xref: /linux/tools/perf/Documentation/perf-probe.txt (revision 906fd46a65383cd639e5eec72a047efc33045d86)
1perf-probe(1)
2=============
3
4NAME
5----
6perf-probe - Define new dynamic tracepoints
7
8SYNOPSIS
9--------
10[verse]
11'perf probe' [options] --add='PROBE' [...]
12or
13'perf probe' [options] PROBE
14or
15'perf probe' [options] --del='[GROUP:]EVENT' [...]
16or
17'perf probe' --list[=[GROUP:]EVENT]
18or
19'perf probe' [options] --line='LINE'
20or
21'perf probe' [options] --vars='PROBEPOINT'
22or
23'perf probe' [options] --funcs
24or
25'perf probe' [options] --definition='PROBE' [...]
26
27DESCRIPTION
28-----------
29This command defines dynamic tracepoint events, by symbol and registers
30without debuginfo, or by C expressions (C line numbers, C function names,
31and C local variables) with debuginfo.
32
33
34OPTIONS
35-------
36-k::
37--vmlinux=PATH::
38	Specify vmlinux path which has debuginfo (Dwarf binary).
39	Only when using this with --definition, you can give an offline
40	vmlinux file.
41
42-m::
43--module=MODNAME|PATH::
44	Specify module name in which perf-probe searches probe points
45	or lines. If a path of module file is passed, perf-probe
46	treat it as an offline module (this means you can add a probe on
47        a module which has not been loaded yet).
48
49-s::
50--source=PATH::
51	Specify path to kernel source.
52
53-v::
54--verbose::
55        Be more verbose (show parsed arguments, etc).
56	Can not use with -q.
57
58-q::
59--quiet::
60	Do not show any warnings or messages.
61	Can not use with -v.
62
63-a::
64--add=::
65	Define a probe event (see PROBE SYNTAX for detail).
66
67-d::
68--del=::
69	Delete probe events. This accepts glob wildcards('*', '?') and character
70	classes(e.g. [a-z], [!A-Z]).
71
72-l::
73--list[=[GROUP:]EVENT]::
74	List up current probe events. This can also accept filtering patterns of
75	event names.
76	When this is used with --cache, perf shows all cached probes instead of
77	the live probes.
78
79-L::
80--line=::
81	Show source code lines which can be probed. This needs an argument
82	which specifies a range of the source code. (see LINE SYNTAX for detail)
83
84-V::
85--vars=::
86	Show available local variables at given probe point. The argument
87	syntax is same as PROBE SYNTAX, but NO ARGs.
88
89--externs::
90	(Only for --vars) Show external defined variables in addition to local
91	variables.
92
93--no-inlines::
94	(Only for --add) Search only for non-inlined functions. The functions
95	which do not have instances are ignored.
96
97-F::
98--funcs[=FILTER]::
99	Show available functions in given module or kernel. With -x/--exec,
100	can also list functions in a user space executable / shared library.
101	This also can accept a FILTER rule argument.
102
103-D::
104--definition=::
105	Show trace-event definition converted from given probe-event instead
106	of write it into tracing/[k,u]probe_events.
107
108--filter=FILTER::
109	(Only for --vars and --funcs) Set filter. FILTER is a combination of glob
110	pattern, see FILTER PATTERN for detail.
111	Default FILTER is "!__k???tab_* & !__crc_*" for --vars, and "!_*"
112	for --funcs.
113	If several filters are specified, only the last filter is used.
114
115-f::
116--force::
117	Forcibly add events with existing name.
118
119-n::
120--dry-run::
121	Dry run. With this option, --add and --del doesn't execute actual
122	adding and removal operations.
123
124--cache::
125	(With --add) Cache the probes. Any events which successfully added
126	are also stored in the cache file.
127	(With --list) Show cached probes.
128	(With --del) Remove cached probes.
129
130--max-probes=NUM::
131	Set the maximum number of probe points for an event. Default is 128.
132
133--target-ns=PID:
134	Obtain mount namespace information from the target pid.  This is
135	used when creating a uprobe for a process that resides in a
136	different mount namespace from the perf(1) utility.
137
138-x::
139--exec=PATH::
140	Specify path to the executable or shared library file for user
141	space tracing. Can also be used with --funcs option.
142
143--demangle::
144	Demangle application symbols. --no-demangle is also available
145	for disabling demangling.
146
147--demangle-kernel::
148	Demangle kernel symbols. --no-demangle-kernel is also available
149	for disabling kernel demangling.
150
151In absence of -m/-x options, perf probe checks if the first argument after
152the options is an absolute path name. If its an absolute path, perf probe
153uses it as a target module/target user space binary to probe.
154
155PROBE SYNTAX
156------------
157Probe points are defined by following syntax.
158
159    1) Define event based on function name
160     [[GROUP:]EVENT=]FUNC[@SRC][:RLN|+OFFS|%return|;PTN] [ARG ...]
161
162    2) Define event based on source file with line number
163     [[GROUP:]EVENT=]SRC:ALN [ARG ...]
164
165    3) Define event based on source file with lazy pattern
166     [[GROUP:]EVENT=]SRC;PTN [ARG ...]
167
168    4) Pre-defined SDT events or cached event with name
169     %[sdt_PROVIDER:]SDTEVENT
170     or,
171     sdt_PROVIDER:SDTEVENT
172
173'EVENT' specifies the name of new event, if omitted, it will be set the name of the probed function, and for return probes, a "\_\_return" suffix is automatically added to the function name. You can also specify a group name by 'GROUP', if omitted, set 'probe' is used for kprobe and 'probe_<bin>' is used for uprobe.
174Note that using existing group name can conflict with other events. Especially, using the group name reserved for kernel modules can hide embedded events in the
175modules.
176'FUNC' specifies a probed function name, and it may have one of the following options; '+OFFS' is the offset from function entry address in bytes, ':RLN' is the relative-line number from function entry line, and '%return' means that it probes function return. And ';PTN' means lazy matching pattern (see LAZY MATCHING). Note that ';PTN' must be the end of the probe point definition.  In addition, '@SRC' specifies a source file which has that function.
177It is also possible to specify a probe point by the source line number or lazy matching by using 'SRC:ALN' or 'SRC;PTN' syntax, where 'SRC' is the source file path, ':ALN' is the line number and ';PTN' is the lazy matching pattern.
178'ARG' specifies the arguments of this probe point, (see PROBE ARGUMENT).
179'SDTEVENT' and 'PROVIDER' is the pre-defined event name which is defined by user SDT (Statically Defined Tracing) or the pre-cached probes with event name.
180Note that before using the SDT event, the target binary (on which SDT events are defined) must be scanned by linkperf:perf-buildid-cache[1] to make SDT events as cached events.
181
182For details of the SDT, see below.
183https://sourceware.org/gdb/onlinedocs/gdb/Static-Probe-Points.html
184
185ESCAPED CHARACTER
186-----------------
187
188In the probe syntax, '=', '@', '+', ':' and ';' are treated as a special character. You can use a backslash ('\') to escape the special characters.
189This is useful if you need to probe on a specific versioned symbols, like @GLIBC_... suffixes, or also you need to specify a source file which includes the special characters.
190Note that usually single backslash is consumed by shell, so you might need to pass double backslash (\\) or wrapping with single quotes (\'AAA\@BBB').
191See EXAMPLES how it is used.
192
193PROBE ARGUMENT
194--------------
195Each probe argument follows below syntax.
196
197 [NAME=]LOCALVAR|$retval|%REG|@SYMBOL[:TYPE][@user]
198
199'NAME' specifies the name of this argument (optional). You can use the name of local variable, local data structure member (e.g. var->field, var.field2), local array with fixed index (e.g. array[1], var->array[0], var->pointer[2]), or kprobe-tracer argument format (e.g. $retval, %ax, etc). Note that the name of this argument will be set as the last member name if you specify a local data structure member (e.g. field2 for 'var->field1.field2'.)
200'$vars' and '$params' special arguments are also available for NAME, '$vars' is expanded to the local variables (including function parameters) which can access at given probe point. '$params' is expanded to only the function parameters.
201'TYPE' casts the type of this argument (optional). If omitted, perf probe automatically set the type based on debuginfo (*). Currently, basic types (u8/u16/u32/u64/s8/s16/s32/s64), hexadecimal integers (x/x8/x16/x32/x64), signedness casting (u/s), "string" and bitfield are supported. (see TYPES for detail)
202On x86 systems %REG is always the short form of the register: for example %AX. %RAX or %EAX is not valid.
203"@user" is a special attribute which means the LOCALVAR will be treated as a user-space memory. This is only valid for kprobe event.
204
205TYPES
206-----
207Basic types (u8/u16/u32/u64/s8/s16/s32/s64) and hexadecimal integers (x8/x16/x32/x64) are integer types. Prefix 's' and 'u' means those types are signed and unsigned respectively, and 'x' means that is shown in hexadecimal format. Traced arguments are shown in decimal (sNN/uNN) or hex (xNN). You can also use 's' or 'u' to specify only signedness and leave its size auto-detected by perf probe. Moreover, you can use 'x' to explicitly specify to be shown in hexadecimal (the size is also auto-detected).
208String type is a special type, which fetches a "null-terminated" string from kernel space. This means it will fail and store NULL if the string container has been paged out. You can specify 'string' type only for the local variable or structure member which is an array of or a pointer to 'char' or 'unsigned char' type.
209Bitfield is another special type, which takes 3 parameters, bit-width, bit-offset, and container-size (usually 32). The syntax is;
210
211 b<bit-width>@<bit-offset>/<container-size>
212
213LINE SYNTAX
214-----------
215Line range is described by following syntax.
216
217 "FUNC[@SRC][:RLN[+NUM|-RLN2]]|SRC[:ALN[+NUM|-ALN2]]"
218
219FUNC specifies the function name of showing lines. 'RLN' is the start line
220number from function entry line, and 'RLN2' is the end line number. As same as
221probe syntax, 'SRC' means the source file path, 'ALN' is start line number,
222and 'ALN2' is end line number in the file. It is also possible to specify how
223many lines to show by using 'NUM'. Moreover, 'FUNC@SRC' combination is good
224for searching a specific function when several functions share same name.
225So, "source.c:100-120" shows lines between 100th to 120th in source.c file. And "func:10+20" shows 20 lines from 10th line of func function.
226
227LAZY MATCHING
228-------------
229The lazy line matching is similar to glob matching but ignoring spaces in both of pattern and target. So this accepts wildcards('*', '?') and character classes(e.g. [a-z], [!A-Z]).
230
231e.g.
232 'a=*' can matches 'a=b', 'a = b', 'a == b' and so on.
233
234This provides some sort of flexibility and robustness to probe point definitions against minor code changes. For example, actual 10th line of schedule() can be moved easily by modifying schedule(), but the same line matching 'rq=cpu_rq*' may still exist in the function.)
235
236FILTER PATTERN
237--------------
238The filter pattern is a glob matching pattern(s) to filter variables.
239In addition, you can use "!" for specifying filter-out rule. You also can give several rules combined with "&" or "|", and fold those rules as one rule by using "(" ")".
240
241e.g.
242 With --filter "foo* | bar*", perf probe -V shows variables which start with "foo" or "bar".
243 With --filter "!foo* & *bar", perf probe -V shows variables which don't start with "foo" and end with "bar", like "fizzbar". But "foobar" is filtered out.
244
245EXAMPLES
246--------
247Display which lines in schedule() can be probed:
248
249 ./perf probe --line schedule
250
251Add a probe on schedule() function 12th line with recording cpu local variable:
252
253 ./perf probe schedule:12 cpu
254 or
255 ./perf probe --add='schedule:12 cpu'
256
257Add one or more probes which has the name start with "schedule".
258
259 ./perf probe schedule*
260 or
261 ./perf probe --add='schedule*'
262
263Add probes on lines in schedule() function which calls update_rq_clock().
264
265 ./perf probe 'schedule;update_rq_clock*'
266 or
267 ./perf probe --add='schedule;update_rq_clock*'
268
269Delete all probes on schedule().
270
271 ./perf probe --del='schedule*'
272
273Add probes at zfree() function on /bin/zsh
274
275 ./perf probe -x /bin/zsh zfree or ./perf probe /bin/zsh zfree
276
277Add probes at malloc() function on libc
278
279 ./perf probe -x /lib/libc.so.6 malloc or ./perf probe /lib/libc.so.6 malloc
280
281Add a uprobe to a target process running in a different mount namespace
282
283 ./perf probe --target-ns <target pid> -x /lib64/libc.so.6 malloc
284
285Add a USDT probe to a target process running in a different mount namespace
286
287 ./perf probe --target-ns <target pid> -x /usr/lib/jvm/java-1.8.0-openjdk-1.8.0.121-0.b13.el7_3.x86_64/jre/lib/amd64/server/libjvm.so %sdt_hotspot:thread__sleep__end
288
289Add a probe on specific versioned symbol by backslash escape
290
291 ./perf probe -x /lib64/libc-2.25.so 'malloc_get_state\@GLIBC_2.2.5'
292
293Add a probe in a source file using special characters by backslash escape
294
295 ./perf probe -x /opt/test/a.out 'foo\+bar.c:4'
296
297
298PERMISSIONS AND SYSCTL
299----------------------
300Since perf probe depends on ftrace (tracefs) and kallsyms (/proc/kallsyms), you have to care about the permission and some sysctl knobs.
301
302 - Since tracefs and kallsyms requires root or privileged user to access it, the following perf probe commands also require it; --add, --del, --list (except for --cache option)
303
304 - The system admin can remount the tracefs with 755 (`sudo mount -o remount,mode=755 /sys/kernel/tracing/`) to allow unprivileged user to run the perf probe --list command.
305
306 - /proc/sys/kernel/kptr_restrict = 2 (restrict all users) also prevents perf probe to retrieve the important information from kallsyms. You also need to set to 1 (restrict non CAP_SYSLOG users) for the above commands. Since the user-space probe doesn't need to access kallsyms, this is only for probing the kernel function (kprobes).
307
308 - Since the perf probe commands read the vmlinux (for kernel) and/or the debuginfo file (including user-space application), you need to ensure that you can read those files.
309
310
311SEE ALSO
312--------
313linkperf:perf-trace[1], linkperf:perf-record[1], linkperf:perf-buildid-cache[1]
314