prof(1) User Commands prof(1)NAMEprof - display profile data
SYNOPSISprof [-ChsVz] [-a | c | n | t] [-o | x] [-g | l] [-m mdata]
[prog]
DESCRIPTION
The prof command interprets a profile file produced by the monitor
function. The symbol table in the object file prog (a.out by default)
is read and correlated with a profile file (mon.out by default). For
each external text symbol the percentage of time spent executing
between the address of that symbol and the address of the next is
printed, together with the number of times that function was called and
the average number of milliseconds per call.
OPTIONS
The mutually exclusive options -a, -c, -n, and -t determine the type of
sorting of the output lines:
-a Sort by increasing symbol address.
-c Sort by decreasing number of calls.
-n Sort lexically by symbol name.
-t Sort by decreasing percentage of total time (default).
The mutually exclusive options -o and -x specify the printing of the
address of each symbol monitored:
-o Print each symbol address (in octal) along with the symbol name.
-x Print each symbol address (in hexadecimal) along with the symbol
name.
The mutually exclusive options -g and -l control the type of symbols to
be reported. The -l option must be used with care; it applies the time
spent in a static function to the preceding (in memory) global func‐
tion, instead of giving the static function a separate entry in the
report. If all static functions are properly located, this feature can
be very useful. If not, the resulting report may be misleading.
Assume that A and B are global functions and only A calls static
function S. If S is located immediately after A in the source code
(that is, if S is properly located), then, with the -l option, the
amount of time spent in A can easily be determined, including the time
spent in S. If, however, both A and B call S, then, if the -l
option is used, the report will be misleading; the time spent during
B's call to S will be attributed to A, making it appear as if more
time had been spent in A than really had. In this case, function S
cannot be properly located.
-g List the time spent in static (non-global) functions separately.
The -g option function is the opposite of the -l function.
-l Suppress printing statically declared functions. If this option
is given, time spent executing in a static function is allocated
to the closest global function loaded before the static function
in the executable. This option is the default. It is the oppo‐
site of the -g function and should be used with care.
The following options may be used in any combination:
-C Demangle C++ symbol names before printing them out.
-h Suppress the heading normally printed on the report. This
is useful if the report is to be processed further.
-m mdata Use file mdata instead of mon.out as the input profile
file.
-s Print a summary of several of the monitoring parameters and
statistics on the standard error output.
-V Print prof version information on the standard error out‐
put.
-z Include all symbols in the profile range, even if associ‐
ated with zero number of calls and zero time.
A single function may be split into subfunctions for profiling by means
of the MARK macro. See prof(5).
ENVIRONMENT VARIABLES
PROFDIR The name of the file created by a profiled program is con‐
trolled by the environment variable PROFDIR. If PROFDIR is
not set, mon.out is produced in the directory current when
the program terminates. If PROFDIR=string, string/pid.prog‐
name is produced, where progname consists of argv[0] with
any path prefix removed, and pid is the process ID of the
program. If PROFDIR is set, but null, no profiling output
is produced.
FILES
mon.out default profile file
a.out default namelist (object) file
ATTRIBUTES
See attributes(5) for descriptions of the following attributes:
┌─────────────────────────────┬─────────────────────────────┐
│ ATTRIBUTE TYPE │ ATTRIBUTE VALUE │
├─────────────────────────────┼─────────────────────────────┤
│Availability │developer/object-file │
└─────────────────────────────┴─────────────────────────────┘
SEE ALSOgprof(1), exit(2), pcsample(2), profil(2), malloc(3C), malloc(3MALLOC),
monitor(3C), attributes(5), prof(5)NOTES
If the executable image has been stripped and does not have the .symtab
symbol table, gprof reads the global dynamic symbol tables .dynsym and
.SUNW_ldynsym, if present. The symbols in the dynamic symbol tables
are a subset of the symbols that are found in .symtab. The .dynsym sym‐
bol table contains the global symbols used by the runtime linker.
.SUNW_ldynsym augments the information in .dynsym with local function
symbols. In the case where .dynsym is found and .SUNW_ldynsym is not,
only the information for the global symbols is available. Without
local symbols, the behavior is as described for the -a option.
The times reported in successive identical runs may show variances
because of varying cache-hit ratios that result from sharing the cache
with other processes. Even if a program seems to be the only one using
the machine, hidden background or asynchronous processes may blur the
data. In rare cases, the clock ticks initiating recording of the pro‐
gram counter may beat with loops in a program, grossly distorting mea‐
surements. Call counts are always recorded precisely, however.
Only programs that call exit or return from main are guaranteed to
produce a profile file, unless a final call to monitor is explicitly
coded.
The times for static functions are attributed to the preceding external
text symbol if the -g option is not used. However, the call counts for
the preceding function are still correct; that is, the static function
call counts are not added to the call counts of the external function.
If more than one of the options -t, -c, -a, and -n is specified, the
last option specified is used and the user is warned.
LD_LIBRARY_PATH must not contain /usr/lib as a component when compiling
a program for profiling. If LD_LIBRARY_PATH contains /usr/lib, the
program will not be linked correctly with the profiling versions of the
system libraries in /usr/lib/libp. See gprof(1).
Functions such as mcount(), _mcount(), moncontrol(), _moncontrol(),
monitor(), and _monitor() may appear in the prof report. These func‐
tions are part of the profiling implementation and thus account for
some amount of the runtime overhead. Since these functions are not
present in an unprofiled application, time accumulated and call counts
for these functions may be ignored when evaluating the performance of
an application.
64-bit profiling
64-bit profiling may be used freely with dynamically linked executa‐
bles, and profiling information is collected for the shared objects if
the objects are compiled for profiling. Care must be applied to inter‐
pret the profile output, since it is possible for symbols from differ‐
ent shared objects to have the same name. If duplicate names are seen
in the profile output, it is better to use the -s (summary) option,
which prefixes a module id before each symbol that is duplicated. The
symbols can then be mapped to appropriate modules by looking at the
modules information in the summary.
If the -a option is used with a dynamically linked executable, the
sorting occurs on a per-shared-object basis. Since there is a high
likelihood of symbols from differed shared objects to have the same
value, this results in an output that is more understandable. A blank
line separates the symbols from different shared objects, if the -s
option is given.
32-bit profiling
32-bit profiling may be used with dynamically linked executables, but
care must be applied. In 32-bit profiling, shared objects cannot be
profiled with prof. Thus, when a profiled, dynamically linked program
is executed, only the main portion of the image is sampled. This means
that all time spent outside of the main object, that is, time spent in
a shared object, will not be included in the profile summary; the total
time reported for the program may be less than the total time used by
the program.
Because the time spent in a shared object cannot be accounted for, the
use of shared objects should be minimized whenever a program is pro‐
filed with prof. If desired, the program should be linked to the pro‐
filed version of a library (or to the standard archive version if no
profiling version is available), instead of the shared object to get
profile information on the functions of a library. Versions of profiled
libraries may be supplied with the system in the /usr/lib/libp direc‐
tory. Refer to compiler driver documentation on profiling.
Consider an extreme case. A profiled program dynamically linked with
the shared C library spends 100 units of time in some libc routine,
say, malloc(). Suppose malloc() is called only from routine B and B
consumes only 1 unit of time. Suppose further that routine A consumes
10 units of time, more than any other routine in the main (profiled)
portion of the image. In this case, prof will conclude that most of
the time is being spent in A and almost no time is being spent in B.
From this it will be almost impossible to tell that the greatest
improvement can be made by looking at routine B and not routine A.
The value of the profiler in this case is severely degraded; the solu‐
tion is to use archives as much as possible for profiling.
SunOS 5.11 25 Aug 2009 prof(1)
