an execution profile of C programs.  The effect of called routines is incorporated into the profile of each caller. 

The profile data is taken from the specified call graph profile_file (gmon.out by default), created by programs compiled with the -pg flag with the cc driver command.  The -pg flag also links in versions of the library routines compiled for profiling when your program is linked by using the −non_shared flag with the cc command. 

The gprof profiling tool reads the symbol table in the specified object_file (a.out by default), correlating it with the call profile file.  If more than one profile_file is specified, gprof output shows the sum of the profile information in the given profile files. 

To produce call graph data, follow these steps:

     1.Compile your program with the −pg switch to the cc command. 

     2.Execute the program to produce a data file. 

     3.Run gprof on the data file. 

The gprof command produces three items:

     1.First, a flat profile is given, similar to that provided by prof.  This listing gives the total execution times and call counts for each of the functions in the program, sorted by decreasing time. 

     2.Next, these execution times are propagated along the edges of the call graph.  Cycles are discovered, and calls into a cycle are made to share the time of the cycle.  A second listing shows the functions sorted according to the time they represent, including the time of their call graph descendents.  Below each function entry is shown its (direct) call graph children, and how their times are propagated to this function.  A similar display above the function shows how this function’s time and the time of its descendents are propagated to its (direct) call graph parents. 

     3.Lastly, cycles are also shown, with an entry for the cycle as a whole and a listing of the members of the cycle and their contributions to the time and call counts of the cycle. 

To obtain call graph data on libraries, link your program by using the −non_shared flag with cc command. To disable profiling of a library, use the −no_pg flag when linking. For example, to enable call graph profiling of your executable program and then disable call graph profiling for system supplied libraries, such as libc, link your program by using the following command:

   % cc -non_shared -pg myprog -no_pg

You can use environment variables to change the default profiling behavior.  The variables are PROFDIR and PROFFLAGS. The general form for setting these variables is:

   For C shell:  setenv varname "value"
   For Bourne shell: varname = "value"; export varname

In the preceding example, varname can be one of the following:

PROFDIR
This environment variable causes PC-sampling data files to be generated with unique names in the directory you specify. You specify a directory path as the value and your prof results are placed in the file path/pid.progname where path is the pathname, pid is the process of the executing program, and progname is the program name. 

PROFFLAGS
This environment variable, which takes −threads as the value, causes a separate data file to be generated for each thread. The name of the data file takes the following form:
pid.sid.progname.  The form of the filename resolves to pid as the process ID of the program, sid as the sequence ID of the thread and progname as the name of the program being profiled. 

You can use the PROFDIR and PROFFLAGS environment variables together. For more information, see the Programmer’s Guide. 

FLAGS

-aSuppresses the printing of statically declared functions.  If this flag is given, all relevant information about the static function (for example, time samples, calls to other functions, calls from other functions) belongs to the function loaded just before the static function in the a.out file. 

-bSuppresses the printing of a description of each field in the profile. 

-e routineSuppresses the printing of the graph profile entry for routine and all its descendants (unless they have other ancestors that are not suppressed).  More than one -e flag may be given.  Only one routine may be given with each -e flag. 

-E routineSuppresses the printing of the graph profile entry for routine and its descendants (same as -e), and also excludes the time spent in routine and its descendants from the total and percentage time computations.  (For example, _gprof_mcount -E _gprof_monstop is the default.) 

-f routinePrints the graph profile entry of only the specified routine and its descendants.  More than one -f flag may be given.  Only one routine may be given with each -f flag.  The -f flag overrides the -e flag. 

-F routinePrints the graph profile entry of only routine and its descendants (same as -f), and also uses only the times of the printed routines in total time and percentage computations.  More than one -F flag may be given.  Only one routine may be given with each -F flag.  The -F flag overrides the -E flag. 

-sProduces a profile file, gmon.sum, which represents the sum of the profile information in all the specified profile files.  This summary profile file may be given to subsequent executions of gprof (probably also with a -s) to accumulate profile data across several runs of an a.out file. 

-zDisplays routines that have zero usage, as indicated by call counts and accumulated time. 

NOTES

     1.Be aware of possible quantization errors when using gprof.  The granularity of the sampling is shown, but remains statistical at best.  The time for each execution of a function can be expressed by the total time for the function divided by the number of times the function is called; thus, the time propagated along the call graph arcs to parents of that function is directly proportional to the number of times that arc is traversed. 

     2.Parents that are not themselves profiled have the time of their profiled children propagated to them, but appear to be spontaneously invoked in the call graph listing; they do not have their time propagated further. 

     3.Similarly, signal catchers, even though profiled, appear to be spontaneous.  Any profiled children of signal catchers should have their times propagated properly, unless the signal catcher was invoked during the execution of the profiling routine, in which case propagation cannot occur. 

     4.The profiled program must call exit or return normally for the profiling information to be saved in the gmon.out file. 

ERRORS

If the program makes more procedure calls than can be recorded, the error "Arc limit exceeded" will occur.  The program will continue to execute to completion, but some call-arc information will be lost.  A possible work-around for this problem is to reduce the portion of the program that is profiled using the monstartup() routine.  See monitor(3) for more information. 

FILES

a.outDefault object file. 

gmon.outDynamic call graph and profile. 

gmon.sumSummarized dynamic call graph and profile. 

RELATED INFORMATION

Programmer’s Guide

Commands:  prof(1), kgmon(8). 

Functions:  profil(2), monitor(3). 

gprof: A Call Graph Execution Profiler, by Graham, S.L., Kessler, P.B., McKusick, M.K.  Proceedings of the SIGPLAN ’82 Symposium on Compiler Construction, SIGPLAN Notices, Vol. 17, No. 6, pp. 120-126, June 1982.