.\" Automatically generated by Pandoc 3.1.3
.\"
.\" Define V font for inline verbatim, using C font in formats
.\" that render this, and otherwise B font.
.ie "\f[CB]x\f[]"x" \{\
. ftr V B
. ftr VI BI
. ftr VB B
. ftr VBI BI
.\}
.el \{\
. ftr V CR
. ftr VI CI
. ftr VB CB
. ftr VBI CBI
.\}
.TH "UFTRACE-REPORT" "1" "Sep, 2018" "Uftrace User Manuals" ""
.hy
.SH NAME
.PP
uftrace-report - Print statistics and summary for trace data
.SH SYNOPSIS
.PP
uftrace report [\f[I]options\f[R]]
.SH DESCRIPTION
.PP
This command collects trace data from a given data file and prints
statistics and summary information.
It shows function statistics by default, but can show task statistics
with the \f[V]--task\f[R] option and show differences between traces
with the \f[V]--diff\f[R] option.
.SH REPORT OPTIONS
.TP
-f \f[I]FIELD\f[R], --output-fields=\f[I]FIELD\f[R]
Customize field in the output.
Possible values are: \f[V]total\f[R], \f[V]total-avg\f[R],
\f[V]total-min\f[R], \f[V]total-max\f[R], \f[V]total-stdv\f[R],
\f[V]self\f[R], \f[V]self-avg\f[R], \f[V]self-min\f[R],
\f[V]self-max\f[R], \f[V]self-stdv\f[R], \f[V]size\f[R], \f[V]call\f[R]
and \f[V]all\f[R].
Multiple fields can be set by using comma.
Special field of `none' can be used (solely) to hide all fields and
`all' can be used to show all fields.
Default is `total,self,call'.
See \f[I]FIELDS\f[R].
\f[V]stdv\f[R] stands for the relative standard deviation (RSD), which
is the standard deviation expressed as a percentage of the mean.
The relative standard deviation is calculated using the formula:
.RS
.PP
[ RSD = ( ) % ]
.PP
where: - ( ) represents the standard deviation.
- ( ) represents the mean of the function\[cq]s time.
.RE
.TP
-s \f[I]KEYS\f[R][,\f[I]KEYS\f[R],\&...], --sort=\f[I]KEYS\f[R][,\f[I]KEYS\f[R],\&...]
Sort functions by given KEYS.
Multiple KEYS can be given, separated by comma (,).
Possible keys are \f[V]total\f[R] (time), \f[V]total-avg\f[R],
\f[V]total-min\f[R], \f[V]total-max\f[R], \f[V]total-stdv\f[R],
\f[V]self\f[R] (time), \f[V]self-avg\f[R], \f[V]self-min\f[R],
\f[V]self-max\f[R], \f[V]self-stdv\f[R], \f[V]size\f[R], \f[V]call\f[R]
and \f[V]func\f[R].
But if either \f[V]--avg-total\f[R] or \f[V]--avg-self\f[R] is used, the
possible keys can be \f[V]avg\f[R], \f[V]min\f[R] and \f[V]max\f[R] that
apply to total or self time respectively.
.TP
--avg-total
Show average, min, max, relative standard deviation of each
function\[cq]s total time.
.TP
--avg-self
Show average, min, max, relative standard deviation of each
function\[cq]s self time.
.TP
--task
Report task summary information rather than function statistics.
Customize field in the output with -f option.
Possible values are: \f[V]total\f[R], \f[V]self\f[R], \f[V]func\f[R] and
\f[V]tid\f[R].
Multiple fields can be set by using comma.
Special field of `none' can be used (solely) to hide all fields.
Default is `total,self,func,tid'.
See \f[I]TASK FIELDS\f[R].
.TP
--diff=\f[I]DATA\f[R]
Report differences between the input trace data and the given DATA.
.TP
--diff-policy=\f[I]POLICY\f[R]
Apply custom diff policy.
Available values are: \[lq]abs\[rq], \[lq]no-abs\[rq],
\[lq]percent\[rq], \[lq]no-percent\[rq], \[lq]compact\[rq] and
\[lq]full\[rq].
The \[lq]abs\[rq] is to sort diff result using absolute value so
positive and negative entries can be shown together while
\[lq]no-abs\[rq] will show positive entries first and then negative
ones.
The \[lq]percent\[rq] is to show diff in percentage while
\[lq]no-percent\[rq] is to show the values.
The \[lq]full\[rq] is to show all three columns of baseline, new data
and difference while \[lq]compact\[rq] only shows the difference.
The default is \[lq]abs\[rq], \[lq]compact\[rq] and
\[lq]no-percent\[rq].
.TP
--sort-column=\f[I]IDX\f[R]
When \f[V]--diff\f[R] is used with \[lq]full\[rq] policy, 3 columns will
be shown for each total time, self time and call count.
This option selects the index of the column to be used as a sort key.
Index 0 is for original data given by the \f[V]--data\f[R] option, index
1 is for data given by the \f[V]--diff\f[R] option, and index 2 is for
(percentage) differences between the two data.
.TP
--srcline
Show source location of each function if available.
.TP
--format=\f[I]TYPE\f[R]
Show format style output.
Currently, normal and html styles are supported.
.SH COMMON OPTIONS
.TP
-F \f[I]FUNC\f[R], --filter=\f[I]FUNC\f[R]
Set filter to trace selected functions and their children functions.
This option can be used more than once.
See \f[V]uftrace-replay\f[R](1) for an explanation of filters.
.TP
-N \f[I]FUNC\f[R], --notrace=\f[I]FUNC\f[R]
Set filter not to trace selected functions and their children functions.
This option can be used more than once.
See \f[V]uftrace-replay\f[R](1) for an explanation of filters.
.TP
-C \f[I]FUNC\f[R], --caller-filter=\f[I]FUNC\f[R]
Set filter to trace callers of selected functions only.
This option can be used more than once.
See \f[V]uftrace-replay\f[R](1) for an explanation of filters.
.TP
-T \f[I]TRG\f[R], --trigger=\f[I]TRG\f[R]
Set trigger on selected functions.
This option can be used more than once.
See \f[V]uftrace-replay\f[R](1) for an explanation of triggers.
.TP
-D \f[I]DEPTH\f[R], --depth \f[I]DEPTH\f[R]
Set trace limit in nesting level.
.TP
-t \f[I]TIME\f[R], --time-filter=\f[I]TIME\f[R]
Do not account functions which run under the time threshold.
If some functions explicitly have the `trace' trigger applied, those are
always accounted regardless of execution time.
.TP
-Z \f[I]SIZE\f[R], --size-filter=\f[I]SIZE\f[R]
Do not show functions smaller than SIZE bytes.
.TP
-L \f[I]LOCATION\f[R], --loc-filter=\f[I]LOCATION\f[R]
Set filter to trace selected source locations.
This option can be used more than once.
.TP
--no-libcall
Do not show library calls.
.TP
--no-event
Do not show any events.
Implies \f[V]--no-sched\f[R].
.TP
--no-sched
Do not show schedule events.
.TP
--no-sched-preempt
Do not show preempt schedule events but show regular(sleeping) schedule
events.
.TP
--match=\f[I]TYPE\f[R]
Use pattern match using TYPE.
Possible types are \f[V]regex\f[R] and \f[V]glob\f[R].
Default is \f[V]regex\f[R].
.TP
--with-syms=\f[I]DIR\f[R]
Read symbol data from the .sym files in \f[I]DIR\f[R] directory instead
of the binary.
This can be useful to deal with stripped binaries.
The file name of the main binary should be the same when saved and used.
.SH COMMON ANALYSIS OPTIONS
.TP
-H \f[I]FUNC\f[R], --hide=\f[I]FUNC\f[R]
Set filter not to trace selected functions.
It doesn\[cq]t affect their subtrees, but hides only the given
functions.
This option can be used more than once.
See \f[V]uftrace-replay\f[R](1) for an explanation of filters.
.TP
--kernel-full
Show all kernel functions, including those called outside of user
functions.
.TP
--kernel-only
Show kernel functions only without user functions.
.TP
--event-full
Show all (user) events outside of user functions.
.TP
--tid=\f[I]TID\f[R][,\f[I]TID\f[R],\&...]
Only print functions called by the given tasks.
To see the list of tasks in the data file, you can use
\f[V]uftrace report --task\f[R] or \f[V]uftrace info\f[R].
This option can also be used more than once.
.TP
--demangle=\f[I]TYPE\f[R]
Use demangled C++ symbol names for filters, triggers, arguments and/or
return values.
Possible values are \[lq]full\[rq], \[lq]simple\[rq] and \[lq]no\[rq].
Default is \[lq]simple\[rq] which ignores function arguments and
template parameters.
.TP
-r \f[I]RANGE\f[R], --time-range=\f[I]RANGE\f[R]
Only show functions executed within the time RANGE.
The RANGE can be <start>\[ti]<stop> (separated by \[lq]\[ti]\[rq]) and
one of <start> and <stop> can be omitted.
The <start> and <stop> are timestamp or elapsed time if they have
<time_unit> postfix, for example `100us'.
The timestamp or elapsed time can be shown with \f[V]-f time\f[R] or
\f[V]-f elapsed\f[R] option respectively in \f[V]uftrace replay\f[R](1).
.SH EXAMPLE
.PP
This command shows information like the following:
.IP
.nf
\f[C]
$ uftrace record fibonacci
$ uftrace report
  Total time   Self time       Calls  Function
  ==========  ==========  ==========  ====================
    3.781 us    0.124 us           1  main
    3.657 us    3.657 us          41  fib
    0.345 us    0.345 us           1  __monstartup
    0.269 us    0.269 us           1  __cxa_atexit

$ uftrace report -s call,self
  Total time   Self time       Calls  Function
  ==========  ==========  ==========  ====================
    3.657 us    3.657 us          41  fib
    0.345 us    0.345 us           1  __monstartup
    0.269 us    0.269 us           1  __cxa_atexit
    3.781 us    0.124 us           1  main

$ uftrace report --avg-self
    Self avg    Self min    Self max   Self stdv  Function
  ==========  ==========  ==========  ==========  ====================
    0.414 us    0.414 us    0.414 us       0.00%   __monstartup
    0.356 us    0.356 us    0.356 us       0.00%   __cxa_atexit
    0.113 us    0.113 us    0.113 us       0.00%   main
    0.087 us    0.032 us    0.369 us      78.13%   fib

$ uftrace report --task
  Total time   Self time     TID   Num funcs  Task name
  ==========  ==========  ======  ==========  ====================
    4.395 us    4.395 us   28662          44  fibonacci

$ uftrace record --srcline abc
$ uftrace report --srcline
  Total time   Self time       Calls  Function [Source]
  ==========  ==========  ==========  ====================
   17.508 us    2.199 us           1  main [./tests/s-abc.c:26]
   15.309 us    2.384 us           1  a [./tests/s-abc.c:11]
   12.925 us    2.633 us           1  b [./tests/s-abc.c:16]
   10.292 us    5.159 us           1  c [./tests/s-abc.c:21]
    5.133 us    5.133 us           1  getpid
    3.437 us    3.437 us           1  __monstartup
    1.959 us    1.959 us           1  __cxa_atexit
\f[R]
.fi
.PP
To see a difference between two data:
.IP
.nf
\f[C]
$ uftrace record abc

$ uftrace report --diff uftrace.data.old
#
# uftrace diff
#  [0] base: uftrace.data       (from uftrace record abc )
#  [1] diff: uftrace.data.old   (from uftrace record abc )
#
  Total time   Self time       Calls  Function
  ==========  ==========  ==========  ====================
   -0.301 us   -0.038 us          +0  main
   -0.263 us   -0.070 us          +0  a
   -0.193 us   -0.042 us          +0  b
   -0.151 us   -0.090 us          +0  c
   -0.131 us   -0.131 us          +0  __cxa_atexit
   -0.061 us   -0.061 us          +0  getpid
\f[R]
.fi
.PP
The above example shows difference sorted by absolute value of total
time.
The following changes it to use (non-absolute) value of self time.
.IP
.nf
\f[C]
$ uftrace report --diff uftrace.data.old -s self --diff-policy no-abs
#
# uftrace diff
#  [0] base: uftrace.data       (from uftrace record abc )
#  [1] diff: uftrace.data.old   (from uftrace record abc )
#
  Total time   Self time       Calls  Function
  ==========  ==========  ==========  ====================
   -0.301 us   -0.038 us          +0  main
   -0.193 us   -0.042 us          +0  b
   -0.061 us   -0.061 us          +0  getpid
   -0.263 us   -0.070 us          +0  a
   -0.151 us   -0.090 us          +0  c
   -0.131 us   -0.131 us          +0  __cxa_atexit
\f[R]
.fi
.PP
By using \[lq]full\[rq] policy, user can see raw data as well like
below.
Also it\[cq]s possible to sort by different column (for raw data).
The example below will sort output by total time of the base data.
.IP
.nf
\f[C]
$ uftrace report --diff uftrace.data.old --sort-column 0 --diff-policy full,percent
#
# uftrace diff
#  [0] base: uftrace.data       (from uftrace record abc )
#  [1] diff: uftrace.data.old   (from uftrace record abc )
#
                 Total time (diff)                   Self time (diff)                  Nr. called (diff)   Function
  ================================   ================================   ================================   ====================
    2.812 us    2.511 us   -10.70%     0.403 us    0.365 us    -9.43%            1          1         +0   main
    2.409 us    2.146 us   -10.92%     0.342 us    0.272 us   -20.47%            1          1         +0   a
    2.067 us    1.874 us    -9.34%     0.410 us    0.368 us   -10.24%            1          1         +0   b
    1.657 us    1.506 us    -9.11%     0.890 us    0.800 us   -10.11%            1          1         +0   c
    0.920 us    0.789 us   -14.24%     0.920 us    0.789 us   -14.24%            1          1         +0   __cxa_atexit
    0.767 us    0.706 us    -7.95%     0.767 us    0.706 us    -7.95%            1          1         +0   getpid
\f[R]
.fi
.SH FIELDS
.PP
The uftrace allows user to customize the report output with a couple of
fields.
By default it uses total, self and call fields, but you can use other
fields in any order like:
.IP
.nf
\f[C]
$ uftrace report -f total,total-max,self-min,call
Total time   Total max    Self min       Calls  Function
==========  ==========  ==========  ==========  ====================
 97.234 us   36.033 us    1.073 us           3  lib_a
 50.552 us   26.690 us    2.828 us           2  lib_b
 46.806 us   46.806 us    3.290 us           1  main
 43.516 us   43.516 us    7.483 us           1  foo
 32.010 us   20.847 us    9.684 us           2  lib_c
\f[R]
.fi
.PP
Each field can be used as sort key:
.IP
.nf
\f[C]
$ uftrace report -f total,total-max,self-min,call -s call
Total time   Total max    Self min       Calls  Function
==========  ==========  ==========  ==========  ====================
 97.234 us   36.033 us    1.073 us           3  lib_a
 50.552 us   26.690 us    2.828 us           2  lib_b
 32.010 us   20.847 us    9.684 us           2  lib_c
 43.516 us   43.516 us    7.483 us           1  foo
 46.806 us   46.806 us    3.290 us           1  main

$ uftrace report -f total,total-max,self-min,total-min,call -s self-min,total-min
Total time   Total max    Self min   Total min       Calls  Function
==========  ==========  ==========  ==========  ==========  ====================
 32.010 us   20.847 us    9.684 us   11.163 us           2  lib_c
 43.516 us   43.516 us    7.483 us   43.516 us           1  foo
 46.806 us   46.806 us    3.290 us   46.806 us           1  main
 50.552 us   26.690 us    2.828 us   23.862 us           2  lib_b
 97.234 us   36.033 us    1.073 us   27.763 us           3  lib_a
\f[R]
.fi
.PP
Each field can be used with \[en]diff option:
.IP
.nf
\f[C]
$ uftrace report --diff uftrace.data.old -f total,total-min
#
# uftrace diff
#  [0] base: uftrace.data       (from uftrace record test/t-lib)
#  [1] diff: uftrace.data.old   (from uftrace record test/t-lib)
#
 Total time     Total min   Function
===========   ===========   ====================
 +34.560 us     +9.884 us   lib_a
 +18.086 us     +8.517 us   lib_b
 +16.887 us    +16.887 us   main
 +15.479 us    +15.479 us   foo
 +10.600 us     +3.127 us   lib_c

$ uftrace report --diff uftrace.data.old -f total,total-min,self-avg --diff-policy full
#
# uftrace diff
#  [0] base: uftrace.data           (from uftrace record --srcline test/t-lib)
#  [1] diff: uftrace.data.old   (from uftrace record --srcline test/t-lib)
#
                  Total time (diff)                      Total min (diff)                       Self avg (diff)   Function
===================================   ===================================   ===================================   ====================
 14.616 us   13.796 us    +0.820 us     4.146 us    3.823 us    +0.323 us     0.443 us    0.459 us    -0.016 us   lib_a
  6.529 us    5.957 us    +0.572 us     6.529 us    5.957 us    +0.572 us     0.436 us    0.356 us    +0.080 us   main
  7.700 us    7.173 us    +0.527 us     3.677 us    3.426 us    +0.251 us     0.365 us    0.363 us    +0.002 us   lib_b
  6.093 us    5.601 us    +0.492 us     6.093 us    5.601 us    +0.492 us     0.741 us    0.476 us    +0.265 us   foo
  5.638 us    5.208 us    +0.430 us     2.346 us    2.187 us    +0.159 us     1.646 us    1.510 us    +0.136 us   lib_c
\f[R]
.fi
.PP
Each field has following meaning:
.IP \[bu] 2
total: total time of each function.
.IP \[bu] 2
total-avg: average of total time of each function.
.IP \[bu] 2
total-min: min of total time of each function.
.IP \[bu] 2
total-max: max of total time of each function.
.IP \[bu] 2
self: self time of each function.
.IP \[bu] 2
self-avg: average of self time of each function.
.IP \[bu] 2
self-min: min of self time of each function.
.IP \[bu] 2
self-max: max of self time of each function.
.IP \[bu] 2
call: called count of each function.
.PP
The default value is `total,self,call'.
If given field name starts with \[lq]+\[rq], then it\[cq]ll be appended
to the default fields.
So \[lq]-f +total-avg\[rq] is as same as \[lq]-f
total,self,call,total-avg\[rq].
And it also accepts a special field name of `none' which disables the
field display and shows function output only.
.SH TASK FIELDS
.IP \[bu] 2
total: total time of each task.
.IP \[bu] 2
self: self time of each task.
.IP \[bu] 2
func: number of functions in the task.
.IP \[bu] 2
tid: task ID.
.PP
The default value is `total,self,func,tid'.
See \f[I]FIELDS\f[R] for field usage.
.SH SEE ALSO
.PP
\f[V]uftrace\f[R](1), \f[V]uftrace-record\f[R](1),
\f[V]uftrace-replay\f[R](1), \f[V]uftrace-tui\f[R](1)
.SH AUTHORS
Namhyung Kim <namhyung@gmail.com>.