Commit | Line | Data |
---|---|---|
eba411c6 MD |
1 | .TH "LTTNG-UST" "3" "February 16, 2012" "" "" |
2 | ||
3 | .SH "NAME" | |
77ca1460 | 4 | lttng-ust \(em Linux Trace Toolkit Next Generation User-Space Tracer 2.x |
eba411c6 MD |
5 | |
6 | .SH "SYNOPSIS" | |
7 | ||
8 | .PP | |
9 | .nf | |
10 | Link liblttng-ust.so with applications, following this manpage. | |
11 | .fi | |
12 | .SH "DESCRIPTION" | |
13 | ||
14 | .PP | |
15 | LTTng-UST, the Linux Trace Toolkit Next Generation Userspace Tracer, is | |
16 | port of the low-overhead tracing capabilities of the LTTng kernel tracer | |
17 | to user-space. The library "liblttng-ust" enables tracing of | |
18 | applications and libraries. | |
19 | ||
20 | .SH "USAGE" | |
21 | .PP | |
22 | The simple way to generate the lttng-ust tracepoint probes is to use the | |
23 | lttng-gen-tp(1) tool. See the lttng-gen-tp(1) manpage for explanation. | |
24 | .PP | |
25 | ||
26 | .PP | |
27 | Here is the way to do it manually, without the lttng-gen-tp(1) helper | |
28 | script, through an example: | |
29 | .PP | |
30 | ||
31 | .SH "CREATION OF TRACEPOINT PROVIDER" | |
32 | ||
33 | .nf | |
34 | ||
35 | To create a tracepoint provider, within a build tree similar to | |
36 | examples/easy-ust installed with lttng-ust documentation, a | |
37 | sample_component_provider.h for the general layout. This manpage will | |
38 | focus on the various types that can be recorded into a trace event: | |
39 | ||
40 | TRACEPOINT_EVENT( | |
41 | /* | |
42 | * provider name, not a variable but a string starting with a | |
43 | * letter and containing either letters, numbers or underscores. | |
44 | * Needs to be the same as TRACEPOINT_PROVIDER. Needs to | |
45 | * follow the namespacing guide-lines in lttng/tracepoint.h: | |
46 | * | |
47 | * Must be included before include tracepoint provider | |
48 | * ex.: project_event | |
49 | * ex.: project_component_event | |
50 | * | |
51 | * Optional company name goes here | |
52 | * ex.: com_efficios_project_component_event | |
53 | * | |
54 | * In this example, "sample" is the project, and "component" is the | |
55 | * component. | |
56 | */ | |
57 | sample_component, | |
58 | ||
59 | /* | |
60 | * tracepoint name, same format as sample provider. Does not | |
61 | * need to be declared before. in this case the name is | |
62 | * "message" | |
63 | */ | |
64 | message, | |
65 | ||
66 | /* | |
67 | * TP_ARGS macro contains the arguments passed for the tracepoint | |
68 | * it is in the following format | |
69 | * TP_ARGS(type1, name1, type2, name2, ... type10, | |
70 | name10) | |
71 | * where there can be from zero to ten elements. | |
72 | * typeN is the datatype, such as int, struct or double **. | |
73 | * name is the variable name (in "int myInt" the name would be | |
74 | * myint) | |
75 | * TP_ARGS() is valid to mean no arguments | |
76 | * TP_ARGS(void) is valid too | |
77 | */ | |
78 | TP_ARGS(int, anint, int, netint, long *, values, | |
79 | char *, text, size_t, textlen, | |
80 | double, doublearg, float, floatarg), | |
81 | ||
82 | /* | |
83 | * TP_FIELDS describes how to write the fields of the trace event. | |
84 | * You can put expressions in the "argument expression" area, | |
85 | * typically using the input arguments from TP_ARGS. | |
86 | */ | |
87 | TP_FIELDS( | |
88 | /* | |
89 | * ctf_integer: standard integer field. | |
90 | * args: (type, field name, argument expression) | |
91 | */ | |
92 | ctf_integer(int, intfield, anint) | |
93 | ctf_integer(long, longfield, anint) | |
94 | ||
95 | /* | |
96 | * ctf_integer_hex: integer field printed as hexadecimal. | |
97 | * args: (type, field name, argument expression) | |
98 | */ | |
99 | ctf_integer_hex(int, intfield2, anint) | |
100 | ||
101 | /* | |
102 | * ctf_integer_network: integer field in network byte | |
103 | * order. (_hex: printed as hexadecimal too) | |
104 | * args: (type, field name, argument expression) | |
105 | */ | |
106 | ctf_integer_network(int, netintfield, netint) | |
107 | ctf_integer_network_hex(int, netintfieldhex, netint) | |
108 | ||
109 | /* | |
110 | * ctf_array: a statically-sized array. | |
111 | * args: (type, field name, argument expression, value) | |
112 | */ | |
113 | ctf_array(long, arrfield1, values, 3) | |
114 | ||
115 | /* | |
116 | * ctf_array_text: a statically-sized array, printed as | |
117 | * a string. No need to be terminated by a null | |
118 | * character. | |
2f65f1f5 | 119 | * Behavior is undefined if "text" argument is NULL. |
eba411c6 MD |
120 | */ |
121 | ctf_array_text(char, arrfield2, text, 10) | |
122 | ||
123 | /* | |
124 | * ctf_sequence: a dynamically-sized array. | |
125 | * args: (type, field name, argument expression, | |
126 | * type of length expression, length expression) | |
efbad5cc MD |
127 | * The "type of length expression" needs to be an |
128 | * unsigned type. As a reminder, "unsigned char" should | |
129 | * be preferred to "char", since the signedness of | |
130 | * "char" is implementation-defined. | |
2f65f1f5 | 131 | * Behavior is undefined if "text" argument is NULL. |
eba411c6 MD |
132 | */ |
133 | ctf_sequence(char, seqfield1, text, | |
134 | size_t, textlen) | |
135 | ||
136 | /* | |
137 | * ctf_sequence_text: a dynamically-sized array, printed | |
138 | * as string. No need to be null-terminated. | |
2f65f1f5 | 139 | * Behavior is undefined if "text" argument is NULL. |
eba411c6 MD |
140 | */ |
141 | ctf_sequence_text(char, seqfield2, text, | |
142 | size_t, textlen) | |
143 | ||
144 | /* | |
145 | * ctf_string: null-terminated string. | |
146 | * args: (field name, argument expression) | |
2f65f1f5 | 147 | * Behavior is undefined if "text" argument is NULL. |
eba411c6 MD |
148 | */ |
149 | ctf_string(stringfield, text) | |
150 | ||
151 | /* | |
152 | * ctf_float: floating-point number. | |
153 | * args: (type, field name, argument expression) | |
154 | */ | |
155 | ctf_float(float, floatfield, floatarg) | |
156 | ctf_float(double, doublefield, doublearg) | |
157 | ) | |
158 | ) | |
159 | .fi | |
160 | ||
5883c06f MD |
161 | .SH "ASSIGNING LOGLEVEL TO EVENTS" |
162 | ||
163 | .nf | |
164 | ||
165 | Optionally, a loglevel can be assigned to a TRACEPOINT_EVENT using the | |
166 | following construct: | |
167 | ||
168 | TRACEPOINT_LOGLEVEL(< [com_company_]project[_component] >, | |
169 | < event >, < loglevel_name >) | |
170 | ||
171 | The first field is the provider name, the second field is the name of | |
172 | the tracepoint, and the third field is the loglevel name. A | |
173 | TRACEPOINT_EVENT should be declared prior to the the TRACEPOINT_LOGLEVEL | |
174 | for a given tracepoint name. The TRACEPOINT_PROVIDER must be already | |
175 | declared before declaring a TRACEPOINT_LOGLEVEL. | |
176 | ||
177 | The loglevels go from 0 to 14. Higher numbers imply the most verbosity | |
178 | (higher event throughput expected. | |
179 | ||
180 | Loglevels 0 through 6, and loglevel 14, match syslog(3) loglevels | |
181 | semantic. Loglevels 7 through 13 offer more fine-grained selection of | |
182 | debug information. | |
183 | ||
184 | TRACE_EMERG 0 | |
185 | system is unusable | |
186 | ||
187 | TRACE_ALERT 1 | |
188 | action must be taken immediately | |
189 | ||
190 | TRACE_CRIT 2 | |
191 | critical conditions | |
192 | ||
193 | TRACE_ERR 3 | |
194 | error conditions | |
195 | ||
196 | TRACE_WARNING 4 | |
197 | warning conditions | |
198 | ||
199 | TRACE_NOTICE 5 | |
200 | normal, but significant, condition | |
201 | ||
202 | TRACE_INFO 6 | |
203 | informational message | |
204 | ||
205 | TRACE_DEBUG_SYSTEM 7 | |
206 | debug information with system-level scope (set of programs) | |
207 | ||
208 | TRACE_DEBUG_PROGRAM 8 | |
209 | debug information with program-level scope (set of processes) | |
210 | ||
211 | TRACE_DEBUG_PROCESS 9 | |
212 | debug information with process-level scope (set of modules) | |
213 | ||
214 | TRACE_DEBUG_MODULE 10 | |
215 | debug information with module (executable/library) scope (set of | |
216 | units) | |
217 | ||
218 | TRACE_DEBUG_UNIT 11 | |
219 | debug information with compilation unit scope (set of functions) | |
220 | ||
221 | TRACE_DEBUG_FUNCTION 12 | |
222 | debug information with function-level scope | |
223 | ||
224 | TRACE_DEBUG_LINE 13 | |
225 | debug information with line-level scope (TRACEPOINT_EVENT default) | |
226 | ||
227 | TRACE_DEBUG 14 | |
228 | debug-level message (trace_printf default) | |
229 | ||
230 | See lttng(1) for information on how to use LTTng-UST loglevels. | |
231 | ||
232 | .fi | |
233 | ||
eba411c6 MD |
234 | .SH "ADDING TRACEPOINTS TO YOUR CODE" |
235 | ||
236 | .nf | |
237 | ||
238 | Include the provider header in each C files you plan to instrument, | |
239 | following the building/linking directives in the next section. | |
240 | ||
241 | For instance, add within a function: | |
242 | ||
243 | tracepoint(ust_tests_hello, tptest, i, netint, values, | |
244 | text, strlen(text), dbl, flt); | |
245 | ||
246 | As a call to the tracepoint. It will only be activated when requested by | |
247 | lttng(1) through lttng-sessiond(8). | |
248 | ||
d646ca64 MD |
249 | Even though LTTng-UST supports tracepoint() call site duplicates having |
250 | the same provider and event name, it is recommended to use a | |
251 | provider event name pair only once within the source code to help | |
252 | mapping events back to their call sites when analyzing the trace. | |
eba411c6 MD |
253 | .fi |
254 | ||
255 | .SH "BUILDING/LINKING THE TRACEPOINT PROVIDER" | |
256 | ||
257 | .nf | |
258 | There are 2 ways to compile the Tracepoint Provider with the | |
259 | application: either statically or dynamically. Please follow | |
260 | carefully: | |
261 | ||
262 | 1.1) Compile the Tracepoint provider with the application, either | |
263 | directly or through a static library (.a): | |
264 | - Into exactly one object of your application: define | |
265 | "TRACEPOINT_DEFINE" and include the tracepoint provider. | |
9b6435af | 266 | - Use "\-I." for the compilation unit containing the tracepoint |
eba411c6 | 267 | provider include (e.g. tp.c). |
9b6435af | 268 | - Link application with "\-ldl". |
eba411c6 | 269 | - If building the provider directly into the application, |
9b6435af | 270 | link the application with "\-llttng-ust". |
eba411c6 | 271 | - If building a static library for the provider, link the static |
38f5f440 | 272 | library with "\-llttng-ust". |
eba411c6 MD |
273 | - Include the tracepoint provider header into all C files using |
274 | the provider. | |
275 | - Example: | |
13fb2d2c | 276 | - tests/hello/ hello.c tp.c ust_tests_hello.h Makefile.example |
eba411c6 MD |
277 | |
278 | 2) Compile the Tracepoint Provider separately from the application, | |
279 | using dynamic linking: | |
280 | - Into exactly one object of your application: define | |
281 | "TRACEPOINT_DEFINE" _and_ also define | |
282 | "TRACEPOINT_PROBE_DYNAMIC_LINKAGE", then include the tracepoint | |
283 | provider header. | |
284 | - Include the tracepoint provider header into all instrumented C | |
285 | files that use the provider. | |
9b6435af AM |
286 | - Compile the tracepoint provider with "\-I.". |
287 | - Link the tracepoint provider with "\-llttng-ust". | |
288 | - Link application with "\-ldl". | |
eba411c6 MD |
289 | - Set a LD_PRELOAD environment to preload the tracepoint provider |
290 | shared object before starting the application when tracing is | |
13fb2d2c JG |
291 | needed. Another way is to dlopen the tracepoint probe when needed |
292 | by the application. | |
eba411c6 | 293 | - Example: |
60302adc | 294 | - doc/examples/demo demo.c tp*.c ust_tests_demo*.h demo-trace |
eba411c6 | 295 | |
13fb2d2c JG |
296 | - Note about dlclose() usage: it is not safe to use dlclose on a |
297 | provider shared object that is being actively used for tracing due | |
298 | to a lack of reference counting from lttng-ust to the used shared | |
299 | object. | |
eba411c6 MD |
300 | - Enable instrumentation and control tracing with the "lttng" command |
301 | from lttng-tools. See lttng-tools doc/quickstart.txt. | |
2bda849d MD |
302 | - Note for C++ support: although an application instrumented with |
303 | tracepoints can be compiled with g++, tracepoint probes should be | |
304 | compiled with gcc (only tested with gcc so far). | |
eba411c6 MD |
305 | |
306 | .fi | |
307 | ||
0a7c55a5 MD |
308 | .SH "USING LTTNG UST WITH DAEMONS" |
309 | ||
310 | .nf | |
311 | Some extra care is needed when using liblttng-ust with daemon | |
312 | applications that call fork(), clone(), or BSD rfork() without a | |
313 | following exec() family system call. The library "liblttng-ust-fork.so" | |
314 | needs to be preloaded for the application (launch with e.g. | |
315 | LD_PRELOAD=liblttng-ust-fork.so appname). | |
316 | ||
317 | .fi | |
318 | ||
94c9c48d MD |
319 | .SH "CONTEXT" |
320 | ||
321 | .PP | |
322 | Context information can be prepended by the tracer before each, or some, | |
323 | events. The following context information is supported by LTTng-UST: | |
324 | .PP | |
325 | ||
326 | .PP | |
327 | .IP "vtid" | |
328 | Virtual thread ID: thread ID as seen from the point of view of the | |
329 | process namespace. | |
330 | .PP | |
331 | ||
332 | .PP | |
333 | .IP "vpid" | |
334 | Virtual process ID: process ID as seen from the point of view of the | |
335 | process namespace. | |
336 | .PP | |
337 | ||
338 | .PP | |
339 | .IP "procname" | |
340 | Thread name, as set by exec() or prctl(). It is recommended that | |
341 | programs set their thread name with prctl() before hitting the first | |
342 | tracepoint for that thread. | |
343 | .PP | |
344 | ||
345 | .PP | |
346 | .IP "pthread_id" | |
347 | Pthread identifier. Can be used on architectures where pthread_t maps | |
348 | nicely to an unsigned long type. | |
349 | .PP | |
350 | ||
eba411c6 MD |
351 | .SH "ENVIRONMENT VARIABLES" |
352 | ||
353 | .PP | |
354 | .IP "LTTNG_UST_DEBUG" | |
355 | Activate liblttng-ust debug output. | |
356 | .PP | |
357 | .IP "LTTNG_UST_REGISTER_TIMEOUT" | |
358 | The environment variable "LTTNG_UST_REGISTER_TIMEOUT" can be used to | |
359 | specify how long the applications should wait for sessiond | |
360 | "registration done" command before proceeding to execute the main | |
361 | program. The default is 3000ms (3 seconds). The timeout value is | |
362 | specified in milliseconds. The value 0 means "don't wait". The value | |
9b6435af | 363 | \-1 means "wait forever". Setting this environment variable to 0 is |
eba411c6 MD |
364 | recommended for applications with time constraints on the process |
365 | startup time. | |
366 | .PP | |
367 | ||
368 | .SH "SEE ALSO" | |
369 | ||
370 | .PP | |
d1eaa4d7 PW |
371 | lttng-gen-tp(1), lttng(1), babeltrace(1), lttng-ust-cyg-profile(3), |
372 | lttng-sessiond(8) | |
eba411c6 MD |
373 | .PP |
374 | .SH "BUGS" | |
375 | ||
376 | .PP | |
60302adc | 377 | No known bugs at this point. |
eba411c6 MD |
378 | |
379 | If you encounter any issues or usability problem, please report it on | |
380 | our mailing list <lttng-dev@lists.lttng.org> to help improve this | |
381 | project. | |
382 | .SH "CREDITS" | |
383 | ||
384 | liblttng-ust is distributed under the GNU Lesser General Public License | |
385 | version 2.1. The headers are distributed under the MIT license. | |
386 | .PP | |
387 | See http://lttng.org for more information on the LTTng project. | |
388 | .PP | |
389 | Mailing list for support and development: <lttng-dev@lists.lttng.org>. | |
390 | .PP | |
391 | You can find us on IRC server irc.oftc.net (OFTC) in #lttng. | |
392 | .PP | |
393 | .SH "THANKS" | |
394 | ||
395 | Thanks to Ericsson for funding this work, providing real-life use-cases, | |
396 | and testing. | |
397 | ||
398 | Special thanks to Michel Dagenais and the DORSAL laboratory at | |
399 | Polytechnique de Montreal for the LTTng journey. | |
400 | .PP | |
401 | .SH "AUTHORS" | |
402 | ||
403 | .PP | |
404 | liblttng-ust was originally written by Mathieu Desnoyers, with additional | |
405 | contributions from various other people. It is currently maintained by | |
406 | Mathieu Desnoyers <mathieu.desnoyers@efficios.com>. | |
407 | .PP |