• Home
  • History
  • Annotate
  • Line#
  • Navigate
  • Raw
  • Download
  • only in /netgear-R7000-V1.0.7.12_1.2.5/components/opensource/linux/linux-2.6.36/tools/perf/Documentation/
1perf-trace-perl(1)
2==================
3
4NAME
5----
6perf-trace-perl - Process trace data with a Perl script
7
8SYNOPSIS
9--------
10[verse]
11'perf trace' [-s [Perl]:script[.pl] ]
12
13DESCRIPTION
14-----------
15
16This perf trace option is used to process perf trace data using perf's
17built-in Perl interpreter.  It reads and processes the input file and
18displays the results of the trace analysis implemented in the given
19Perl script, if any.
20
21STARTER SCRIPTS
22---------------
23
24You can avoid reading the rest of this document by running 'perf trace
25-g perl' in the same directory as an existing perf.data trace file.
26That will generate a starter script containing a handler for each of
27the event types in the trace file; it simply prints every available
28field for each event in the trace file.
29
30You can also look at the existing scripts in
31~/libexec/perf-core/scripts/perl for typical examples showing how to
32do basic things like aggregate event data, print results, etc.  Also,
33the check-perf-trace.pl script, while not interesting for its results,
34attempts to exercise all of the main scripting features.
35
36EVENT HANDLERS
37--------------
38
39When perf trace is invoked using a trace script, a user-defined
40'handler function' is called for each event in the trace.  If there's
41no handler function defined for a given event type, the event is
42ignored (or passed to a 'trace_handled' function, see below) and the
43next event is processed.
44
45Most of the event's field values are passed as arguments to the
46handler function; some of the less common ones aren't - those are
47available as calls back into the perf executable (see below).
48
49As an example, the following perf record command can be used to record
50all sched_wakeup events in the system:
51
52 # perf record -a -e sched:sched_wakeup
53
54Traces meant to be processed using a script should be recorded with
55the above option: -a to enable system-wide collection.
56
57The format file for the sched_wakep event defines the following fields
58(see /sys/kernel/debug/tracing/events/sched/sched_wakeup/format):
59
60----
61 format:
62        field:unsigned short common_type;
63        field:unsigned char common_flags;
64        field:unsigned char common_preempt_count;
65        field:int common_pid;
66        field:int common_lock_depth;
67
68        field:char comm[TASK_COMM_LEN];
69        field:pid_t pid;
70        field:int prio;
71        field:int success;
72        field:int target_cpu;
73----
74
75The handler function for this event would be defined as:
76
77----
78sub sched::sched_wakeup
79{
80   my ($event_name, $context, $common_cpu, $common_secs,
81       $common_nsecs, $common_pid, $common_comm,
82       $comm, $pid, $prio, $success, $target_cpu) = @_;
83}
84----
85
86The handler function takes the form subsystem::event_name.
87
88The $common_* arguments in the handler's argument list are the set of
89arguments passed to all event handlers; some of the fields correspond
90to the common_* fields in the format file, but some are synthesized,
91and some of the common_* fields aren't common enough to to be passed
92to every event as arguments but are available as library functions.
93
94Here's a brief description of each of the invariant event args:
95
96 $event_name 	  	    the name of the event as text
97 $context		    an opaque 'cookie' used in calls back into perf
98 $common_cpu		    the cpu the event occurred on
99 $common_secs		    the secs portion of the event timestamp
100 $common_nsecs		    the nsecs portion of the event timestamp
101 $common_pid		    the pid of the current task
102 $common_comm		    the name of the current process
103
104All of the remaining fields in the event's format file have
105counterparts as handler function arguments of the same name, as can be
106seen in the example above.
107
108The above provides the basics needed to directly access every field of
109every event in a trace, which covers 90% of what you need to know to
110write a useful trace script.  The sections below cover the rest.
111
112SCRIPT LAYOUT
113-------------
114
115Every perf trace Perl script should start by setting up a Perl module
116search path and 'use'ing a few support modules (see module
117descriptions below):
118
119----
120 use lib "$ENV{'PERF_EXEC_PATH'}/scripts/perl/Perf-Trace-Util/lib";
121 use lib "./Perf-Trace-Util/lib";
122 use Perf::Trace::Core;
123 use Perf::Trace::Context;
124 use Perf::Trace::Util;
125----
126
127The rest of the script can contain handler functions and support
128functions in any order.
129
130Aside from the event handler functions discussed above, every script
131can implement a set of optional functions:
132
133*trace_begin*, if defined, is called before any event is processed and
134gives scripts a chance to do setup tasks:
135
136----
137 sub trace_begin
138 {
139 }
140----
141
142*trace_end*, if defined, is called after all events have been
143 processed and gives scripts a chance to do end-of-script tasks, such
144 as display results:
145
146----
147sub trace_end
148{
149}
150----
151
152*trace_unhandled*, if defined, is called after for any event that
153 doesn't have a handler explicitly defined for it.  The standard set
154 of common arguments are passed into it:
155
156----
157sub trace_unhandled
158{
159    my ($event_name, $context, $common_cpu, $common_secs,
160        $common_nsecs, $common_pid, $common_comm) = @_;
161}
162----
163
164The remaining sections provide descriptions of each of the available
165built-in perf trace Perl modules and their associated functions.
166
167AVAILABLE MODULES AND FUNCTIONS
168-------------------------------
169
170The following sections describe the functions and variables available
171via the various Perf::Trace::* Perl modules.  To use the functions and
172variables from the given module, add the corresponding 'use
173Perf::Trace::XXX' line to your perf trace script.
174
175Perf::Trace::Core Module
176~~~~~~~~~~~~~~~~~~~~~~~~
177
178These functions provide some essential functions to user scripts.
179
180The *flag_str* and *symbol_str* functions provide human-readable
181strings for flag and symbolic fields.  These correspond to the strings
182and values parsed from the 'print fmt' fields of the event format
183files:
184
185  flag_str($event_name, $field_name, $field_value) - returns the string represention corresponding to $field_value for the flag field $field_name of event $event_name
186  symbol_str($event_name, $field_name, $field_value) - returns the string represention corresponding to $field_value for the symbolic field $field_name of event $event_name
187
188Perf::Trace::Context Module
189~~~~~~~~~~~~~~~~~~~~~~~~~~~
190
191Some of the 'common' fields in the event format file aren't all that
192common, but need to be made accessible to user scripts nonetheless.
193
194Perf::Trace::Context defines a set of functions that can be used to
195access this data in the context of the current event.  Each of these
196functions expects a $context variable, which is the same as the
197$context variable passed into every event handler as the second
198argument.
199
200 common_pc($context) - returns common_preempt count for the current event
201 common_flags($context) - returns common_flags for the current event
202 common_lock_depth($context) - returns common_lock_depth for the current event
203
204Perf::Trace::Util Module
205~~~~~~~~~~~~~~~~~~~~~~~~
206
207Various utility functions for use with perf trace:
208
209  nsecs($secs, $nsecs) - returns total nsecs given secs/nsecs pair
210  nsecs_secs($nsecs) - returns whole secs portion given nsecs
211  nsecs_nsecs($nsecs) - returns nsecs remainder given nsecs
212  nsecs_str($nsecs) - returns printable string in the form secs.nsecs
213  avg($total, $n) - returns average given a sum and a total number of values
214
215SEE ALSO
216--------
217linkperf:perf-trace[1]
218