1 The LTTng Documentation
2 =======================
3 Philippe Proulx <pproulx@efficios.com>
7 include::../common/copyright.txt[]
10 include::../common/welcome.txt[]
13 include::../common/audience.txt[]
17 === What's in this documentation?
19 The LTTng Documentation is divided into the following sections:
21 * **<<nuts-and-bolts,Nuts and bolts>>** explains the
22 rudiments of software tracing and the rationale behind the
25 You can skip this section if you’re familiar with software tracing and
26 with the LTTng project.
28 * **<<installing-lttng,Installation>>** describes the steps to
29 install the LTTng packages on common Linux distributions and from
32 You can skip this section if you already properly installed LTTng on
35 * **<<getting-started,Quick start>>** is a concise guide to
36 getting started quickly with LTTng kernel and user space tracing.
38 We recommend this section if you're new to LTTng or to software tracing
41 You can skip this section if you're not new to LTTng.
43 * **<<core-concepts,Core concepts>>** explains the concepts at
46 It's a good idea to become familiar with the core concepts
47 before attempting to use the toolkit.
49 * **<<plumbing,Components of LTTng>>** describes the various components
50 of the LTTng machinery, like the daemons, the libraries, and the
51 command-line interface.
52 * **<<instrumenting,Instrumentation>>** shows different ways to
53 instrument user applications and the Linux kernel.
55 Instrumenting source code is essential to provide a meaningful
58 You can skip this section if you do not have a programming background.
60 * **<<controlling-tracing,Tracing control>>** is divided into topics
61 which demonstrate how to use the vast array of features that
62 LTTng{nbsp}{revision} offers.
63 * **<<reference,Reference>>** contains reference tables.
64 * **<<glossary,Glossary>>** is a specialized dictionary of terms related
65 to LTTng or to the field of software tracing.
68 include::../common/convention.txt[]
71 include::../common/acknowledgements.txt[]
75 == What's new in LTTng {revision}?
77 * **Tracing control**:
78 ** You can attach <<java-application-context,Java application-specific
79 context fields>> to a <<channel,channel>> with the
80 man:lttng-add-context(1) command:
85 lttng add-context --jul --type='$app.retriever:cur_msg_id'
89 Here, `$app` is the prefix of all application-specific context fields,
90 `retriever` names a _context information retriever_ defined at the
91 application level, and `cur_msg_id` names a context field read from this
94 Both the `java.util.logging` and Apache log4j <<domain,tracing domains>>
97 ** You can use Java application-specific <<adding-context,context>>
98 fields in the <<enabling-disabling-events,filter expression>> of an
104 lttng enable-event --log4j my_logger \
105 --filter='$app.retriever:cur_msg_id == 23'
109 ** New `lttng status` command which is the equivalent of +lttng list
110 __CUR__+, where +__CUR__+ is the name of the current
111 <<tracing-session,tracing session>>.
113 See man:lttng-status(1).
115 ** New `lttng metadata regenerate` command to
116 <<metadata-regenerate,regenerate the metadata file of an LTTng
117 trace>> at any moment. This command is meant to be used to resample
118 the wall time following a major
119 https://en.wikipedia.org/wiki/Network_Time_Protocol[NTP] correction
120 so that a system which boots with an incorrect wall time can be
121 traced before its wall time is NTP-corrected.
123 See man:lttng-metadata(1).
125 ** New command-line interface warnings when <<event,event records>> or
126 whole sub-buffers are
127 <<channel-overwrite-mode-vs-discard-mode,lost>>. The warning messages
128 are printed when a <<tracing-session,tracing session>> is
129 <<basic-tracing-session-control,stopped>> (man:lttng-stop(1)
132 * **User space tracing**:
133 ** Shared object base address dump in order to map <<event,event
134 records>> to original source location (file and line number).
136 If you attach the `ip` and `vpid` <<adding-context,context fields>> to a
137 user space <<channel,channel>> and if you use the
138 <<liblttng-ust-dl,path:{liblttng-ust-dl.so} helper>>, you can retrieve
139 the source location where a given event record was generated.
141 The http://diamon.org/babeltrace/[Babeltrace] trace viewer supports this
142 state dump and those context fields since version 1.4 to print the
143 source location of a given event record. http://tracecompass.org/[Trace
144 Compass] also supports this since version 2.0.
146 ** A <<java-application,Java application>> which uses
147 `java.util.logging` now adds an LTTng-UST log handler to the desired
150 The previous workflow was to initialize the LTTng-UST Java agent
151 by calling `LTTngAgent.getLTTngAgent()`. This had the effect of adding
152 an LTTng-UST log handler to the root loggers.
154 ** A <<java-application,Java application>> which uses Apache log4j now
155 adds an LTTng-UST log appender to the desired log4j loggers.
157 The previous workflow was to initialize the LTTng-UST Java agent
158 by calling `LTTngAgent.getLTTngAgent()`. This had the effect of adding
159 an LTTng-UST appender to the root loggers.
161 ** Any <<java-application,Java application>> can provide
162 <<java-application-context,dynamic context fields>> while running
163 thanks to a new API provided by the <<lttng-ust-agents,LTTng-UST Java
164 agent>>. You can require LTTng to record specific context fields in
165 event records, and you can use them in the filter expression of
166 <<event,event rules>>.
168 * **Linux kernel tracing**:
169 ** The LTTng kernel modules can now be built into a Linux kernel image,
170 that is, not as loadable modules.
173 https://github.com/lttng/lttng-modules/blob/stable-{revision}/README.md#kernel-built-in-support[`README.md`]
176 ** New instrumentation:
177 *** ARM64 architecture support.
179 *** x86 `irq_vectors`.
180 ** New <<adding-context,context fields>>:
183 *** `need_reschedule`
184 *** `migratable` (specific to RT-Preempt)
185 ** Clock source plugin support for advanced cases where a custom source
186 of time is needed to timestamp LTTng event records.
188 See https://github.com/lttng/lttng-modules/blob/stable-{revision}/lttng-clock.h[`lttng-clock.h`]
189 for an overview of the small API.
192 ** The link:/man[man pages] of the man:lttng(1) command-line tool are
193 split into one man page per command (à la Git), for example:
198 man lttng-enable-event
202 You can also use the `--help` option of any man:lttng(1) command to
205 The content and formatting of all the LTTng man pages has improved
212 What is LTTng? As its name suggests, the _Linux Trace Toolkit: next
213 generation_ is a modern toolkit for tracing Linux systems and
214 applications. So your first question might be:
221 As the history of software engineering progressed and led to what
222 we now take for granted--complex, numerous and
223 interdependent software applications running in parallel on
224 sophisticated operating systems like Linux--the authors of such
225 components, software developers, began feeling a natural
226 urge to have tools that would ensure the robustness and good performance
227 of their masterpieces.
229 One major achievement in this field is, inarguably, the
230 https://www.gnu.org/software/gdb/[GNU debugger (GDB)],
231 an essential tool for developers to find and fix bugs. But even the best
232 debugger won't help make your software run faster, and nowadays, faster
233 software means either more work done by the same hardware, or cheaper
234 hardware for the same work.
236 A _profiler_ is often the tool of choice to identify performance
237 bottlenecks. Profiling is suitable to identify _where_ performance is
238 lost in a given software. The profiler outputs a profile, a statistical
239 summary of observed events, which you may use to discover which
240 functions took the most time to execute. However, a profiler won't
241 report _why_ some identified functions are the bottleneck. Bottlenecks
242 might only occur when specific conditions are met, conditions that are
243 sometimes impossible to capture by a statistical profiler, or impossible
244 to reproduce with an application altered by the overhead of an
245 event-based profiler. For a thorough investigation of software
246 performance issues, a history of execution is essential, with the
247 recorded values of variables and context fields you choose, and
248 with as little influence as possible on the instrumented software. This
249 is where tracing comes in handy.
251 _Tracing_ is a technique used to understand what goes on in a running
252 software system. The software used for tracing is called a _tracer_,
253 which is conceptually similar to a tape recorder. When recording,
254 specific instrumentation points placed in the software source code
255 generate events that are saved on a giant tape: a _trace_ file. You
256 can trace user applications and the operating system at the same time,
257 opening the possibility of resolving a wide range of problems that would
258 otherwise be extremely challenging.
260 Tracing is often compared to _logging_. However, tracers and loggers are
261 two different tools, serving two different purposes. Tracers are
262 designed to record much lower-level events that occur much more
263 frequently than log messages, often in the range of thousands per
264 second, with very little execution overhead. Logging is more appropriate
265 for a very high-level analysis of less frequent events: user accesses,
266 exceptional conditions (errors and warnings, for example), database
267 transactions, instant messaging communications, and such. Simply put,
268 logging is one of the many use cases that can be satisfied with tracing.
270 The list of recorded events inside a trace file can be read manually
271 like a log file for the maximum level of detail, but it is generally
272 much more interesting to perform application-specific analyses to
273 produce reduced statistics and graphs that are useful to resolve a
274 given problem. Trace viewers and analyzers are specialized tools
277 In the end, this is what LTTng is: a powerful, open source set of
278 tools to trace the Linux kernel and user applications at the same time.
279 LTTng is composed of several components actively maintained and
280 developed by its link:/community/#where[community].
283 [[lttng-alternatives]]
284 === Alternatives to noch:{LTTng}
286 Excluding proprietary solutions, a few competing software tracers
289 * https://github.com/dtrace4linux/linux[dtrace4linux] is a port of
290 Sun Microsystems's DTrace to Linux. The cmd:dtrace tool interprets
291 user scripts and is responsible for loading code into the
292 Linux kernel for further execution and collecting the outputted data.
293 * https://en.wikipedia.org/wiki/Berkeley_Packet_Filter[eBPF] is a
294 subsystem in the Linux kernel in which a virtual machine can execute
295 programs passed from the user space to the kernel. You can attach
296 such programs to tracepoints and KProbes thanks to a system call, and
297 they can output data to the user space when executed thanks to
298 different mechanisms (pipe, VM register values, and eBPF maps, to name
300 * https://www.kernel.org/doc/Documentation/trace/ftrace.txt[ftrace]
301 is the de facto function tracer of the Linux kernel. Its user
302 interface is a set of special files in sysfs.
303 * https://perf.wiki.kernel.org/[perf] is
304 a performance analyzing tool for Linux which supports hardware
305 performance counters, tracepoints, as well as other counters and
306 types of probes. perf's controlling utility is the cmd:perf command
308 * http://linux.die.net/man/1/strace[strace]
309 is a command-line utility which records system calls made by a
310 user process, as well as signal deliveries and changes of process
311 state. strace makes use of https://en.wikipedia.org/wiki/Ptrace[ptrace]
312 to fulfill its function.
313 * http://www.sysdig.org/[sysdig], like SystemTap, uses scripts to
314 analyze Linux kernel events. You write scripts, or _chisels_ in
315 sysdig's jargon, in Lua and sysdig executes them while the system is
316 being traced or afterwards. sysdig's interface is the cmd:sysdig
317 command-line tool as well as the curses-based cmd:csysdig tool.
318 * https://sourceware.org/systemtap/[SystemTap] is a Linux kernel and
319 user space tracer which uses custom user scripts to produce plain text
320 traces. SystemTap converts the scripts to the C language, and then
321 compiles them as Linux kernel modules which are loaded to produce
322 trace data. SystemTap's primary user interface is the cmd:stap
325 The main distinctive features of LTTng is that it produces correlated
326 kernel and user space traces, as well as doing so with the lowest
327 overhead amongst other solutions. It produces trace files in the
328 http://diamon.org/ctf[CTF] format, a file format optimized
329 for the production and analyses of multi-gigabyte data.
331 LTTng is the result of more than 10 years of active open source
332 development by a community of passionate developers.
333 LTTng{nbsp}{revision} is currently available on major desktop and server
336 The main interface for tracing control is a single command-line tool
337 named cmd:lttng. The latter can create several tracing sessions, enable
338 and disable events on the fly, filter events efficiently with custom
339 user expressions, start and stop tracing, and much more. LTTng can
340 record the traces on the file system or send them over the network, and
341 keep them totally or partially. You can view the traces once tracing
342 becomes inactive or in real-time.
344 <<installing-lttng,Install LTTng now>> and
345 <<getting-started,start tracing>>!
351 **LTTng** is a set of software <<plumbing,components>> which interact to
352 <<instrumenting,instrument>> the Linux kernel and user applications, and
353 to <<controlling-tracing,control tracing>> (start and stop
354 tracing, enable and disable event rules, and the rest). Those
355 components are bundled into the following packages:
357 * **LTTng-tools**: Libraries and command-line interface to
359 * **LTTng-modules**: Linux kernel modules to instrument and
361 * **LTTng-UST**: Libraries and Java/Python packages to instrument and
362 trace user applications.
364 Most distributions mark the LTTng-modules and LTTng-UST packages as
365 optional when installing LTTng-tools (which is always required). In the
366 following sections, we always provide the steps to install all three,
369 * You only need to install LTTng-modules if you intend to trace the
371 * You only need to install LTTng-UST if you intend to trace user
375 .Availability of LTTng{nbsp}{revision} for major Linux distributions.
377 |Distribution |Available in releases |Alternatives
380 |<<ubuntu,Ubuntu{nbsp}16.10 _Yakkety Yak_>>
381 |LTTng{nbsp}{revision} for Ubuntu{nbsp}14.04 _Trusty Tahr_
382 and Ubuntu{nbsp}16.04 _Xenial Xerus_:
383 <<ubuntu-ppa,use the LTTng Stable{nbsp}{revision} PPA>>.
385 <<building-from-source,Build LTTng{nbsp}{revision} from source>> for
386 other Ubuntu releases.
390 |LTTng{nbsp}{revision} for Fedora{nbsp}25 and Fedora{nbsp}26 (not
393 <<building-from-source,Build LTTng{nbsp}{revision} from source>> for
394 other Fedora releases.
397 |<<debian,Debian "stretch" (testing)>>
399 <<building-from-source,Build LTTng{nbsp}{revision} from source>> for
400 previous Debian releases.
404 |LTTng{nbsp}2.7 for openSUSE Leap{nbsp}42.1.
406 <<building-from-source,Build LTTng{nbsp}{revision} from source>> for
407 other openSUSE releases.
410 |Latest AUR packages.
414 |<<alpine-linux,Alpine Linux "edge">>
415 |LTTng{nbsp}{revision} for Alpine Linux{nbsp}3.5 (not released yet).
417 <<building-from-source,Build LTTng{nbsp}{revision} from source>> for
418 other Alpine Linux releases.
421 |See http://packages.efficios.com/[EfficiOS Enterprise Packages].
426 |LTTng{nbsp}{revision} for Buildroot{nbsp}2016.11 (not released yet).
428 LTTng{nbsp}2.7 for Buildroot{nbsp}2016.02, Buildroot{nbsp}2016.05,
429 and Buildroot{nbsp}2016.08.
431 <<building-from-source,Build LTTng{nbsp}{revision} from source>> for
432 other Buildroot releases.
434 |OpenEmbedded and Yocto
435 |<<oe-yocto,Yocto Project{nbsp}2.2 _Morty_>> (`openembedded-core` layer)
436 |<<building-from-source,Build LTTng{nbsp}{revision} from source>> for
437 other OpenEmbedded releases.
442 === [[ubuntu-official-repositories]]Ubuntu
444 LTTng{nbsp}{revision} is available on Ubuntu{nbsp}16.10 _Yakkety Yak_.
445 For previous releases of Ubuntu, <<ubuntu-ppa,use the LTTng
446 Stable{nbsp}{revision} PPA>>.
448 To install LTTng{nbsp}{revision} on Ubuntu 16.10{nbsp}_Yakkety Yak_:
450 . Install the main LTTng{nbsp}{revision} packages:
455 sudo apt-get install lttng-tools
456 sudo apt-get install lttng-modules-dkms
457 sudo apt-get install liblttng-ust-dev
461 . **If you need to instrument and trace
462 <<java-application,Java applications>>**, install the LTTng-UST
468 sudo apt-get install liblttng-ust-agent-java
472 . **If you need to instrument and trace
473 <<python-application,Python applications>>**, install the
474 LTTng-UST Python agent:
479 sudo apt-get install python3-lttngust
485 ==== noch:{LTTng} Stable {revision} PPA
487 The https://launchpad.net/~lttng/+archive/ubuntu/stable-{revision}[LTTng
488 Stable{nbsp}{revision} PPA] offers the latest stable
489 LTTng{nbsp}{revision} packages for:
491 * Ubuntu{nbsp}14.04 _Trusty Tahr_
492 * Ubuntu{nbsp}16.04 _Xenial Xerus_
494 To install LTTng{nbsp}{revision} from the LTTng Stable{nbsp}{revision} PPA:
496 . Add the LTTng Stable{nbsp}{revision} PPA repository and update the
502 sudo apt-add-repository ppa:lttng/stable-2.8
507 . Install the main LTTng{nbsp}{revision} packages:
512 sudo apt-get install lttng-tools
513 sudo apt-get install lttng-modules-dkms
514 sudo apt-get install liblttng-ust-dev
518 . **If you need to instrument and trace
519 <<java-application,Java applications>>**, install the LTTng-UST
525 sudo apt-get install liblttng-ust-agent-java
529 . **If you need to instrument and trace
530 <<python-application,Python applications>>**, install the
531 LTTng-UST Python agent:
536 sudo apt-get install python3-lttngust
544 To install LTTng{nbsp}{revision} on Debian "stretch" (testing):
546 . Install the main LTTng{nbsp}{revision} packages:
551 sudo apt-get install lttng-modules-dkms
552 sudo apt-get install liblttng-ust-dev
553 sudo apt-get install lttng-tools
557 . **If you need to instrument and trace <<java-application,Java
558 applications>>**, install the LTTng-UST Java agent:
563 sudo apt-get install liblttng-ust-agent-java
567 . **If you need to instrument and trace <<python-application,Python
568 applications>>**, install the LTTng-UST Python agent:
573 sudo apt-get install python3-lttngust
581 To install LTTng{nbsp}{revision} (tracing control and user space
582 tracing) on Alpine Linux "edge":
584 . Make sure your system is
585 https://wiki.alpinelinux.org/wiki/Edge[configured for "edge"].
586 . Enable the _testing_ repository by uncommenting the corresponding
587 line in path:{/etc/apk/repositories}.
588 . Add the LTTng packages:
593 sudo apk add lttng-tools
594 sudo apk add lttng-ust-dev
598 To install LTTng-modules{nbsp}{revision} (Linux kernel tracing support)
599 on Alpine Linux "edge":
601 . Add the vanilla Linux kernel:
606 apk add linux-vanilla linux-vanilla-dev
610 . Reboot with the vanilla Linux kernel.
611 . Download, build, and install the latest LTTng-modules{nbsp}{revision}:
617 wget http://lttng.org/files/lttng-modules/lttng-modules-latest-2.8.tar.bz2 &&
618 tar -xf lttng-modules-latest-2.8.tar.bz2 &&
619 cd lttng-modules-2.8.* &&
621 sudo make modules_install &&
628 === OpenEmbedded and Yocto
630 LTTng{nbsp}{revision} recipes are available in the
631 http://layers.openembedded.org/layerindex/branch/master/layer/openembedded-core/[`openembedded-core`]
632 layer for Yocto Project{nbsp}2.2 _Morty_ under the following names:
638 With BitBake, the simplest way to include LTTng recipes in your target
639 image is to add them to `IMAGE_INSTALL_append` in path:{conf/local.conf}:
642 IMAGE_INSTALL_append = " lttng-tools lttng-modules lttng-ust"
647 . Select a machine and an image recipe.
648 . Click **Edit image recipe**.
649 . Under the **All recipes** tab, search for **lttng**.
650 . Check the desired LTTng recipes.
653 .Java and Python application instrumentation and tracing
655 If you need to instrument and trace <<java-application,Java
656 applications>> on openSUSE, you need to build and install
657 LTTng-UST{nbsp}{revision} <<building-from-source,from source>> and pass
658 the `--enable-java-agent-jul`, `--enable-java-agent-log4j`, or
659 `--enable-java-agent-all` options to the `configure` script, depending
660 on which Java logging framework you use.
662 If you need to instrument and trace <<python-application,Python
663 applications>> on openSUSE, you need to build and install
664 LTTng-UST{nbsp}{revision} from source and pass the
665 `--enable-python-agent` option to the `configure` script.
669 [[enterprise-distributions]]
670 === RHEL, SUSE, and other enterprise distributions
672 To install LTTng on enterprise Linux distributions, such as Red Hat
673 Enterprise Linux (RHEL) and SUSE Linux Enterprise Server (SUSE), please
674 see http://packages.efficios.com/[EfficiOS Enterprise Packages].
677 [[building-from-source]]
678 === Build from source
680 To build and install LTTng{nbsp}{revision} from source:
682 . Using your distribution's package manager, or from source, install
683 the following dependencies of LTTng-tools and LTTng-UST:
686 * https://sourceforge.net/projects/libuuid/[libuuid]
687 * http://directory.fsf.org/wiki/Popt[popt]
688 * http://liburcu.org/[Userspace RCU]
689 * http://www.xmlsoft.org/[libxml2]
692 . Download, build, and install the latest LTTng-modules{nbsp}{revision}:
698 wget http://lttng.org/files/lttng-modules/lttng-modules-latest-2.8.tar.bz2 &&
699 tar -xf lttng-modules-latest-2.8.tar.bz2 &&
700 cd lttng-modules-2.8.* &&
702 sudo make modules_install &&
707 . Download, build, and install the latest LTTng-UST{nbsp}{revision}:
713 wget http://lttng.org/files/lttng-ust/lttng-ust-latest-2.8.tar.bz2 &&
714 tar -xf lttng-ust-latest-2.8.tar.bz2 &&
715 cd lttng-ust-2.8.* &&
725 .Java and Python application tracing
727 If you need to instrument and trace <<java-application,Java
728 applications>>, pass the `--enable-java-agent-jul`,
729 `--enable-java-agent-log4j`, or `--enable-java-agent-all` options to the
730 `configure` script, depending on which Java logging framework you use.
732 If you need to instrument and trace <<python-application,Python
733 applications>>, pass the `--enable-python-agent` option to the
734 `configure` script. You can set the `PYTHON` environment variable to the
735 path to the Python interpreter for which to install the LTTng-UST Python
743 By default, LTTng-UST libraries are installed to
744 dir:{/usr/local/lib}, which is the de facto directory in which to
745 keep self-compiled and third-party libraries.
747 When <<building-tracepoint-providers-and-user-application,linking an
748 instrumented user application with `liblttng-ust`>>:
750 * Append `/usr/local/lib` to the env:LD_LIBRARY_PATH environment
752 * Pass the `-L/usr/local/lib` and `-Wl,-rpath,/usr/local/lib` options to
753 man:gcc(1), man:g++(1), or man:clang(1).
757 . Download, build, and install the latest LTTng-tools{nbsp}{revision}:
763 wget http://lttng.org/files/lttng-tools/lttng-tools-latest-2.8.tar.bz2 &&
764 tar -xf lttng-tools-latest-2.8.tar.bz2 &&
765 cd lttng-tools-2.8.* &&
773 TIP: The https://github.com/eepp/vlttng[vlttng tool] can do all the
774 previous steps automatically for a given version of LTTng and confine
775 the installed files in a specific directory. This can be useful to test
776 LTTng without installing it on your system.
782 This is a short guide to get started quickly with LTTng kernel and user
785 Before you follow this guide, make sure to <<installing-lttng,install>>
788 This tutorial walks you through the steps to:
790 . <<tracing-the-linux-kernel,Trace the Linux kernel>>.
791 . <<tracing-your-own-user-application,Trace a user application>> written
793 . <<viewing-and-analyzing-your-traces,View and analyze the
797 [[tracing-the-linux-kernel]]
798 === Trace the Linux kernel
800 The following command lines start with cmd:sudo because you need root
801 privileges to trace the Linux kernel. You can avoid using cmd:sudo if
802 your Unix user is a member of the <<lttng-sessiond,tracing group>>.
804 . Create a <<tracing-session,tracing session>>:
809 sudo lttng create my-kernel-session
813 . List the available kernel tracepoints and system calls:
822 . Create an <<event,event rule>> which matches the desired event names,
823 for example `sched_switch` and `sched_process_fork`:
828 sudo lttng enable-event --kernel sched_switch,sched_process_fork
832 You can also create an event rule which _matches_ all the Linux kernel
833 tracepoints (this will generate a lot of data when tracing):
838 sudo lttng enable-event --kernel --all
851 . Do some operation on your system for a few seconds. For example,
852 load a website, or list the files of a directory.
853 . Stop tracing and destroy the tracing session:
863 The man:lttng-destroy(1) command does not destroy the trace data; it
864 only destroys the state of the tracing session.
866 By default, LTTng saves the traces in
867 +$LTTNG_HOME/lttng-traces/__name__-__date__-__time__+,
868 where +__name__+ is the tracing session name. Note that the
869 env:LTTNG_HOME environment variable defaults to `$HOME` if not set.
871 See <<viewing-and-analyzing-your-traces,View and analyze the
872 recorded events>> to view the recorded events.
875 [[tracing-your-own-user-application]]
876 === Trace a user application
878 This section steps you through a simple example to trace a
879 _Hello world_ program written in C.
881 To create the traceable user application:
883 . Create the tracepoint provider header file, which defines the
884 tracepoints and the events they can generate:
890 #undef TRACEPOINT_PROVIDER
891 #define TRACEPOINT_PROVIDER hello_world
893 #undef TRACEPOINT_INCLUDE
894 #define TRACEPOINT_INCLUDE "./hello-tp.h"
896 #if !defined(_HELLO_TP_H) || defined(TRACEPOINT_HEADER_MULTI_READ)
899 #include <lttng/tracepoint.h>
909 ctf_string(my_string_field, my_string_arg)
910 ctf_integer(int, my_integer_field, my_integer_arg)
914 #endif /* _HELLO_TP_H */
916 #include <lttng/tracepoint-event.h>
920 . Create the tracepoint provider package source file:
926 #define TRACEPOINT_CREATE_PROBES
927 #define TRACEPOINT_DEFINE
929 #include "hello-tp.h"
933 . Build the tracepoint provider package:
938 gcc -c -I. hello-tp.c
942 . Create the _Hello World_ application source file:
949 #include "hello-tp.h"
951 int main(int argc, char *argv[])
955 puts("Hello, World!\nPress Enter to continue...");
958 * The following getchar() call is only placed here for the purpose
959 * of this demonstration, to pause the application in order for
960 * you to have time to list its tracepoints. It is not
966 * A tracepoint() call.
968 * Arguments, as defined in hello-tp.h:
970 * 1. Tracepoint provider name (required)
971 * 2. Tracepoint name (required)
972 * 3. my_integer_arg (first user-defined argument)
973 * 4. my_string_arg (second user-defined argument)
975 * Notice the tracepoint provider and tracepoint names are
976 * NOT strings: they are in fact parts of variables that the
977 * macros in hello-tp.h create.
979 tracepoint(hello_world, my_first_tracepoint, 23, "hi there!");
981 for (x = 0; x < argc; ++x) {
982 tracepoint(hello_world, my_first_tracepoint, x, argv[x]);
985 puts("Quitting now!");
986 tracepoint(hello_world, my_first_tracepoint, x * x, "x^2");
993 . Build the application:
1002 . Link the application with the tracepoint provider package,
1003 `liblttng-ust`, and `libdl`:
1008 gcc -o hello hello.o hello-tp.o -llttng-ust -ldl
1012 Here's the whole build process:
1015 .User space tracing tutorial's build steps.
1016 image::ust-flow.png[]
1018 To trace the user application:
1020 . Run the application with a few arguments:
1025 ./hello world and beyond
1034 Press Enter to continue...
1038 . Start an LTTng <<lttng-sessiond,session daemon>>:
1043 lttng-sessiond --daemonize
1047 Note that a session daemon might already be running, for example as
1048 a service that the distribution's service manager started.
1050 . List the available user space tracepoints:
1055 lttng list --userspace
1059 You see the `hello_world:my_first_tracepoint` tracepoint listed
1060 under the `./hello` process.
1062 . Create a <<tracing-session,tracing session>>:
1067 lttng create my-user-space-session
1071 . Create an <<event,event rule>> which matches the
1072 `hello_world:my_first_tracepoint` event name:
1077 lttng enable-event --userspace hello_world:my_first_tracepoint
1090 . Go back to the running `hello` application and press Enter. The
1091 program executes all `tracepoint()` instrumentation points and exits.
1092 . Stop tracing and destroy the tracing session:
1102 The man:lttng-destroy(1) command does not destroy the trace data; it
1103 only destroys the state of the tracing session.
1105 By default, LTTng saves the traces in
1106 +$LTTNG_HOME/lttng-traces/__name__-__date__-__time__+,
1107 where +__name__+ is the tracing session name. Note that the
1108 env:LTTNG_HOME environment variable defaults to `$HOME` if not set.
1110 See <<viewing-and-analyzing-your-traces,View and analyze the
1111 recorded events>> to view the recorded events.
1114 [[viewing-and-analyzing-your-traces]]
1115 === View and analyze the recorded events
1117 Once you have completed the <<tracing-the-linux-kernel,Trace the Linux
1118 kernel>> and <<tracing-your-own-user-application,Trace a user
1119 application>> tutorials, you can inspect the recorded events.
1121 Many tools are available to read LTTng traces:
1123 * **cmd:babeltrace** is a command-line utility which converts trace
1124 formats; it supports the format that LTTng produces, CTF, as well as a
1125 basic text output which can be ++grep++ed. The cmd:babeltrace command
1126 is part of the http://diamon.org/babeltrace[Babeltrace] project.
1127 * Babeltrace also includes
1128 **https://www.python.org/[Python] bindings** so
1129 that you can easily open and read an LTTng trace with your own script,
1130 benefiting from the power of Python.
1131 * http://tracecompass.org/[**Trace Compass**]
1132 is a graphical user interface for viewing and analyzing any type of
1133 logs or traces, including LTTng's.
1134 * https://github.com/lttng/lttng-analyses[**LTTng analyses**] is a
1135 project which includes many high-level analyses of LTTng kernel
1136 traces, like scheduling statistics, interrupt frequency distribution,
1137 top CPU usage, and more.
1139 NOTE: This section assumes that the traces recorded during the previous
1140 tutorials were saved to their default location, in the
1141 dir:{$LTTNG_HOME/lttng-traces} directory. Note that the env:LTTNG_HOME
1142 environment variable defaults to `$HOME` if not set.
1145 [[viewing-and-analyzing-your-traces-bt]]
1146 ==== Use the cmd:babeltrace command-line tool
1148 The simplest way to list all the recorded events of a trace is to pass
1149 its path to cmd:babeltrace with no options:
1153 babeltrace ~/lttng-traces/my-user-space-session*
1156 cmd:babeltrace finds all traces recursively within the given path and
1157 prints all their events, merging them in chronological order.
1159 You can pipe the output of cmd:babeltrace into a tool like man:grep(1) for
1164 babeltrace ~/lttng-traces/my-kernel-session* | grep sys_
1167 You can pipe the output of cmd:babeltrace into a tool like man:wc(1) to
1168 count the recorded events:
1172 babeltrace ~/lttng-traces/my-kernel-session* | grep sys_read | wc --lines
1176 [[viewing-and-analyzing-your-traces-bt-python]]
1177 ==== Use the Babeltrace Python bindings
1179 The <<viewing-and-analyzing-your-traces-bt,text output of cmd:babeltrace>>
1180 is useful to isolate events by simple matching using man:grep(1) and
1181 similar utilities. However, more elaborate filters, such as keeping only
1182 event records with a field value falling within a specific range, are
1183 not trivial to write using a shell. Moreover, reductions and even the
1184 most basic computations involving multiple event records are virtually
1185 impossible to implement.
1187 Fortunately, Babeltrace ships with Python 3 bindings which makes it easy
1188 to read the event records of an LTTng trace sequentially and compute the
1189 desired information.
1191 The following script accepts an LTTng Linux kernel trace path as its
1192 first argument and prints the short names of the top 5 running processes
1193 on CPU 0 during the whole trace:
1198 from collections import Counter
1204 if len(sys.argv) != 2:
1205 msg = 'Usage: python3 {} TRACEPATH'.format(sys.argv[0])
1206 print(msg, file=sys.stderr)
1209 # A trace collection contains one or more traces
1210 col = babeltrace.TraceCollection()
1212 # Add the trace provided by the user (LTTng traces always have
1214 if col.add_trace(sys.argv[1], 'ctf') is None:
1215 raise RuntimeError('Cannot add trace')
1217 # This counter dict contains execution times:
1219 # task command name -> total execution time (ns)
1220 exec_times = Counter()
1222 # This contains the last `sched_switch` timestamp
1226 for event in col.events:
1227 # Keep only `sched_switch` events
1228 if event.name != 'sched_switch':
1231 # Keep only events which happened on CPU 0
1232 if event['cpu_id'] != 0:
1236 cur_ts = event.timestamp
1242 # Previous task command (short) name
1243 prev_comm = event['prev_comm']
1245 # Initialize entry in our dict if not yet done
1246 if prev_comm not in exec_times:
1247 exec_times[prev_comm] = 0
1249 # Compute previous command execution time
1250 diff = cur_ts - last_ts
1252 # Update execution time of this command
1253 exec_times[prev_comm] += diff
1255 # Update last timestamp
1259 for name, ns in exec_times.most_common(5):
1261 print('{:20}{} s'.format(name, s))
1266 if __name__ == '__main__':
1267 sys.exit(0 if top5proc() else 1)
1274 python3 top5proc.py ~/lttng-traces/my-kernel-session*/kernel
1280 swapper/0 48.607245889 s
1281 chromium 7.192738188 s
1282 pavucontrol 0.709894415 s
1283 Compositor 0.660867933 s
1284 Xorg.bin 0.616753786 s
1287 Note that `swapper/0` is the "idle" process of CPU 0 on Linux; since we
1288 weren't using the CPU that much when tracing, its first position in the
1293 == [[understanding-lttng]]Core concepts
1295 From a user's perspective, the LTTng system is built on a few concepts,
1296 or objects, on which the <<lttng-cli,cmd:lttng command-line tool>>
1297 operates by sending commands to the <<lttng-sessiond,session daemon>>.
1298 Understanding how those objects relate to eachother is key in mastering
1301 The core concepts are:
1303 * <<tracing-session,Tracing session>>
1304 * <<domain,Tracing domain>>
1305 * <<channel,Channel and ring buffer>>
1306 * <<"event","Instrumentation point, event rule, event, and event record">>
1312 A _tracing session_ is a stateful dialogue between you and
1313 a <<lttng-sessiond,session daemon>>. You can
1314 <<creating-destroying-tracing-sessions,create a new tracing
1315 session>> with the `lttng create` command.
1317 Anything that you do when you control LTTng tracers happens within a
1318 tracing session. In particular, a tracing session:
1321 * Has its own set of trace files.
1322 * Has its own state of activity (started or stopped).
1323 * Has its own <<tracing-session-mode,mode>> (local, network streaming,
1325 * Has its own <<channel,channels>> which have their own
1326 <<event,event rules>>.
1329 .A _tracing session_ contains <<channel,channels>> that are members of <<domain,tracing domains>> and contain <<event,event rules>>.
1330 image::concepts.png[]
1332 Those attributes and objects are completely isolated between different
1335 A tracing session is analogous to a cash machine session:
1336 the operations you do on the banking system through the cash machine do
1337 not alter the data of other users of the same system. In the case of
1338 the cash machine, a session lasts as long as your bank card is inside.
1339 In the case of LTTng, a tracing session lasts from the `lttng create`
1340 command to the `lttng destroy` command.
1343 .Each Unix user has its own set of tracing sessions.
1344 image::many-sessions.png[]
1347 [[tracing-session-mode]]
1348 ==== Tracing session mode
1350 LTTng can send the generated trace data to different locations. The
1351 _tracing session mode_ dictates where to send it. The following modes
1352 are available in LTTng{nbsp}{revision}:
1355 LTTng writes the traces to the file system of the machine being traced
1358 Network streaming mode::
1359 LTTng sends the traces over the network to a
1360 <<lttng-relayd,relay daemon>> running on a remote system.
1363 LTTng does not write the traces by default. Instead, you can request
1364 LTTng to <<taking-a-snapshot,take a snapshot>>, that is, a copy of the
1365 current tracing buffers, and to write it to the target's file system
1366 or to send it over the network to a <<lttng-relayd,relay daemon>>
1367 running on a remote system.
1370 This mode is similar to the network streaming mode, but a live
1371 trace viewer can connect to the distant relay daemon to
1372 <<lttng-live,view event records as LTTng generates them>> by
1379 A _tracing domain_ is a namespace for event sources. A tracing domain
1380 has its own properties and features.
1382 There are currently five available tracing domains:
1386 * `java.util.logging` (JUL)
1390 You must specify a tracing domain when using some commands to avoid
1391 ambiguity. For example, since all the domains support named tracepoints
1392 as event sources (instrumentation points that you manually insert in the
1393 source code), you need to specify a tracing domain when
1394 <<enabling-disabling-events,creating an event rule>> because all the
1395 tracing domains could have tracepoints with the same names.
1397 Some features are reserved to specific tracing domains. Dynamic function
1398 entry and return instrumentation points, for example, are currently only
1399 supported in the Linux kernel tracing domain, but support for other
1400 tracing domains could be added in the future.
1402 You can create <<channel,channels>> in the Linux kernel and user space
1403 tracing domains. The other tracing domains have a single default
1408 === Channel and ring buffer
1410 A _channel_ is an object which is responsible for a set of ring buffers.
1411 Each ring buffer is divided into multiple sub-buffers. When an LTTng
1412 tracer emits an event, it can record it to one or more
1413 sub-buffers. The attributes of a channel determine what to do when
1414 there's no space left for a new event record because all sub-buffers
1415 are full, where to send a full sub-buffer, and other behaviours.
1417 A channel is always associated to a <<domain,tracing domain>>. The
1418 `java.util.logging` (JUL), log4j, and Python tracing domains each have
1419 a default channel which you cannot configure.
1421 A channel also owns <<event,event rules>>. When an LTTng tracer emits
1422 an event, it records it to the sub-buffers of all
1423 the enabled channels with a satisfied event rule, as long as those
1424 channels are part of active <<tracing-session,tracing sessions>>.
1427 [[channel-buffering-schemes]]
1428 ==== Per-user vs. per-process buffering schemes
1430 A channel has at least one ring buffer _per CPU_. LTTng always
1431 records an event to the ring buffer associated to the CPU on which it
1434 Two _buffering schemes_ are available when you
1435 <<enabling-disabling-channels,create a channel>> in the
1436 user space <<domain,tracing domain>>:
1438 Per-user buffering::
1439 Allocate one set of ring buffers--one per CPU--shared by all the
1440 instrumented processes of each Unix user.
1444 .Per-user buffering scheme.
1445 image::per-user-buffering.png[]
1448 Per-process buffering::
1449 Allocate one set of ring buffers--one per CPU--for each
1450 instrumented process.
1454 .Per-process buffering scheme.
1455 image::per-process-buffering.png[]
1458 The per-process buffering scheme tends to consume more memory than the
1459 per-user option because systems generally have more instrumented
1460 processes than Unix users running instrumented processes. However, the
1461 per-process buffering scheme ensures that one process having a high
1462 event throughput won't fill all the shared sub-buffers of the same
1465 The Linux kernel tracing domain has only one available buffering scheme
1466 which is to allocate a single set of ring buffers for the whole system.
1467 This scheme is similar to the per-user option, but with a single, global
1468 user "running" the kernel.
1471 [[channel-overwrite-mode-vs-discard-mode]]
1472 ==== Overwrite vs. discard event loss modes
1474 When an event occurs, LTTng records it to a specific sub-buffer (yellow
1475 arc in the following animation) of a specific channel's ring buffer.
1476 When there's no space left in a sub-buffer, the tracer marks it as
1477 consumable (red) and another, empty sub-buffer starts receiving the
1478 following event records. A <<lttng-consumerd,consumer daemon>>
1479 eventually consumes the marked sub-buffer (returns to white).
1482 [role="docsvg-channel-subbuf-anim"]
1487 In an ideal world, sub-buffers are consumed faster than they are filled,
1488 as is the case in the previous animation. In the real world,
1489 however, all sub-buffers can be full at some point, leaving no space to
1490 record the following events.
1492 By design, LTTng is a _non-blocking_ tracer: when no empty sub-buffer is
1493 available, it is acceptable to lose event records when the alternative
1494 would be to cause substantial delays in the instrumented application's
1495 execution. LTTng privileges performance over integrity; it aims at
1496 perturbing the traced system as little as possible in order to make
1497 tracing of subtle race conditions and rare interrupt cascades possible.
1499 When it comes to losing event records because no empty sub-buffer is
1500 available, the channel's _event loss mode_ determines what to do. The
1501 available event loss modes are:
1504 Drop the newest event records until a the tracer
1505 releases a sub-buffer.
1508 Clear the sub-buffer containing the oldest event records and start
1509 writing the newest event records there.
1511 This mode is sometimes called _flight recorder mode_ because it's
1513 https://en.wikipedia.org/wiki/Flight_recorder[flight recorder]:
1514 always keep a fixed amount of the latest data.
1516 Which mechanism you should choose depends on your context: prioritize
1517 the newest or the oldest event records in the ring buffer?
1519 Beware that, in overwrite mode, the tracer abandons a whole sub-buffer
1520 as soon as a there's no space left for a new event record, whereas in
1521 discard mode, the tracer only discards the event record that doesn't
1524 In discard mode, LTTng increments a count of lost event records when
1525 an event record is lost and saves this count to the trace. In
1526 overwrite mode, LTTng keeps no information when it overwrites a
1527 sub-buffer before consuming it.
1529 There are a few ways to decrease your probability of losing event
1531 <<channel-subbuf-size-vs-subbuf-count,Sub-buffer count and size>> shows
1532 how you can fine-une the sub-buffer count and size of a channel to
1533 virtually stop losing event records, though at the cost of greater
1537 [[channel-subbuf-size-vs-subbuf-count]]
1538 ==== Sub-buffer count and size
1540 When you <<enabling-disabling-channels,create a channel>>, you can
1541 set its number of sub-buffers and their size.
1543 Note that there is noticeable CPU overhead introduced when
1544 switching sub-buffers (marking a full one as consumable and switching
1545 to an empty one for the following events to be recorded). Knowing this,
1546 the following list presents a few practical situations along with how
1547 to configure the sub-buffer count and size for them:
1549 * **High event throughput**: In general, prefer bigger sub-buffers to
1550 lower the risk of losing event records.
1552 Having bigger sub-buffers also ensures a lower sub-buffer switching
1555 The number of sub-buffers is only meaningful if you create the channel
1556 in overwrite mode: in this case, if a sub-buffer overwrite happens, the
1557 other sub-buffers are left unaltered.
1559 * **Low event throughput**: In general, prefer smaller sub-buffers
1560 since the risk of losing event records is low.
1562 Because events occur less frequently, the sub-buffer switching frequency
1563 should remain low and thus the tracer's overhead should not be a
1566 * **Low memory system**: If your target system has a low memory
1567 limit, prefer fewer first, then smaller sub-buffers.
1569 Even if the system is limited in memory, you want to keep the
1570 sub-buffers as big as possible to avoid a high sub-buffer switching
1573 Note that LTTng uses http://diamon.org/ctf/[CTF] as its trace format,
1574 which means event data is very compact. For example, the average
1575 LTTng kernel event record weights about 32{nbsp}bytes. Thus, a
1576 sub-buffer size of 1{nbsp}MiB is considered big.
1578 The previous situations highlight the major trade-off between a few big
1579 sub-buffers and more, smaller sub-buffers: sub-buffer switching
1580 frequency vs. how much data is lost in overwrite mode. Assuming a
1581 constant event throughput and using the overwrite mode, the two
1582 following configurations have the same ring buffer total size:
1585 [role="docsvg-channel-subbuf-size-vs-count-anim"]
1590 * **2 sub-buffers of 4{nbsp}MiB each**: Expect a very low sub-buffer
1591 switching frequency, but if a sub-buffer overwrite happens, half of
1592 the event records so far (4{nbsp}MiB) are definitely lost.
1593 * **8 sub-buffers of 1{nbsp}MiB each**: Expect 4{nbsp}times the tracer's
1594 overhead as the previous configuration, but if a sub-buffer
1595 overwrite happens, only the eighth of event records so far are
1598 In discard mode, the sub-buffers count parameter is pointless: use two
1599 sub-buffers and set their size according to the requirements of your
1603 [[channel-switch-timer]]
1604 ==== Switch timer period
1606 The _switch timer period_ is an important configurable attribute of
1607 a channel to ensure periodic sub-buffer flushing.
1609 When the _switch timer_ expires, a sub-buffer switch happens. You can
1610 set the switch timer period attribute when you
1611 <<enabling-disabling-channels,create a channel>> to ensure that event
1612 data is consumed and committed to trace files or to a distant relay
1613 daemon periodically in case of a low event throughput.
1616 [role="docsvg-channel-switch-timer"]
1621 This attribute is also convenient when you use big sub-buffers to cope
1622 with a sporadic high event throughput, even if the throughput is
1626 [[channel-read-timer]]
1627 ==== Read timer period
1629 By default, the LTTng tracers use a notification mechanism to signal a
1630 full sub-buffer so that a consumer daemon can consume it. When such
1631 notifications must be avoided, for example in real-time applications,
1632 you can use the channel's _read timer_ instead. When the read timer
1633 fires, the <<lttng-consumerd,consumer daemon>> checks for full,
1634 consumable sub-buffers.
1637 [[tracefile-rotation]]
1638 ==== Trace file count and size
1640 By default, trace files can grow as large as needed. You can set the
1641 maximum size of each trace file that a channel writes when you
1642 <<enabling-disabling-channels,create a channel>>. When the size of
1643 a trace file reaches the channel's fixed maximum size, LTTng creates
1644 another file to contain the next event records. LTTng appends a file
1645 count to each trace file name in this case.
1647 If you set the trace file size attribute when you create a channel, the
1648 maximum number of trace files that LTTng creates is _unlimited_ by
1649 default. To limit them, you can also set a maximum number of trace
1650 files. When the number of trace files reaches the channel's fixed
1651 maximum count, the oldest trace file is overwritten. This mechanism is
1652 called _trace file rotation_.
1656 === Instrumentation point, event rule, event, and event record
1658 An _event rule_ is a set of conditions which must be **all** satisfied
1659 for LTTng to record an occuring event.
1661 You set the conditions when you <<enabling-disabling-events,create
1664 You always attach an event rule to <<channel,channel>> when you create
1667 When an event passes the conditions of an event rule, LTTng records it
1668 in one of the attached channel's sub-buffers.
1670 The available conditions, as of LTTng{nbsp}{revision}, are:
1672 * The event rule _is enabled_.
1673 * The instrumentation point's type _is{nbsp}T_.
1674 * The instrumentation point's name (sometimes called _event name_)
1675 _matches{nbsp}N_, but _is not{nbsp}E_.
1676 * The instrumentation point's log level _is as severe as{nbsp}L_, or
1677 _is exactly{nbsp}L_.
1678 * The fields of the event's payload _satisfy_ a filter
1679 expression{nbsp}__F__.
1681 As you can see, all the conditions but the dynamic filter are related to
1682 the event rule's status or to the instrumentation point, not to the
1683 occurring events. This is why, without a filter, checking if an event
1684 passes an event rule is not a dynamic task: when you create or modify an
1685 event rule, all the tracers of its tracing domain enable or disable the
1686 instrumentation points themselves once. This is possible because the
1687 attributes of an instrumentation point (type, name, and log level) are
1688 defined statically. In other words, without a dynamic filter, the tracer
1689 _does not evaluate_ the arguments of an instrumentation point unless it
1690 matches an enabled event rule.
1692 Note that, for LTTng to record an event, the <<channel,channel>> to
1693 which a matching event rule is attached must also be enabled, and the
1694 tracing session owning this channel must be active.
1697 .Logical path from an instrumentation point to an event record.
1698 image::event-rule.png[]
1700 .Event, event record, or event rule?
1702 With so many similar terms, it's easy to get confused.
1704 An **event** is the consequence of the execution of an _instrumentation
1705 point_, like a tracepoint that you manually place in some source code,
1706 or a Linux kernel KProbe. An event is said to _occur_ at a specific
1707 time. Different actions can be taken upon the occurance of an event,
1708 like record the event's payload to a buffer.
1710 An **event record** is the representation of an event in a sub-buffer. A
1711 tracer is responsible for capturing the payload of an event, current
1712 context variables, the event's ID, and the event's timestamp. LTTng
1713 can append this sub-buffer to a trace file.
1715 An **event rule** is a set of conditions which must all be satisfied for
1716 LTTng to record an occuring event. Events still occur without
1717 satisfying event rules, but LTTng does not record them.
1722 == Components of noch:{LTTng}
1724 The second _T_ in _LTTng_ stands for _toolkit_: it would be wrong
1725 to call LTTng a simple _tool_ since it is composed of multiple
1726 interacting components. This section describes those components,
1727 explains their respective roles, and shows how they connect together to
1728 form the LTTng ecosystem.
1730 The following diagram shows how the most important components of LTTng
1731 interact with user applications, the Linux kernel, and you:
1734 .Control and trace data paths between LTTng components.
1735 image::plumbing.png[]
1737 The LTTng project incorporates:
1739 * **LTTng-tools**: Libraries and command-line interface to
1740 control tracing sessions.
1741 ** <<lttng-sessiond,Session daemon>> (man:lttng-sessiond(8)).
1742 ** <<lttng-consumerd,Consumer daemon>> (man:lttng-consumerd(8)).
1743 ** <<lttng-relayd,Relay daemon>> (man:lttng-relayd(8)).
1744 ** <<liblttng-ctl-lttng,Tracing control library>> (`liblttng-ctl`).
1745 ** <<lttng-cli,Tracing control command-line tool>> (man:lttng(1)).
1746 * **LTTng-UST**: Libraries and Java/Python packages to trace user
1748 ** <<lttng-ust,User space tracing library>> (`liblttng-ust`) and its
1749 headers to instrument and trace any native user application.
1750 ** <<prebuilt-ust-helpers,Preloadable user space tracing helpers>>:
1751 *** `liblttng-ust-libc-wrapper`
1752 *** `liblttng-ust-pthread-wrapper`
1753 *** `liblttng-ust-cyg-profile`
1754 *** `liblttng-ust-cyg-profile-fast`
1755 *** `liblttng-ust-dl`
1756 ** User space tracepoint provider source files generator command-line
1757 tool (man:lttng-gen-tp(1)).
1758 ** <<lttng-ust-agents,LTTng-UST Java agent>> to instrument and trace
1759 Java applications using `java.util.logging` or
1760 Apache log4j 1.2 logging.
1761 ** <<lttng-ust-agents,LTTng-UST Python agent>> to instrument
1762 Python applications using the standard `logging` package.
1763 * **LTTng-modules**: <<lttng-modules,Linux kernel modules>> to trace
1765 ** LTTng kernel tracer module.
1766 ** Tracing ring buffer kernel modules.
1767 ** Probe kernel modules.
1768 ** LTTng logger kernel module.
1772 === Tracing control command-line interface
1775 .The tracing control command-line interface.
1776 image::plumbing-lttng-cli.png[]
1778 The _man:lttng(1) command-line tool_ is the standard user interface to
1779 control LTTng <<tracing-session,tracing sessions>>. The cmd:lttng tool
1780 is part of LTTng-tools.
1782 The cmd:lttng tool is linked with
1783 <<liblttng-ctl-lttng,`liblttng-ctl`>> to communicate with
1784 one or more <<lttng-sessiond,session daemons>> behind the scenes.
1786 The cmd:lttng tool has a Git-like interface:
1790 lttng <general options> <command> <command options>
1793 The <<controlling-tracing,Tracing control>> section explores the
1794 available features of LTTng using the cmd:lttng tool.
1797 [[liblttng-ctl-lttng]]
1798 === Tracing control library
1801 .The tracing control library.
1802 image::plumbing-liblttng-ctl.png[]
1804 The _LTTng control library_, `liblttng-ctl`, is used to communicate
1805 with a <<lttng-sessiond,session daemon>> using a C API that hides the
1806 underlying protocol's details. `liblttng-ctl` is part of LTTng-tools.
1808 The <<lttng-cli,cmd:lttng command-line tool>>
1809 is linked with `liblttng-ctl`.
1811 You can use `liblttng-ctl` in C or $$C++$$ source code by including its
1816 #include <lttng/lttng.h>
1819 Some objects are referenced by name (C string), such as tracing
1820 sessions, but most of them require to create a handle first using
1821 `lttng_create_handle()`.
1823 The best available developer documentation for `liblttng-ctl` is, as of
1824 LTTng{nbsp}{revision}, its installed header files. Every function and
1825 structure is thoroughly documented.
1829 === User space tracing library
1832 .The user space tracing library.
1833 image::plumbing-liblttng-ust.png[]
1835 The _user space tracing library_, `liblttng-ust` (see man:lttng-ust(3)),
1836 is the LTTng user space tracer. It receives commands from a
1837 <<lttng-sessiond,session daemon>>, for example to
1838 enable and disable specific instrumentation points, and writes event
1839 records to ring buffers shared with a
1840 <<lttng-consumerd,consumer daemon>>.
1841 `liblttng-ust` is part of LTTng-UST.
1843 Public C header files are installed beside `liblttng-ust` to
1844 instrument any <<c-application,C or $$C++$$ application>>.
1846 <<lttng-ust-agents,LTTng-UST agents>>, which are regular Java and Python
1847 packages, use their own library providing tracepoints which is
1848 linked with `liblttng-ust`.
1850 An application or library does not have to initialize `liblttng-ust`
1851 manually: its constructor does the necessary tasks to properly register
1852 to a session daemon. The initialization phase also enables the
1853 instrumentation points matching the <<event,event rules>> that you
1857 [[lttng-ust-agents]]
1858 === User space tracing agents
1861 .The user space tracing agents.
1862 image::plumbing-lttng-ust-agents.png[]
1864 The _LTTng-UST Java and Python agents_ are regular Java and Python
1865 packages which add LTTng tracing capabilities to the
1866 native logging frameworks. The LTTng-UST agents are part of LTTng-UST.
1868 In the case of Java, the
1869 https://docs.oracle.com/javase/7/docs/api/java/util/logging/package-summary.html[`java.util.logging`
1870 core logging facilities] and
1871 https://logging.apache.org/log4j/1.2/[Apache log4j 1.2] are supported.
1872 Note that Apache Log4{nbsp}2 is not supported.
1874 In the case of Python, the standard
1875 https://docs.python.org/3/library/logging.html[`logging`] package
1876 is supported. Both Python 2 and Python 3 modules can import the
1877 LTTng-UST Python agent package.
1879 The applications using the LTTng-UST agents are in the
1880 `java.util.logging` (JUL),
1881 log4j, and Python <<domain,tracing domains>>.
1883 Both agents use the same mechanism to trace the log statements. When an
1884 agent is initialized, it creates a log handler that attaches to the root
1885 logger. The agent also registers to a <<lttng-sessiond,session daemon>>.
1886 When the application executes a log statement, it is passed to the
1887 agent's log handler by the root logger. The agent's log handler calls a
1888 native function in a tracepoint provider package shared library linked
1889 with <<lttng-ust,`liblttng-ust`>>, passing the formatted log message and
1890 other fields, like its logger name and its log level. This native
1891 function contains a user space instrumentation point, hence tracing the
1894 The log level condition of an
1895 <<event,event rule>> is considered when tracing
1896 a Java or a Python application, and it's compatible with the standard
1897 JUL, log4j, and Python log levels.
1901 === LTTng kernel modules
1904 .The LTTng kernel modules.
1905 image::plumbing-lttng-modules.png[]
1907 The _LTTng kernel modules_ are a set of Linux kernel modules
1908 which implement the kernel tracer of the LTTng project. The LTTng
1909 kernel modules are part of LTTng-modules.
1911 The LTTng kernel modules include:
1913 * A set of _probe_ modules.
1915 Each module attaches to a specific subsystem
1916 of the Linux kernel using its tracepoint instrument points. There are
1917 also modules to attach to the entry and return points of the Linux
1918 system call functions.
1920 * _Ring buffer_ modules.
1922 A ring buffer implementation is provided as kernel modules. The LTTng
1923 kernel tracer writes to the ring buffer; a
1924 <<lttng-consumerd,consumer daemon>> reads from the ring buffer.
1926 * The _LTTng kernel tracer_ module.
1927 * The _LTTng logger_ module.
1929 The LTTng logger module implements the special path:{/proc/lttng-logger}
1930 file so that any executable can generate LTTng events by opening and
1931 writing to this file.
1933 See <<proc-lttng-logger-abi,LTTng logger>>.
1935 Generally, you do not have to load the LTTng kernel modules manually
1936 (using man:modprobe(8), for example): a root <<lttng-sessiond,session
1937 daemon>> loads the necessary modules when starting. If you have extra
1938 probe modules, you can specify to load them to the session daemon on
1941 The LTTng kernel modules are installed in
1942 +/usr/lib/modules/__release__/extra+ by default, where +__release__+ is
1943 the kernel release (see `uname --kernel-release`).
1950 .The session daemon.
1951 image::plumbing-sessiond.png[]
1953 The _session daemon_, man:lttng-sessiond(8), is a daemon responsible for
1954 managing tracing sessions and for controlling the various components of
1955 LTTng. The session daemon is part of LTTng-tools.
1957 The session daemon sends control requests to and receives control
1960 * The <<lttng-ust,user space tracing library>>.
1962 Any instance of the user space tracing library first registers to
1963 a session daemon. Then, the session daemon can send requests to
1964 this instance, such as:
1967 ** Get the list of tracepoints.
1968 ** Share an <<event,event rule>> so that the user space tracing library
1969 can enable or disable tracepoints. Amongst the possible conditions
1970 of an event rule is a filter expression which `liblttng-ust` evalutes
1971 when an event occurs.
1972 ** Share <<channel,channel>> attributes and ring buffer locations.
1975 The session daemon and the user space tracing library use a Unix
1976 domain socket for their communication.
1978 * The <<lttng-ust-agents,user space tracing agents>>.
1980 Any instance of a user space tracing agent first registers to
1981 a session daemon. Then, the session daemon can send requests to
1982 this instance, such as:
1985 ** Get the list of loggers.
1986 ** Enable or disable a specific logger.
1989 The session daemon and the user space tracing agent use a TCP connection
1990 for their communication.
1992 * The <<lttng-modules,LTTng kernel tracer>>.
1993 * The <<lttng-consumerd,consumer daemon>>.
1995 The session daemon sends requests to the consumer daemon to instruct
1996 it where to send the trace data streams, amongst other information.
1998 * The <<lttng-relayd,relay daemon>>.
2000 The session daemon receives commands from the
2001 <<liblttng-ctl-lttng,tracing control library>>.
2003 The root session daemon loads the appropriate
2004 <<lttng-modules,LTTng kernel modules>> on startup. It also spawns
2005 a <<lttng-consumerd,consumer daemon>> as soon as you create
2006 an <<event,event rule>>.
2008 The session daemon does not send and receive trace data: this is the
2009 role of the <<lttng-consumerd,consumer daemon>> and
2010 <<lttng-relayd,relay daemon>>. It does, however, generate the
2011 http://diamon.org/ctf/[CTF] metadata stream.
2013 Each Unix user can have its own session daemon instance. The
2014 tracing sessions managed by different session daemons are completely
2017 The root user's session daemon is the only one which is
2018 allowed to control the LTTng kernel tracer, and its spawned consumer
2019 daemon is the only one which is allowed to consume trace data from the
2020 LTTng kernel tracer. Note, however, that any Unix user which is a member
2021 of the <<tracing-group,tracing group>> is allowed
2022 to create <<channel,channels>> in the
2023 Linux kernel <<domain,tracing domain>>, and thus to trace the Linux
2026 The <<lttng-cli,cmd:lttng command-line tool>> automatically starts a
2027 session daemon when using its `create` command if none is currently
2028 running. You can also start the session daemon manually.
2035 .The consumer daemon.
2036 image::plumbing-consumerd.png[]
2038 The _consumer daemon_, man:lttng-consumerd(8), is a daemon which shares
2039 ring buffers with user applications or with the LTTng kernel modules to
2040 collect trace data and send it to some location (on disk or to a
2041 <<lttng-relayd,relay daemon>> over the network). The consumer daemon
2042 is part of LTTng-tools.
2044 You do not start a consumer daemon manually: a consumer daemon is always
2045 spawned by a <<lttng-sessiond,session daemon>> as soon as you create an
2046 <<event,event rule>>, that is, before you start tracing. When you kill
2047 its owner session daemon, the consumer daemon also exits because it is
2048 the session daemon's child process. Command-line options of
2049 man:lttng-sessiond(8) target the consumer daemon process.
2051 There are up to two running consumer daemons per Unix user, whereas only
2052 one session daemon can run per user. This is because each process can be
2053 either 32-bit or 64-bit: if the target system runs a mixture of 32-bit
2054 and 64-bit processes, it is more efficient to have separate
2055 corresponding 32-bit and 64-bit consumer daemons. The root user is an
2056 exception: it can have up to _three_ running consumer daemons: 32-bit
2057 and 64-bit instances for its user applications, and one more
2058 reserved for collecting kernel trace data.
2066 image::plumbing-relayd.png[]
2068 The _relay daemon_, man:lttng-relayd(8), is a daemon acting as a bridge
2069 between remote session and consumer daemons, local trace files, and a
2070 remote live trace viewer. The relay daemon is part of LTTng-tools.
2072 The main purpose of the relay daemon is to implement a receiver of
2073 <<sending-trace-data-over-the-network,trace data over the network>>.
2074 This is useful when the target system does not have much file system
2075 space to record trace files locally.
2077 The relay daemon is also a server to which a
2078 <<lttng-live,live trace viewer>> can
2079 connect. The live trace viewer sends requests to the relay daemon to
2080 receive trace data as the target system emits events. The
2081 communication protocol is named _LTTng live_; it is used over TCP
2084 Note that you can start the relay daemon on the target system directly.
2085 This is the setup of choice when the use case is to view events as
2086 the target system emits them without the need of a remote system.
2090 == [[using-lttng]]Instrumentation
2092 There are many examples of tracing and monitoring in our everyday life:
2094 * You have access to real-time and historical weather reports and
2095 forecasts thanks to weather stations installed around the country.
2096 * You know your heart is safe thanks to an electrocardiogram.
2097 * You make sure not to drive your car too fast and to have enough fuel
2098 to reach your destination thanks to gauges visible on your dashboard.
2100 All the previous examples have something in common: they rely on
2101 **instruments**. Without the electrodes attached to the surface of your
2102 body's skin, cardiac monitoring is futile.
2104 LTTng, as a tracer, is no different from those real life examples. If
2105 you're about to trace a software system or, in other words, record its
2106 history of execution, you better have **instrumentation points** in the
2107 subject you're tracing, that is, the actual software.
2109 Various ways were developed to instrument a piece of software for LTTng
2110 tracing. The most straightforward one is to manually place
2111 instrumentation points, called _tracepoints_, in the software's source
2112 code. It is also possible to add instrumentation points dynamically in
2113 the Linux kernel <<domain,tracing domain>>.
2115 If you're only interested in tracing the Linux kernel, your
2116 instrumentation needs are probably already covered by LTTng's built-in
2117 <<lttng-modules,Linux kernel tracepoints>>. You may also wish to trace a
2118 user application which is already instrumented for LTTng tracing.
2119 In such cases, you can skip this whole section and read the topics of
2120 the <<controlling-tracing,Tracing control>> section.
2122 Many methods are available to instrument a piece of software for LTTng
2125 * <<c-application,User space instrumentation for C and $$C++$$
2127 * <<prebuilt-ust-helpers,Prebuilt user space tracing helpers>>.
2128 * <<java-application,User space Java agent>>.
2129 * <<python-application,User space Python agent>>.
2130 * <<proc-lttng-logger-abi,LTTng logger>>.
2131 * <<instrumenting-linux-kernel,LTTng kernel tracepoints>>.
2135 === [[cxx-application]]User space instrumentation for C and $$C++$$ applications
2137 The procedure to instrument a C or $$C++$$ user application with
2138 the <<lttng-ust,LTTng user space tracing library>>, `liblttng-ust`, is:
2140 . <<tracepoint-provider,Create the source files of a tracepoint provider
2142 . <<probing-the-application-source-code,Add tracepoints to
2143 the application's source code>>.
2144 . <<building-tracepoint-providers-and-user-application,Build and link
2145 a tracepoint provider package and the user application>>.
2147 If you need quick, man:printf(3)-like instrumentation, you can skip
2148 those steps and use <<tracef,`tracef()`>> or <<tracelog,`tracelog()`>>
2151 IMPORTANT: You need to <<installing-lttng,install>> LTTng-UST to
2152 instrument a user application with `liblttng-ust`.
2155 [[tracepoint-provider]]
2156 ==== Create the source files of a tracepoint provider package
2158 A _tracepoint provider_ is a set of compiled functions which provide
2159 **tracepoints** to an application, the type of instrumentation point
2160 supported by LTTng-UST. Those functions can emit events with
2161 user-defined fields and serialize those events as event records to one
2162 or more LTTng-UST <<channel,channel>> sub-buffers. The `tracepoint()`
2163 macro, which you <<probing-the-application-source-code,insert in a user
2164 application's source code>>, calls those functions.
2166 A _tracepoint provider package_ is an object file (`.o`) or a shared
2167 library (`.so`) which contains one or more tracepoint providers.
2168 Its source files are:
2170 * One or more <<tpp-header,tracepoint provider header>> (`.h`).
2171 * A <<tpp-source,tracepoint provider package source>> (`.c`).
2173 A tracepoint provider package is dynamically linked with `liblttng-ust`,
2174 the LTTng user space tracer, at run time.
2177 .User application linked with `liblttng-ust` and containing a tracepoint provider.
2178 image::ust-app.png[]
2180 NOTE: If you need quick, man:printf(3)-like instrumentation, you can
2181 skip creating and using a tracepoint provider and use
2182 <<tracef,`tracef()`>> or <<tracelog,`tracelog()`>> instead.
2186 ===== Create a tracepoint provider header file template
2188 A _tracepoint provider header file_ contains the tracepoint
2189 definitions of a tracepoint provider.
2191 To create a tracepoint provider header file:
2193 . Start from this template:
2197 .Tracepoint provider header file template (`.h` file extension).
2199 #undef TRACEPOINT_PROVIDER
2200 #define TRACEPOINT_PROVIDER provider_name
2202 #undef TRACEPOINT_INCLUDE
2203 #define TRACEPOINT_INCLUDE "./tp.h"
2205 #if !defined(_TP_H) || defined(TRACEPOINT_HEADER_MULTI_READ)
2208 #include <lttng/tracepoint.h>
2211 * Use TRACEPOINT_EVENT(), TRACEPOINT_EVENT_CLASS(),
2212 * TRACEPOINT_EVENT_INSTANCE(), and TRACEPOINT_LOGLEVEL() here.
2217 #include <lttng/tracepoint-event.h>
2223 * `provider_name` with the name of your tracepoint provider.
2224 * `"tp.h"` with the name of your tracepoint provider header file.
2226 . Below the `#include <lttng/tracepoint.h>` line, put your
2227 <<defining-tracepoints,tracepoint definitions>>.
2229 Your tracepoint provider name must be unique amongst all the possible
2230 tracepoint provider names used on the same target system. We
2231 suggest to include the name of your project or company in the name,
2232 for example, `org_lttng_my_project_tpp`.
2234 TIP: [[lttng-gen-tp]]You can use the man:lttng-gen-tp(1) tool to create
2235 this boilerplate for you. When using cmd:lttng-gen-tp, all you need to
2236 write are the <<defining-tracepoints,tracepoint definitions>>.
2239 [[defining-tracepoints]]
2240 ===== Create a tracepoint definition
2242 A _tracepoint definition_ defines, for a given tracepoint:
2244 * Its **input arguments**. They are the macro parameters that the
2245 `tracepoint()` macro accepts for this particular tracepoint
2246 in the user application's source code.
2247 * Its **output event fields**. They are the sources of event fields
2248 that form the payload of any event that the execution of the
2249 `tracepoint()` macro emits for this particular tracepoint.
2251 You can create a tracepoint definition by using the
2252 `TRACEPOINT_EVENT()` macro below the `#include <lttng/tracepoint.h>`
2254 <<tpp-header,tracepoint provider header file template>>.
2256 The syntax of the `TRACEPOINT_EVENT()` macro is:
2259 .`TRACEPOINT_EVENT()` macro syntax.
2262 /* Tracepoint provider name */
2265 /* Tracepoint name */
2268 /* Input arguments */
2273 /* Output event fields */
2282 * `provider_name` with your tracepoint provider name.
2283 * `tracepoint_name` with your tracepoint name.
2284 * `arguments` with the <<tpp-def-input-args,input arguments>>.
2285 * `fields` with the <<tpp-def-output-fields,output event field>>
2288 This tracepoint emits events named `provider_name:tracepoint_name`.
2291 .Event name's length limitation
2293 The concatenation of the tracepoint provider name and the
2294 tracepoint name must not exceed **254 characters**. If it does, the
2295 instrumented application compiles and runs, but LTTng throws multiple
2296 warnings and you could experience serious issues.
2299 [[tpp-def-input-args]]The syntax of the `TP_ARGS()` macro is:
2302 .`TP_ARGS()` macro syntax.
2311 * `type` with the C type of the argument.
2312 * `arg_name` with the argument name.
2314 You can repeat `type` and `arg_name` up to 10 times to have
2315 more than one argument.
2317 .`TP_ARGS()` usage with three arguments.
2329 The `TP_ARGS()` and `TP_ARGS(void)` forms are valid to create a
2330 tracepoint definition with no input arguments.
2332 [[tpp-def-output-fields]]The `TP_FIELDS()` macro contains a list of
2333 `ctf_*()` macros. Each `ctf_*()` macro defines one event field. See
2334 man:lttng-ust(3) for a complete description of the available `ctf_*()`
2335 macros. A `ctf_*()` macro specifies the type, size, and byte order of
2338 Each `ctf_*()` macro takes an _argument expression_ parameter. This is a
2339 C expression that the tracer evalutes at the `tracepoint()` macro site
2340 in the application's source code. This expression provides a field's
2341 source of data. The argument expression can include input argument names
2342 listed in the `TP_ARGS()` macro.
2344 Each `ctf_*()` macro also takes a _field name_ parameter. Field names
2345 must be unique within a given tracepoint definition.
2347 Here's a complete tracepoint definition example:
2349 .Tracepoint definition.
2351 The following tracepoint definition defines a tracepoint which takes
2352 three input arguments and has four output event fields.
2356 #include "my-custom-structure.h"
2362 const struct my_custom_structure*, my_custom_structure,
2367 ctf_string(query_field, query)
2368 ctf_float(double, ratio_field, ratio)
2369 ctf_integer(int, recv_size, my_custom_structure->recv_size)
2370 ctf_integer(int, send_size, my_custom_structure->send_size)
2375 You can refer to this tracepoint definition with the `tracepoint()`
2376 macro in your application's source code like this:
2380 tracepoint(my_provider, my_tracepoint,
2381 my_structure, some_ratio, the_query);
2385 NOTE: The LTTng tracer only evaluates tracepoint arguments at run time
2386 if they satisfy an enabled <<event,event rule>>.
2389 [[using-tracepoint-classes]]
2390 ===== Use a tracepoint class
2392 A _tracepoint class_ is a class of tracepoints which share the same
2393 output event field definitions. A _tracepoint instance_ is one
2394 instance of such a defined tracepoint class, with its own tracepoint
2397 The <<defining-tracepoints,`TRACEPOINT_EVENT()` macro>> is actually a
2398 shorthand which defines both a tracepoint class and a tracepoint
2399 instance at the same time.
2401 When you build a tracepoint provider package, the C or $$C++$$ compiler
2402 creates one serialization function for each **tracepoint class**. A
2403 serialization function is responsible for serializing the event fields
2404 of a tracepoint to a sub-buffer when tracing.
2406 For various performance reasons, when your situation requires multiple
2407 tracepoint definitions with different names, but with the same event
2408 fields, we recommend that you manually create a tracepoint class
2409 and instantiate as many tracepoint instances as needed. One positive
2410 effect of such a design, amongst other advantages, is that all
2411 tracepoint instances of the same tracepoint class reuse the same
2412 serialization function, thus reducing
2413 https://en.wikipedia.org/wiki/Cache_pollution[cache pollution].
2415 .Use a tracepoint class and tracepoint instances.
2417 Consider the following three tracepoint definitions:
2429 ctf_integer(int, userid, userid)
2430 ctf_integer(size_t, len, len)
2442 ctf_integer(int, userid, userid)
2443 ctf_integer(size_t, len, len)
2455 ctf_integer(int, userid, userid)
2456 ctf_integer(size_t, len, len)
2461 In this case, we create three tracepoint classes, with one implicit
2462 tracepoint instance for each of them: `get_account`, `get_settings`, and
2463 `get_transaction`. However, they all share the same event field names
2464 and types. Hence three identical, yet independent serialization
2465 functions are created when you build the tracepoint provider package.
2467 A better design choice is to define a single tracepoint class and three
2468 tracepoint instances:
2472 /* The tracepoint class */
2473 TRACEPOINT_EVENT_CLASS(
2474 /* Tracepoint provider name */
2477 /* Tracepoint class name */
2480 /* Input arguments */
2486 /* Output event fields */
2488 ctf_integer(int, userid, userid)
2489 ctf_integer(size_t, len, len)
2493 /* The tracepoint instances */
2494 TRACEPOINT_EVENT_INSTANCE(
2495 /* Tracepoint provider name */
2498 /* Tracepoint class name */
2501 /* Tracepoint name */
2504 /* Input arguments */
2510 TRACEPOINT_EVENT_INSTANCE(
2519 TRACEPOINT_EVENT_INSTANCE(
2532 [[assigning-log-levels]]
2533 ===== Assign a log level to a tracepoint definition
2535 You can assign an optional _log level_ to a
2536 <<defining-tracepoints,tracepoint definition>>.
2538 Assigning different levels of severity to tracepoint definitions can
2539 be useful: when you <<enabling-disabling-events,create an event rule>>,
2540 you can target tracepoints having a log level as severe as a specific
2543 The concept of LTTng-UST log levels is similar to the levels found
2544 in typical logging frameworks:
2546 * In a logging framework, the log level is given by the function
2547 or method name you use at the log statement site: `debug()`,
2548 `info()`, `warn()`, `error()`, and so on.
2549 * In LTTng-UST, you statically assign the log level to a tracepoint
2550 definition; any `tracepoint()` macro invocation which refers to
2551 this definition has this log level.
2553 You can assign a log level to a tracepoint definition with the
2554 `TRACEPOINT_LOGLEVEL()` macro. You must use this macro _after_ the
2555 <<defining-tracepoints,`TRACEPOINT_EVENT()`>> or
2556 <<using-tracepoint-classes,`TRACEPOINT_INSTANCE()`>> macro for a given
2559 The syntax of the `TRACEPOINT_LOGLEVEL()` macro is:
2562 .`TRACEPOINT_LOGLEVEL()` macro syntax.
2564 TRACEPOINT_LOGLEVEL(provider_name, tracepoint_name, log_level)
2569 * `provider_name` with the tracepoint provider name.
2570 * `tracepoint_name` with the tracepoint name.
2571 * `log_level` with the log level to assign to the tracepoint
2572 definition named `tracepoint_name` in the `provider_name`
2573 tracepoint provider.
2575 See man:lttng-ust(3) for a list of available log level names.
2577 .Assign the `TRACE_DEBUG_UNIT` log level to a tracepoint definition.
2581 /* Tracepoint definition */
2590 ctf_integer(int, userid, userid)
2591 ctf_integer(size_t, len, len)
2595 /* Log level assignment */
2596 TRACEPOINT_LOGLEVEL(my_app, get_transaction, TRACE_DEBUG_UNIT)
2602 ===== Create a tracepoint provider package source file
2604 A _tracepoint provider package source file_ is a C source file which
2605 includes a <<tpp-header,tracepoint provider header file>> to expand its
2606 macros into event serialization and other functions.
2608 You can always use the following tracepoint provider package source
2612 .Tracepoint provider package source file template.
2614 #define TRACEPOINT_CREATE_PROBES
2619 Replace `tp.h` with the name of your <<tpp-header,tracepoint provider
2620 header file>> name. You may also include more than one tracepoint
2621 provider header file here to create a tracepoint provider package
2622 holding more than one tracepoint providers.
2625 [[probing-the-application-source-code]]
2626 ==== Add tracepoints to an application's source code
2628 Once you <<tpp-header,create a tracepoint provider header file>>, you
2629 can use the `tracepoint()` macro in your application's
2630 source code to insert the tracepoints that this header
2631 <<defining-tracepoints,defined>> defines.
2633 The `tracepoint()` macro takes at least two parameters: the tracepoint
2634 provider name and the tracepoint name. The corresponding tracepoint
2635 definition defines the other parameters.
2637 .`tracepoint()` usage.
2639 The following <<defining-tracepoints,tracepoint definition>> defines a
2640 tracepoint which takes two input arguments and has two output event
2644 .Tracepoint provider header file.
2646 #include "my-custom-structure.h"
2653 const char*, cmd_name
2656 ctf_string(cmd_name, cmd_name)
2657 ctf_integer(int, number_of_args, argc)
2662 You can refer to this tracepoint definition with the `tracepoint()`
2663 macro in your application's source code like this:
2666 .Application's source file.
2670 int main(int argc, char* argv[])
2672 tracepoint(my_provider, my_tracepoint, argc, argv[0]);
2678 Note how the application's source code includes
2679 the tracepoint provider header file containing the tracepoint
2680 definitions to use, path:{tp.h}.
2683 .`tracepoint()` usage with a complex tracepoint definition.
2685 Consider this complex tracepoint definition, where multiple event
2686 fields refer to the same input arguments in their argument expression
2690 .Tracepoint provider header file.
2692 /* For `struct stat` */
2693 #include <sys/types.h>
2694 #include <sys/stat.h>
2706 ctf_integer(int, my_constant_field, 23 + 17)
2707 ctf_integer(int, my_int_arg_field, my_int_arg)
2708 ctf_integer(int, my_int_arg_field2, my_int_arg * my_int_arg)
2709 ctf_integer(int, sum4_field, my_str_arg[0] + my_str_arg[1] +
2710 my_str_arg[2] + my_str_arg[3])
2711 ctf_string(my_str_arg_field, my_str_arg)
2712 ctf_integer_hex(off_t, size_field, st->st_size)
2713 ctf_float(double, size_dbl_field, (double) st->st_size)
2714 ctf_sequence_text(char, half_my_str_arg_field, my_str_arg,
2715 size_t, strlen(my_str_arg) / 2)
2720 You can refer to this tracepoint definition with the `tracepoint()`
2721 macro in your application's source code like this:
2724 .Application's source file.
2726 #define TRACEPOINT_DEFINE
2733 stat("/etc/fstab", &s);
2734 tracepoint(my_provider, my_tracepoint, 23, "Hello, World!", &s);
2740 If you look at the event record that LTTng writes when tracing this
2741 program, assuming the file size of path:{/etc/fstab} is 301{nbsp}bytes,
2742 it should look like this:
2744 .Event record fields
2746 |Field's name |Field's value
2747 |`my_constant_field` |40
2748 |`my_int_arg_field` |23
2749 |`my_int_arg_field2` |529
2751 |`my_str_arg_field` |`Hello, World!`
2752 |`size_field` |0x12d
2753 |`size_dbl_field` |301.0
2754 |`half_my_str_arg_field` |`Hello,`
2758 Sometimes, the arguments you pass to `tracepoint()` are expensive to
2759 compute--they use the call stack, for example. To avoid this
2760 computation when the tracepoint is disabled, you can use the
2761 `tracepoint_enabled()` and `do_tracepoint()` macros.
2763 The syntax of the `tracepoint_enabled()` and `do_tracepoint()` macros
2767 .`tracepoint_enabled()` and `do_tracepoint()` macros syntax.
2769 tracepoint_enabled(provider_name, tracepoint_name)
2770 do_tracepoint(provider_name, tracepoint_name, ...)
2775 * `provider_name` with the tracepoint provider name.
2776 * `tracepoint_name` with the tracepoint name.
2778 `tracepoint_enabled()` returns a non-zero value if the tracepoint named
2779 `tracepoint_name` from the provider named `provider_name` is enabled
2782 `do_tracepoint()` is like `tracepoint()`, except that it doesn't check
2783 if the tracepoint is enabled. Using `tracepoint()` with
2784 `tracepoint_enabled()` is dangerous since `tracepoint()` also contains
2785 the `tracepoint_enabled()` check, thus a race condition is
2786 possible in this situation:
2789 .Possible race condition when using `tracepoint_enabled()` with `tracepoint()`.
2791 if (tracepoint_enabled(my_provider, my_tracepoint)) {
2792 stuff = prepare_stuff();
2795 tracepoint(my_provider, my_tracepoint, stuff);
2798 If the tracepoint is enabled after the condition, then `stuff` is not
2799 prepared: the emitted event will either contain wrong data, or the whole
2800 application could crash (segmentation fault, for example).
2802 NOTE: Neither `tracepoint_enabled()` nor `do_tracepoint()` have an
2803 `STAP_PROBEV()` call. If you need it, you must emit
2807 [[building-tracepoint-providers-and-user-application]]
2808 ==== Build and link a tracepoint provider package and an application
2810 Once you have one or more <<tpp-header,tracepoint provider header
2811 files>> and a <<tpp-source,tracepoint provider package source file>>,
2812 you can create the tracepoint provider package by compiling its source
2813 file. From here, multiple build and run scenarios are possible. The
2814 following table shows common application and library configurations
2815 along with the required command lines to achieve them.
2817 In the following diagrams, we use the following file names:
2820 Executable application.
2823 Application's object file.
2826 Tracepoint provider package object file.
2829 Tracepoint provider package archive file.
2832 Tracepoint provider package shared object file.
2835 User library object file.
2838 User library shared object file.
2840 The red star indicates that this object file is instrumented
2841 (contains code which uses the `tracepoint()` macro). The spring
2842 symbol between the application and a library means the application is
2843 linked with the library at build time.
2845 We assume that path:{.} is part of the env:LD_LIBRARY_PATH environment
2846 variable in the following instructions.
2848 [role="growable ust-scenarios",cols="asciidoc,asciidoc"]
2849 .Common tracepoint provider package scenarios.
2851 |Scenario |Instructions
2854 The instrumented application is statically linked with
2855 the tracepoint provider package object.
2857 image::ust-sit+app-linked-with-tp-o+app-instrumented.png[]
2860 include::../common/ust-sit-step-tp-o.txt[]
2862 To build the instrumented application:
2864 . In path:{app.c}, before including path:{tpp.h}, add the following line:
2869 #define TRACEPOINT_DEFINE
2873 . Compile the application source file:
2882 . Build the application:
2887 gcc -o app app.o tpp.o -llttng-ust -ldl
2891 To run the instrumented application:
2893 * Start the application:
2903 The instrumented application is statically linked with the
2904 tracepoint provider package archive file.
2906 image::ust-sit+app-linked-with-tp-a+app-instrumented.png[]
2909 To create the tracepoint provider package archive file:
2911 . Compile the <<tpp-source,tracepoint provider package source file>>:
2920 . Create the tracepoint provider package archive file:
2929 To build the instrumented application:
2931 . In path:{app.c}, before including path:{tpp.h}, add the following line:
2936 #define TRACEPOINT_DEFINE
2940 . Compile the application source file:
2949 . Build the application:
2954 gcc -o app app.o tpp.a -llttng-ust -ldl
2958 To run the instrumented application:
2960 * Start the application:
2970 The instrumented application is linked with the tracepoint provider
2971 package shared object.
2973 image::ust-sit+app-linked-with-tp-so+app-instrumented.png[]
2976 include::../common/ust-sit-step-tp-so.txt[]
2978 To build the instrumented application:
2980 . In path:{app.c}, before including path:{tpp.h}, add the following line:
2985 #define TRACEPOINT_DEFINE
2989 . Compile the application source file:
2998 . Build the application:
3003 gcc -o app app.o -ldl -L. -ltpp
3007 To run the instrumented application:
3009 * Start the application:
3019 The tracepoint provider package shared object is preloaded before the
3020 instrumented application starts.
3022 image::ust-sit+tp-so-preloaded+app-instrumented.png[]
3025 include::../common/ust-sit-step-tp-so.txt[]
3027 To build the instrumented application:
3029 . In path:{app.c}, before including path:{tpp.h}, add the
3035 #define TRACEPOINT_DEFINE
3036 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3040 . Compile the application source file:
3049 . Build the application:
3054 gcc -o app app.o -ldl
3058 To run the instrumented application with tracing support:
3060 * Preload the tracepoint provider package shared object and
3061 start the application:
3066 LD_PRELOAD=./libtpp.so ./app
3070 To run the instrumented application without tracing support:
3072 * Start the application:
3082 The instrumented application dynamically loads the tracepoint provider
3083 package shared object.
3085 See the <<dlclose-warning,warning about `dlclose()`>>.
3087 image::ust-sit+app-dlopens-tp-so+app-instrumented.png[]
3090 include::../common/ust-sit-step-tp-so.txt[]
3092 To build the instrumented application:
3094 . In path:{app.c}, before including path:{tpp.h}, add the
3100 #define TRACEPOINT_DEFINE
3101 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3105 . Compile the application source file:
3114 . Build the application:
3119 gcc -o app app.o -ldl
3123 To run the instrumented application:
3125 * Start the application:
3135 The application is linked with the instrumented user library.
3137 The instrumented user library is statically linked with the tracepoint
3138 provider package object file.
3140 image::ust-sit+app-linked-with-lib+lib-linked-with-tp-o+lib-instrumented.png[]
3143 include::../common/ust-sit-step-tp-o-fpic.txt[]
3145 To build the instrumented user library:
3147 . In path:{emon.c}, before including path:{tpp.h}, add the
3153 #define TRACEPOINT_DEFINE
3157 . Compile the user library source file:
3162 gcc -I. -fpic -c emon.c
3166 . Build the user library shared object:
3171 gcc -shared -o libemon.so emon.o tpp.o -llttng-ust -ldl
3175 To build the application:
3177 . Compile the application source file:
3186 . Build the application:
3191 gcc -o app app.o -L. -lemon
3195 To run the application:
3197 * Start the application:
3207 The application is linked with the instrumented user library.
3209 The instrumented user library is linked with the tracepoint provider
3210 package shared object.
3212 image::ust-sit+app-linked-with-lib+lib-linked-with-tp-so+lib-instrumented.png[]
3215 include::../common/ust-sit-step-tp-so.txt[]
3217 To build the instrumented user library:
3219 . In path:{emon.c}, before including path:{tpp.h}, add the
3225 #define TRACEPOINT_DEFINE
3229 . Compile the user library source file:
3234 gcc -I. -fpic -c emon.c
3238 . Build the user library shared object:
3243 gcc -shared -o libemon.so emon.o -ldl -L. -ltpp
3247 To build the application:
3249 . Compile the application source file:
3258 . Build the application:
3263 gcc -o app app.o -L. -lemon
3267 To run the application:
3269 * Start the application:
3279 The tracepoint provider package shared object is preloaded before the
3282 The application is linked with the instrumented user library.
3284 image::ust-sit+tp-so-preloaded+app-linked-with-lib+lib-instrumented.png[]
3287 include::../common/ust-sit-step-tp-so.txt[]
3289 To build the instrumented user library:
3291 . In path:{emon.c}, before including path:{tpp.h}, add the
3297 #define TRACEPOINT_DEFINE
3298 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3302 . Compile the user library source file:
3307 gcc -I. -fpic -c emon.c
3311 . Build the user library shared object:
3316 gcc -shared -o libemon.so emon.o -ldl
3320 To build the application:
3322 . Compile the application source file:
3331 . Build the application:
3336 gcc -o app app.o -L. -lemon
3340 To run the application with tracing support:
3342 * Preload the tracepoint provider package shared object and
3343 start the application:
3348 LD_PRELOAD=./libtpp.so ./app
3352 To run the application without tracing support:
3354 * Start the application:
3364 The application is linked with the instrumented user library.
3366 The instrumented user library dynamically loads the tracepoint provider
3367 package shared object.
3369 See the <<dlclose-warning,warning about `dlclose()`>>.
3371 image::ust-sit+app-linked-with-lib+lib-dlopens-tp-so+lib-instrumented.png[]
3374 include::../common/ust-sit-step-tp-so.txt[]
3376 To build the instrumented user library:
3378 . In path:{emon.c}, before including path:{tpp.h}, add the
3384 #define TRACEPOINT_DEFINE
3385 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3389 . Compile the user library source file:
3394 gcc -I. -fpic -c emon.c
3398 . Build the user library shared object:
3403 gcc -shared -o libemon.so emon.o -ldl
3407 To build the application:
3409 . Compile the application source file:
3418 . Build the application:
3423 gcc -o app app.o -L. -lemon
3427 To run the application:
3429 * Start the application:
3439 The application dynamically loads the instrumented user library.
3441 The instrumented user library is linked with the tracepoint provider
3442 package shared object.
3444 See the <<dlclose-warning,warning about `dlclose()`>>.
3446 image::ust-sit+app-dlopens-lib+lib-linked-with-tp-so+lib-instrumented.png[]
3449 include::../common/ust-sit-step-tp-so.txt[]
3451 To build the instrumented user library:
3453 . In path:{emon.c}, before including path:{tpp.h}, add the
3459 #define TRACEPOINT_DEFINE
3463 . Compile the user library source file:
3468 gcc -I. -fpic -c emon.c
3472 . Build the user library shared object:
3477 gcc -shared -o libemon.so emon.o -ldl -L. -ltpp
3481 To build the application:
3483 . Compile the application source file:
3492 . Build the application:
3497 gcc -o app app.o -ldl -L. -lemon
3501 To run the application:
3503 * Start the application:
3513 The application dynamically loads the instrumented user library.
3515 The instrumented user library dynamically loads the tracepoint provider
3516 package shared object.
3518 See the <<dlclose-warning,warning about `dlclose()`>>.
3520 image::ust-sit+app-dlopens-lib+lib-dlopens-tp-so+lib-instrumented.png[]
3523 include::../common/ust-sit-step-tp-so.txt[]
3525 To build the instrumented user library:
3527 . In path:{emon.c}, before including path:{tpp.h}, add the
3533 #define TRACEPOINT_DEFINE
3534 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3538 . Compile the user library source file:
3543 gcc -I. -fpic -c emon.c
3547 . Build the user library shared object:
3552 gcc -shared -o libemon.so emon.o -ldl
3556 To build the application:
3558 . Compile the application source file:
3567 . Build the application:
3572 gcc -o app app.o -ldl -L. -lemon
3576 To run the application:
3578 * Start the application:
3588 The tracepoint provider package shared object is preloaded before the
3591 The application dynamically loads the instrumented user library.
3593 image::ust-sit+tp-so-preloaded+app-dlopens-lib+lib-instrumented.png[]
3596 include::../common/ust-sit-step-tp-so.txt[]
3598 To build the instrumented user library:
3600 . In path:{emon.c}, before including path:{tpp.h}, add the
3606 #define TRACEPOINT_DEFINE
3607 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3611 . Compile the user library source file:
3616 gcc -I. -fpic -c emon.c
3620 . Build the user library shared object:
3625 gcc -shared -o libemon.so emon.o -ldl
3629 To build the application:
3631 . Compile the application source file:
3640 . Build the application:
3645 gcc -o app app.o -L. -lemon
3649 To run the application with tracing support:
3651 * Preload the tracepoint provider package shared object and
3652 start the application:
3657 LD_PRELOAD=./libtpp.so ./app
3661 To run the application without tracing support:
3663 * Start the application:
3673 The application is statically linked with the tracepoint provider
3674 package object file.
3676 The application is linked with the instrumented user library.
3678 image::ust-sit+app-linked-with-tp-o+app-linked-with-lib+lib-instrumented.png[]
3681 include::../common/ust-sit-step-tp-o.txt[]
3683 To build the instrumented user library:
3685 . In path:{emon.c}, before including path:{tpp.h}, add the
3691 #define TRACEPOINT_DEFINE
3695 . Compile the user library source file:
3700 gcc -I. -fpic -c emon.c
3704 . Build the user library shared object:
3709 gcc -shared -o libemon.so emon.o
3713 To build the application:
3715 . Compile the application source file:
3724 . Build the application:
3729 gcc -o app app.o tpp.o -llttng-ust -ldl -L. -lemon
3733 To run the instrumented application:
3735 * Start the application:
3745 The application is statically linked with the tracepoint provider
3746 package object file.
3748 The application dynamically loads the instrumented user library.
3750 image::ust-sit+app-linked-with-tp-o+app-dlopens-lib+lib-instrumented.png[]
3753 include::../common/ust-sit-step-tp-o.txt[]
3755 To build the application:
3757 . In path:{app.c}, before including path:{tpp.h}, add the following line:
3762 #define TRACEPOINT_DEFINE
3766 . Compile the application source file:
3775 . Build the application:
3780 gcc -Wl,--export-dynamic -o app app.o tpp.o \
3785 The `--export-dynamic` option passed to the linker is necessary for the
3786 dynamically loaded library to ``see'' the tracepoint symbols defined in
3789 To build the instrumented user library:
3791 . Compile the user library source file:
3796 gcc -I. -fpic -c emon.c
3800 . Build the user library shared object:
3805 gcc -shared -o libemon.so emon.o
3809 To run the application:
3811 * Start the application:
3823 .Do not use man:dlclose(3) on a tracepoint provider package
3825 Never use man:dlclose(3) on any shared object which:
3827 * Is linked with, statically or dynamically, a tracepoint provider
3829 * Calls man:dlopen(3) itself to dynamically open a tracepoint provider
3830 package shared object.
3832 This is currently considered **unsafe** due to a lack of reference
3833 counting from LTTng-UST to the shared object.
3835 A known workaround (available since glibc 2.2) is to use the
3836 `RTLD_NODELETE` flag when calling man:dlopen(3) initially. This has the
3837 effect of not unloading the loaded shared object, even if man:dlclose(3)
3840 You can also preload the tracepoint provider package shared object with
3841 the env:LD_PRELOAD environment variable to overcome this limitation.
3845 [[using-lttng-ust-with-daemons]]
3846 ===== Use noch:{LTTng-UST} with daemons
3848 If your instrumented application calls man:fork(2), man:clone(2),
3849 or BSD's man:rfork(2), without a following man:exec(3)-family
3850 system call, you must preload the path:{liblttng-ust-fork.so} shared
3851 object when starting the application.
3855 LD_PRELOAD=liblttng-ust-fork.so ./my-app
3858 If your tracepoint provider package is
3859 a shared library which you also preload, you must put both
3860 shared objects in env:LD_PRELOAD:
3864 LD_PRELOAD=liblttng-ust-fork.so:/path/to/tp.so ./my-app
3868 [[lttng-ust-pkg-config]]
3869 ===== Use noch:{pkg-config}
3871 On some distributions, LTTng-UST ships with a
3872 https://www.freedesktop.org/wiki/Software/pkg-config/[pkg-config]
3873 metadata file. If this is your case, then you can use cmd:pkg-config to
3874 build an application on the command line:
3878 gcc -o my-app my-app.o tp.o $(pkg-config --cflags --libs lttng-ust)
3882 [[instrumenting-32-bit-app-on-64-bit-system]]
3883 ===== [[advanced-instrumenting-techniques]]Build a 32-bit instrumented application for a 64-bit target system
3885 In order to trace a 32-bit application running on a 64-bit system,
3886 LTTng must use a dedicated 32-bit
3887 <<lttng-consumerd,consumer daemon>>.
3889 The following steps show how to build and install a 32-bit consumer
3890 daemon, which is _not_ part of the default 64-bit LTTng build, how to
3891 build and install the 32-bit LTTng-UST libraries, and how to build and
3892 link an instrumented 32-bit application in that context.
3894 To build a 32-bit instrumented application for a 64-bit target system,
3895 assuming you have a fresh target system with no installed Userspace RCU
3898 . Download, build, and install a 32-bit version of Userspace RCU:
3904 wget http://lttng.org/files/urcu/userspace-rcu-latest-0.9.tar.bz2 &&
3905 tar -xf userspace-rcu-latest-0.9.tar.bz2 &&
3906 cd userspace-rcu-0.9.* &&
3907 ./configure --libdir=/usr/local/lib32 CFLAGS=-m32 &&
3909 sudo make install &&
3914 . Using your distribution's package manager, or from source, install
3915 the following 32-bit versions of the following dependencies of
3916 LTTng-tools and LTTng-UST:
3919 * https://sourceforge.net/projects/libuuid/[libuuid]
3920 * http://directory.fsf.org/wiki/Popt[popt]
3921 * http://www.xmlsoft.org/[libxml2]
3924 . Download, build, and install a 32-bit version of the latest
3925 LTTng-UST{nbsp}{revision}:
3931 wget http://lttng.org/files/lttng-ust/lttng-ust-latest-2.8.tar.bz2 &&
3932 tar -xf lttng-ust-latest-2.8.tar.bz2 &&
3933 cd lttng-ust-2.8.* &&
3934 ./configure --libdir=/usr/local/lib32 \
3935 CFLAGS=-m32 CXXFLAGS=-m32 \
3936 LDFLAGS='-L/usr/local/lib32 -L/usr/lib32' &&
3938 sudo make install &&
3945 Depending on your distribution,
3946 32-bit libraries could be installed at a different location than
3947 `/usr/lib32`. For example, Debian is known to install
3948 some 32-bit libraries in `/usr/lib/i386-linux-gnu`.
3950 In this case, make sure to set `LDFLAGS` to all the
3951 relevant 32-bit library paths, for example:
3955 LDFLAGS='-L/usr/lib/i386-linux-gnu -L/usr/lib32'
3959 . Download the latest LTTng-tools{nbsp}{revision}, build, and install
3960 the 32-bit consumer daemon:
3966 wget http://lttng.org/files/lttng-tools/lttng-tools-latest-2.8.tar.bz2 &&
3967 tar -xf lttng-tools-latest-2.8.tar.bz2 &&
3968 cd lttng-tools-2.8.* &&
3969 ./configure --libdir=/usr/local/lib32 CFLAGS=-m32 CXXFLAGS=-m32 \
3970 LDFLAGS='-L/usr/local/lib32 -L/usr/lib32' &&
3972 cd src/bin/lttng-consumerd &&
3973 sudo make install &&
3978 . From your distribution or from source,
3979 <<installing-lttng,install>> the 64-bit versions of
3980 LTTng-UST and Userspace RCU.
3981 . Download, build, and install the 64-bit version of the
3982 latest LTTng-tools{nbsp}{revision}:
3988 wget http://lttng.org/files/lttng-tools/lttng-tools-latest-2.8.tar.bz2 &&
3989 tar -xf lttng-tools-latest-2.8.tar.bz2 &&
3990 cd lttng-tools-2.8.* &&
3991 ./configure --with-consumerd32-libdir=/usr/local/lib32 \
3992 --with-consumerd32-bin=/usr/local/lib32/lttng/libexec/lttng-consumerd &&
3994 sudo make install &&
3999 . Pass the following options to man:gcc(1), man:g++(1), or man:clang(1)
4000 when linking your 32-bit application:
4003 -m32 -L/usr/lib32 -L/usr/local/lib32 \
4004 -Wl,-rpath,/usr/lib32,-rpath,/usr/local/lib32
4007 For example, let's rebuild the quick start example in
4008 <<tracing-your-own-user-application,Trace a user application>> as an
4009 instrumented 32-bit application:
4014 gcc -m32 -c -I. hello-tp.c
4016 gcc -m32 -o hello hello.o hello-tp.o \
4017 -L/usr/lib32 -L/usr/local/lib32 \
4018 -Wl,-rpath,/usr/lib32,-rpath,/usr/local/lib32 \
4023 No special action is required to execute the 32-bit application and
4024 to trace it: use the command-line man:lttng(1) tool as usual.
4031 man:tracef(3) is a small LTTng-UST API designed for quick,
4032 man:printf(3)-like instrumentation without the burden of
4033 <<tracepoint-provider,creating>> and
4034 <<building-tracepoint-providers-and-user-application,building>>
4035 a tracepoint provider package.
4037 To use `tracef()` in your application:
4039 . In the C or C++ source files where you need to use `tracef()`,
4040 include `<lttng/tracef.h>`:
4045 #include <lttng/tracef.h>
4049 . In the application's source code, use `tracef()` like you would use
4057 tracef("my message: %d (%s)", my_integer, my_string);
4063 . Link your application with `liblttng-ust`:
4068 gcc -o app app.c -llttng-ust
4072 To trace the events that `tracef()` calls emit:
4074 * <<enabling-disabling-events,Create an event rule>> which matches the
4075 `lttng_ust_tracef:*` event name:
4080 lttng enable-event --userspace 'lttng_ust_tracef:*'
4085 .Limitations of `tracef()`
4087 The `tracef()` utility function was developed to make user space tracing
4088 super simple, albeit with notable disadvantages compared to
4089 <<defining-tracepoints,user-defined tracepoints>>:
4091 * All the emitted events have the same tracepoint provider and
4092 tracepoint names, respectively `lttng_ust_tracef` and `event`.
4093 * There is no static type checking.
4094 * The only event record field you actually get, named `msg`, is a string
4095 potentially containing the values you passed to `tracef()`
4096 using your own format string. This also means that you cannot filter
4097 events with a custom expression at run time because there are no
4099 * Since `tracef()` uses the C standard library's man:vasprintf(3)
4100 function behind the scenes to format the strings at run time, its
4101 expected performance is lower than with user-defined tracepoints,
4102 which do not require a conversion to a string.
4104 Taking this into consideration, `tracef()` is useful for some quick
4105 prototyping and debugging, but you should not consider it for any
4106 permanent and serious applicative instrumentation.
4112 ==== Use `tracelog()`
4114 The man:tracelog(3) API is very similar to <<tracef,`tracef()`>>, with
4115 the difference that it accepts an additional log level parameter.
4117 The goal of `tracelog()` is to ease the migration from logging to
4120 To use `tracelog()` in your application:
4122 . In the C or C++ source files where you need to use `tracelog()`,
4123 include `<lttng/tracelog.h>`:
4128 #include <lttng/tracelog.h>
4132 . In the application's source code, use `tracelog()` like you would use
4133 man:printf(3), except for the first parameter which is the log
4141 tracelog(TRACE_WARNING, "my message: %d (%s)",
4142 my_integer, my_string);
4148 See man:lttng-ust(3) for a list of available log level names.
4150 . Link your application with `liblttng-ust`:
4155 gcc -o app app.c -llttng-ust
4159 To trace the events that `tracelog()` calls emit with a log level
4160 _as severe as_ a specific log level:
4162 * <<enabling-disabling-events,Create an event rule>> which matches the
4163 `lttng_ust_tracelog:*` event name and a minimum level
4169 lttng enable-event --userspace 'lttng_ust_tracelog:*'
4170 --loglevel=TRACE_WARNING
4174 To trace the events that `tracelog()` calls emit with a
4175 _specific log level_:
4177 * Create an event rule which matches the `lttng_ust_tracelog:*`
4178 event name and a specific log level:
4183 lttng enable-event --userspace 'lttng_ust_tracelog:*'
4184 --loglevel-only=TRACE_INFO
4189 [[prebuilt-ust-helpers]]
4190 === Prebuilt user space tracing helpers
4192 The LTTng-UST package provides a few helpers in the form or preloadable
4193 shared objects which automatically instrument system functions and
4196 The helper shared objects are normally found in dir:{/usr/lib}. If you
4197 built LTTng-UST <<building-from-source,from source>>, they are probably
4198 located in dir:{/usr/local/lib}.
4200 The installed user space tracing helpers in LTTng-UST{nbsp}{revision}
4203 path:{liblttng-ust-libc-wrapper.so}::
4204 path:{liblttng-ust-pthread-wrapper.so}::
4205 <<liblttng-ust-libc-pthread-wrapper,C{nbsp}standard library
4206 memory and POSIX threads function tracing>>.
4208 path:{liblttng-ust-cyg-profile.so}::
4209 path:{liblttng-ust-cyg-profile-fast.so}::
4210 <<liblttng-ust-cyg-profile,Function entry and exit tracing>>.
4212 path:{liblttng-ust-dl.so}::
4213 <<liblttng-ust-dl,Dynamic linker tracing>>.
4215 To use a user space tracing helper with any user application:
4217 * Preload the helper shared object when you start the application:
4222 LD_PRELOAD=liblttng-ust-libc-wrapper.so my-app
4226 You can preload more than one helper:
4231 LD_PRELOAD=liblttng-ust-libc-wrapper.so:liblttng-ust-dl.so my-app
4237 [[liblttng-ust-libc-pthread-wrapper]]
4238 ==== Instrument C standard library memory and POSIX threads functions
4240 The path:{liblttng-ust-libc-wrapper.so} and
4241 path:{liblttng-ust-pthread-wrapper.so} helpers
4242 add instrumentation to some C standard library and POSIX
4246 .Functions instrumented by preloading path:{liblttng-ust-libc-wrapper.so}.
4248 |TP provider name |TP name |Instrumented function
4250 .6+|`lttng_ust_libc` |`malloc` |man:malloc(3)
4251 |`calloc` |man:calloc(3)
4252 |`realloc` |man:realloc(3)
4253 |`free` |man:free(3)
4254 |`memalign` |man:memalign(3)
4255 |`posix_memalign` |man:posix_memalign(3)
4259 .Functions instrumented by preloading path:{liblttng-ust-pthread-wrapper.so}.
4261 |TP provider name |TP name |Instrumented function
4263 .4+|`lttng_ust_pthread` |`pthread_mutex_lock_req` |man:pthread_mutex_lock(3p) (request time)
4264 |`pthread_mutex_lock_acq` |man:pthread_mutex_lock(3p) (acquire time)
4265 |`pthread_mutex_trylock` |man:pthread_mutex_trylock(3p)
4266 |`pthread_mutex_unlock` |man:pthread_mutex_unlock(3p)
4269 When you preload the shared object, it replaces the functions listed
4270 in the previous tables by wrappers which contain tracepoints and call
4271 the replaced functions.
4274 [[liblttng-ust-cyg-profile]]
4275 ==== Instrument function entry and exit
4277 The path:{liblttng-ust-cyg-profile*.so} helpers can add instrumentation
4278 to the entry and exit points of functions.
4280 man:gcc(1) and man:clang(1) have an option named
4281 https://gcc.gnu.org/onlinedocs/gcc/Code-Gen-Options.html[`-finstrument-functions`]
4282 which generates instrumentation calls for entry and exit to functions.
4283 The LTTng-UST function tracing helpers,
4284 path:{liblttng-ust-cyg-profile.so} and
4285 path:{liblttng-ust-cyg-profile-fast.so}, take advantage of this feature
4286 to add tracepoints to the two generated functions (which contain
4287 `cyg_profile` in their names, hence the helper's name).
4289 To use the LTTng-UST function tracing helper, the source files to
4290 instrument must be built using the `-finstrument-functions` compiler
4293 There are two versions of the LTTng-UST function tracing helper:
4295 * **path:{liblttng-ust-cyg-profile-fast.so}** is a lightweight variant
4296 that you should only use when it can be _guaranteed_ that the
4297 complete event stream is recorded without any lost event record.
4298 Any kind of duplicate information is left out.
4300 Assuming no event record is lost, having only the function addresses on
4301 entry is enough to create a call graph, since an event record always
4302 contains the ID of the CPU that generated it.
4304 You can use a tool like
4305 https://sourceware.org/binutils/docs/binutils/addr2line.html[cmd:addr2line]
4306 to convert function addresses back to source file names and
4309 * **path:{liblttng-ust-cyg-profile.so}** is a more robust variant
4310 which also works in use cases where event records might get discarded or
4311 not recorded from application startup.
4312 In these cases, the trace analyzer needs more information to be
4313 able to reconstruct the program flow.
4315 See man:lttng-ust-cyg-profile(3) to learn more about the instrumentation
4316 points of this helper.
4318 All the tracepoints that this helper provides have the
4319 log level `TRACE_DEBUG_FUNCTION` (see man:lttng-ust(3)).
4321 TIP: It's sometimes a good idea to limit the number of source files that
4322 you compile with the `-finstrument-functions` option to prevent LTTng
4323 from writing an excessive amount of trace data at run time. When using
4324 man:gcc(1), you can use the
4325 `-finstrument-functions-exclude-function-list` option to avoid
4326 instrument entries and exits of specific function names.
4331 ==== Instrument the dynamic linker
4333 The path:{liblttng-ust-dl.so} helper adds instrumentation to the
4334 man:dlopen(3) and man:dlclose(3) function calls.
4336 See man:lttng-ust-dl(3) to learn more about the instrumentation points
4341 [[java-application]]
4342 === User space Java agent
4344 You can instrument any Java application which uses one of the following
4347 * The https://docs.oracle.com/javase/7/docs/api/java/util/logging/package-summary.html[**`java.util.logging`**]
4348 (JUL) core logging facilities.
4349 * http://logging.apache.org/log4j/1.2/[**Apache log4j 1.2**], since
4350 LTTng 2.6. Note that Apache Log4j{nbsp}2 is not supported.
4353 .LTTng-UST Java agent imported by a Java application.
4354 image::java-app.png[]
4356 Note that the methods described below are new in LTTng{nbsp}{revision}.
4357 Previous LTTng versions use another technique.
4359 NOTE: We use http://openjdk.java.net/[OpenJDK]{nbsp}8 for development
4360 and https://ci.lttng.org/[continuous integration], thus this version is
4361 directly supported. However, the LTTng-UST Java agent is also tested
4362 with OpenJDK{nbsp}7.
4367 ==== Use the LTTng-UST Java agent for `java.util.logging`
4369 To use the LTTng-UST Java agent in a Java application which uses
4370 `java.util.logging` (JUL):
4372 . In the Java application's source code, import the LTTng-UST
4373 log handler package for `java.util.logging`:
4378 import org.lttng.ust.agent.jul.LttngLogHandler;
4382 . Create an LTTng-UST JUL log handler:
4387 Handler lttngUstLogHandler = new LttngLogHandler();
4391 . Add this handler to the JUL loggers which should emit LTTng events:
4396 Logger myLogger = Logger.getLogger("some-logger");
4398 myLogger.addHandler(lttngUstLogHandler);
4402 . Use `java.util.logging` log statements and configuration as usual.
4403 The loggers with an attached LTTng-UST log handler can emit
4406 . Before exiting the application, remove the LTTng-UST log handler from
4407 the loggers attached to it and call its `close()` method:
4412 myLogger.removeHandler(lttngUstLogHandler);
4413 lttngUstLogHandler.close();
4417 This is not strictly necessary, but it is recommended for a clean
4418 disposal of the handler's resources.
4420 . Include the LTTng-UST Java agent's common and JUL-specific JAR files,
4421 path:{lttng-ust-agent-common.jar} and path:{lttng-ust-agent-jul.jar},
4423 https://docs.oracle.com/javase/tutorial/essential/environment/paths.html[class
4424 path] when you build the Java application.
4426 The JAR files are typically located in dir:{/usr/share/java}.
4428 IMPORTANT: The LTTng-UST Java agent must be
4429 <<installing-lttng,installed>> for the logging framework your
4432 .Use the LTTng-UST Java agent for `java.util.logging`.
4437 import java.io.IOException;
4438 import java.util.logging.Handler;
4439 import java.util.logging.Logger;
4440 import org.lttng.ust.agent.jul.LttngLogHandler;
4444 private static final int answer = 42;
4446 public static void main(String[] argv) throws Exception
4449 Logger logger = Logger.getLogger("jello");
4451 // Create an LTTng-UST log handler
4452 Handler lttngUstLogHandler = new LttngLogHandler();
4454 // Add the LTTng-UST log handler to our logger
4455 logger.addHandler(lttngUstLogHandler);
4458 logger.info("some info");
4459 logger.warning("some warning");
4461 logger.finer("finer information; the answer is " + answer);
4463 logger.severe("error!");
4465 // Not mandatory, but cleaner
4466 logger.removeHandler(lttngUstLogHandler);
4467 lttngUstLogHandler.close();
4476 javac -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-jul.jar Test.java
4479 <<creating-destroying-tracing-sessions,Create a tracing session>>,
4480 <<enabling-disabling-events,create an event rule>> matching the
4481 `jello` JUL logger, and <<basic-tracing-session-control,start tracing>>:
4486 lttng enable-event --jul jello
4490 Run the compiled class:
4494 java -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-jul.jar:. Test
4497 <<basic-tracing-session-control,Stop tracing>> and inspect the
4507 You can use the opt:lttng-enable-event(1):--loglevel or
4508 opt:lttng-enable-event(1):--loglevel-only option of the
4509 man:lttng-enable-event(1) command to target a range of JUL log levels
4510 or a specific JUL log level.
4515 ==== Use the LTTng-UST Java agent for Apache log4j
4517 To use the LTTng-UST Java agent in a Java application which uses
4520 . In the Java application's source code, import the LTTng-UST
4521 log appender package for Apache log4j:
4526 import org.lttng.ust.agent.log4j.LttngLogAppender;
4530 . Create an LTTng-UST log4j log appender:
4535 Appender lttngUstLogAppender = new LttngLogAppender();
4539 . Add this appender to the log4j loggers which should emit LTTng events:
4544 Logger myLogger = Logger.getLogger("some-logger");
4546 myLogger.addAppender(lttngUstLogAppender);
4550 . Use Apache log4j log statements and configuration as usual. The
4551 loggers with an attached LTTng-UST log appender can emit LTTng events.
4553 . Before exiting the application, remove the LTTng-UST log appender from
4554 the loggers attached to it and call its `close()` method:
4559 myLogger.removeAppender(lttngUstLogAppender);
4560 lttngUstLogAppender.close();
4564 This is not strictly necessary, but it is recommended for a clean
4565 disposal of the appender's resources.
4567 . Include the LTTng-UST Java agent's common and log4j-specific JAR
4568 files, path:{lttng-ust-agent-common.jar} and
4569 path:{lttng-ust-agent-log4j.jar}, in the
4570 https://docs.oracle.com/javase/tutorial/essential/environment/paths.html[class
4571 path] when you build the Java application.
4573 The JAR files are typically located in dir:{/usr/share/java}.
4575 IMPORTANT: The LTTng-UST Java agent must be
4576 <<installing-lttng,installed>> for the logging framework your
4579 .Use the LTTng-UST Java agent for Apache log4j.
4584 import org.apache.log4j.Appender;
4585 import org.apache.log4j.Logger;
4586 import org.lttng.ust.agent.log4j.LttngLogAppender;
4590 private static final int answer = 42;
4592 public static void main(String[] argv) throws Exception
4595 Logger logger = Logger.getLogger("jello");
4597 // Create an LTTng-UST log appender
4598 Appender lttngUstLogAppender = new LttngLogAppender();
4600 // Add the LTTng-UST log appender to our logger
4601 logger.addAppender(lttngUstLogAppender);
4604 logger.info("some info");
4605 logger.warn("some warning");
4607 logger.debug("debug information; the answer is " + answer);
4609 logger.fatal("error!");
4611 // Not mandatory, but cleaner
4612 logger.removeAppender(lttngUstLogAppender);
4613 lttngUstLogAppender.close();
4619 Build this example (`$LOG4JPATH` is the path to the Apache log4j JAR
4624 javac -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-log4j.jar:$LOG4JPATH Test.java
4627 <<creating-destroying-tracing-sessions,Create a tracing session>>,
4628 <<enabling-disabling-events,create an event rule>> matching the
4629 `jello` log4j logger, and <<basic-tracing-session-control,start tracing>>:
4634 lttng enable-event --log4j jello
4638 Run the compiled class:
4642 java -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-log4j.jar:$LOG4JPATH:. Test
4645 <<basic-tracing-session-control,Stop tracing>> and inspect the
4655 You can use the opt:lttng-enable-event(1):--loglevel or
4656 opt:lttng-enable-event(1):--loglevel-only option of the
4657 man:lttng-enable-event(1) command to target a range of Apache log4j log levels
4658 or a specific log4j log level.
4662 [[java-application-context]]
4663 ==== Provide application-specific context fields in a Java application
4665 A Java application-specific context field is a piece of state provided
4666 by the application which <<adding-context,you can add>>, using the
4667 man:lttng-add-context(1) command, to each <<event,event record>>
4668 produced by the log statements of this application.
4670 For example, a given object might have a current request ID variable.
4671 You can create a context information retriever for this object and
4672 assign a name to this current request ID. You can then, using the
4673 man:lttng-add-context(1) command, add this context field by name to
4674 the JUL or log4j <<channel,channel>>.
4676 To provide application-specific context fields in a Java application:
4678 . In the Java application's source code, import the LTTng-UST
4679 Java agent context classes and interfaces:
4684 import org.lttng.ust.agent.context.ContextInfoManager;
4685 import org.lttng.ust.agent.context.IContextInfoRetriever;
4689 . Create a context information retriever class, that is, a class which
4690 implements the `IContextInfoRetriever` interface:
4695 class MyContextInfoRetriever implements IContextInfoRetriever
4698 public Object retrieveContextInfo(String key)
4700 if (key.equals("intCtx")) {
4702 } else if (key.equals("strContext")) {
4703 return "context value!";
4712 This `retrieveContextInfo()` method is the only member of the
4713 `IContextInfoRetriever` interface. Its role is to return the current
4714 value of a state by name to create a context field. The names of the
4715 context fields and which state variables they return depends on your
4718 All primitive types and objects are supported as context fields.
4719 When `retrieveContextInfo()` returns an object, the context field
4720 serializer calls its `toString()` method to add a string field to
4721 event records. The method can also return `null`, which means that
4722 no context field is available for the required name.
4724 . Register an instance of your context information retriever class to
4725 the context information manager singleton:
4730 IContextInfoRetriever cir = new MyContextInfoRetriever();
4731 ContextInfoManager cim = ContextInfoManager.getInstance();
4732 cim.registerContextInfoRetriever("retrieverName", cir);
4736 . Before exiting the application, remove your context information
4737 retriever from the context information manager singleton:
4742 ContextInfoManager cim = ContextInfoManager.getInstance();
4743 cim.unregisterContextInfoRetriever("retrieverName");
4747 This is not strictly necessary, but it is recommended for a clean
4748 disposal of some manager's resources.
4750 . Build your Java application with LTTng-UST Java agent support as
4751 usual, following the procedure for either the <<jul,JUL>> or
4752 <<log4j,Apache log4j>> framework.
4755 .Provide application-specific context fields in a Java application.
4760 import java.util.logging.Handler;
4761 import java.util.logging.Logger;
4762 import org.lttng.ust.agent.jul.LttngLogHandler;
4763 import org.lttng.ust.agent.context.ContextInfoManager;
4764 import org.lttng.ust.agent.context.IContextInfoRetriever;
4768 // Our context information retriever class
4769 private static class MyContextInfoRetriever
4770 implements IContextInfoRetriever
4773 public Object retrieveContextInfo(String key) {
4774 if (key.equals("intCtx")) {
4776 } else if (key.equals("strContext")) {
4777 return "context value!";
4784 private static final int answer = 42;
4786 public static void main(String args[]) throws Exception
4788 // Get the context information manager instance
4789 ContextInfoManager cim = ContextInfoManager.getInstance();
4791 // Create and register our context information retriever
4792 IContextInfoRetriever cir = new MyContextInfoRetriever();
4793 cim.registerContextInfoRetriever("myRetriever", cir);
4796 Logger logger = Logger.getLogger("jello");
4798 // Create an LTTng-UST log handler
4799 Handler lttngUstLogHandler = new LttngLogHandler();
4801 // Add the LTTng-UST log handler to our logger
4802 logger.addHandler(lttngUstLogHandler);
4805 logger.info("some info");
4806 logger.warning("some warning");
4808 logger.finer("finer information; the answer is " + answer);
4810 logger.severe("error!");
4812 // Not mandatory, but cleaner
4813 logger.removeHandler(lttngUstLogHandler);
4814 lttngUstLogHandler.close();
4815 cim.unregisterContextInfoRetriever("myRetriever");
4824 javac -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-jul.jar Test.java
4827 <<creating-destroying-tracing-sessions,Create a tracing session>>
4828 and <<enabling-disabling-events,create an event rule>> matching the
4834 lttng enable-event --jul jello
4837 <<adding-context,Add the application-specific context fields>> to the
4842 lttng add-context --jul --type='$app.myRetriever:intCtx'
4843 lttng add-context --jul --type='$app.myRetriever:strContext'
4846 <<basic-tracing-session-control,Start tracing>>:
4853 Run the compiled class:
4857 java -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-jul.jar:. Test
4860 <<basic-tracing-session-control,Stop tracing>> and inspect the
4872 [[python-application]]
4873 === User space Python agent
4875 You can instrument a Python 2 or Python 3 application which uses the
4876 standard https://docs.python.org/3/library/logging.html[`logging`]
4879 Each log statement emits an LTTng event once the
4880 application module imports the
4881 <<lttng-ust-agents,LTTng-UST Python agent>> package.
4884 .A Python application importing the LTTng-UST Python agent.
4885 image::python-app.png[]
4887 To use the LTTng-UST Python agent:
4889 . In the Python application's source code, import the LTTng-UST Python
4899 The LTTng-UST Python agent automatically adds its logging handler to the
4900 root logger at import time.
4902 Any log statement that the application executes before this import does
4903 not emit an LTTng event.
4905 IMPORTANT: The LTTng-UST Python agent must be
4906 <<installing-lttng,installed>>.
4908 . Use log statements and logging configuration as usual.
4909 Since the LTTng-UST Python agent adds a handler to the _root_
4910 logger, you can trace any log statement from any logger.
4912 .Use the LTTng-UST Python agent.
4923 logging.basicConfig()
4924 logger = logging.getLogger('my-logger')
4927 logger.debug('debug message')
4928 logger.info('info message')
4929 logger.warn('warn message')
4930 logger.error('error message')
4931 logger.critical('critical message')
4935 if __name__ == '__main__':
4939 NOTE: `logging.basicConfig()`, which adds to the root logger a basic
4940 logging handler which prints to the standard error stream, is not
4941 strictly required for LTTng-UST tracing to work, but in versions of
4942 Python preceding 3.2, you could see a warning message which indicates
4943 that no handler exists for the logger `my-logger`.
4945 <<creating-destroying-tracing-sessions,Create a tracing session>>,
4946 <<enabling-disabling-events,create an event rule>> matching the
4947 `my-logger` Python logger, and <<basic-tracing-session-control,start
4953 lttng enable-event --python my-logger
4957 Run the Python script:
4964 <<basic-tracing-session-control,Stop tracing>> and inspect the recorded
4974 You can use the opt:lttng-enable-event(1):--loglevel or
4975 opt:lttng-enable-event(1):--loglevel-only option of the
4976 man:lttng-enable-event(1) command to target a range of Python log levels
4977 or a specific Python log level.
4979 When an application imports the LTTng-UST Python agent, the agent tries
4980 to register to a <<lttng-sessiond,session daemon>>. Note that you must
4981 <<start-sessiond,start the session daemon>> _before_ you run the Python
4982 application. If a session daemon is found, the agent tries to register
4983 to it during 5{nbsp}seconds, after which the application continues
4984 without LTTng tracing support. You can override this timeout value with
4985 the env:LTTNG_UST_PYTHON_REGISTER_TIMEOUT environment variable
4988 If the session daemon stops while a Python application with an imported
4989 LTTng-UST Python agent runs, the agent retries to connect and to
4990 register to a session daemon every 3{nbsp}seconds. You can override this
4991 delay with the env:LTTNG_UST_PYTHON_REGISTER_RETRY_DELAY environment
4996 [[proc-lttng-logger-abi]]
4999 The `lttng-tracer` Linux kernel module, part of
5000 <<lttng-modules,LTTng-modules>>, creates the special LTTng logger file
5001 path:{/proc/lttng-logger} when it's loaded. Any application can write
5002 text data to this file to emit an LTTng event.
5005 .An application writes to the LTTng logger file to emit an LTTng event.
5006 image::lttng-logger.png[]
5008 The LTTng logger is the quickest method--not the most efficient,
5009 however--to add instrumentation to an application. It is designed
5010 mostly to instrument shell scripts:
5014 echo "Some message, some $variable" > /proc/lttng-logger
5017 Any event that the LTTng logger emits is named `lttng_logger` and
5018 belongs to the Linux kernel <<domain,tracing domain>>. However, unlike
5019 other instrumentation points in the kernel tracing domain, **any Unix
5020 user** can <<enabling-disabling-events,create an event rule>> which
5021 matches its event name, not only the root user or users in the tracing
5024 To use the LTTng logger:
5026 * From any application, write text data to the path:{/proc/lttng-logger}
5029 The `msg` field of `lttng_logger` event records contains the
5032 NOTE: The maximum message length of an LTTng logger event is
5033 1024{nbsp}bytes. Writing more than this makes the LTTng logger emit more
5034 than one event to contain the remaining data.
5036 You should not use the LTTng logger to trace a user application which
5037 can be instrumented in a more efficient way, namely:
5039 * <<c-application,C and $$C++$$ applications>>.
5040 * <<java-application,Java applications>>.
5041 * <<python-application,Python applications>>.
5043 .Use the LTTng logger.
5048 echo 'Hello, World!' > /proc/lttng-logger
5050 df --human-readable --print-type / > /proc/lttng-logger
5053 <<creating-destroying-tracing-sessions,Create a tracing session>>,
5054 <<enabling-disabling-events,create an event rule>> matching the
5055 `lttng_logger` Linux kernel tracepoint, and
5056 <<basic-tracing-session-control,start tracing>>:
5061 lttng enable-event --kernel lttng_logger
5065 Run the Bash script:
5072 <<basic-tracing-session-control,Stop tracing>> and inspect the recorded
5083 [[instrumenting-linux-kernel]]
5084 === LTTng kernel tracepoints
5086 NOTE: This section shows how to _add_ instrumentation points to the
5087 Linux kernel. The kernel's subsystems are already thoroughly
5088 instrumented at strategic places for LTTng when you
5089 <<installing-lttng,install>> the <<lttng-modules,LTTng-modules>>
5093 There are two methods to instrument the Linux kernel:
5095 . <<linux-add-lttng-layer,Add an LTTng layer>> over an existing ftrace
5096 tracepoint which uses the `TRACE_EVENT()` API.
5098 Choose this if you want to instrumentation a Linux kernel tree with an
5099 instrumentation point compatible with ftrace, perf, and SystemTap.
5101 . Use an <<linux-lttng-tracepoint-event,LTTng-only approach>> to
5102 instrument an out-of-tree kernel module.
5104 Choose this if you don't need ftrace, perf, or SystemTap support.
5108 [[linux-add-lttng-layer]]
5109 ==== [[instrumenting-linux-kernel-itself]][[mainline-trace-event]][[lttng-adaptation-layer]]Add an LTTng layer to an existing ftrace tracepoint
5111 This section shows how to add an LTTng layer to existing ftrace
5112 instrumentation using the `TRACE_EVENT()` API.
5114 This section does not document the `TRACE_EVENT()` macro. You can
5115 read the following articles to learn more about this API:
5117 * http://lwn.net/Articles/379903/[Using the TRACE_EVENT() macro (Part 1)]
5118 * http://lwn.net/Articles/381064/[Using the TRACE_EVENT() macro (Part 2)]
5119 * http://lwn.net/Articles/383362/[Using the TRACE_EVENT() macro (Part 3)]
5121 The following procedure assumes that your ftrace tracepoints are
5122 correctly defined in their own header and that they are created in
5123 one source file using the `CREATE_TRACE_POINTS` definition.
5125 To add an LTTng layer over an existing ftrace tracepoint:
5127 . Make sure the following kernel configuration options are
5133 * `CONFIG_HIGH_RES_TIMERS`
5134 * `CONFIG_TRACEPOINTS`
5137 . Build the Linux source tree with your custom ftrace tracepoints.
5138 . Boot the resulting Linux image on your target system.
5140 Confirm that the tracepoints exist by looking for their names in the
5141 dir:{/sys/kernel/debug/tracing/events/subsys} directory, where `subsys`
5142 is your subsystem's name.
5144 . Get a copy of the latest LTTng-modules{nbsp}{revision}:
5150 wget http://lttng.org/files/lttng-modules/lttng-modules-latest-2.8.tar.bz2 &&
5151 tar -xf lttng-modules-latest-2.8.tar.bz2 &&
5152 cd lttng-modules-2.8.*
5156 . In dir:{instrumentation/events/lttng-module}, relative to the root
5157 of the LTTng-modules source tree, create a header file named
5158 +__subsys__.h+ for your custom subsystem +__subsys__+ and write your
5159 LTTng-modules tracepoint definitions using the LTTng-modules
5162 Start with this template:
5166 .path:{instrumentation/events/lttng-module/my_subsys.h}
5169 #define TRACE_SYSTEM my_subsys
5171 #if !defined(_LTTNG_MY_SUBSYS_H) || defined(TRACE_HEADER_MULTI_READ)
5172 #define _LTTNG_MY_SUBSYS_H
5174 #include "../../../probes/lttng-tracepoint-event.h"
5175 #include <linux/tracepoint.h>
5177 LTTNG_TRACEPOINT_EVENT(
5179 * Format is identical to TRACE_EVENT()'s version for the three
5180 * following macro parameters:
5183 TP_PROTO(int my_int, const char *my_string),
5184 TP_ARGS(my_int, my_string),
5186 /* LTTng-modules specific macros */
5188 ctf_integer(int, my_int_field, my_int)
5189 ctf_string(my_bar_field, my_bar)
5193 #endif /* !defined(_LTTNG_MY_SUBSYS_H) || defined(TRACE_HEADER_MULTI_READ) */
5195 #include "../../../probes/define_trace.h"
5199 The entries in the `TP_FIELDS()` section are the list of fields for the
5200 LTTng tracepoint. This is similar to the `TP_STRUCT__entry()` part of
5201 ftrace's `TRACE_EVENT()` macro.
5203 See <<lttng-modules-tp-fields,Tracepoint fields macros>> for a
5204 complete description of the available `ctf_*()` macros.
5206 . Create the LTTng-modules probe's kernel module C source file,
5207 +probes/lttng-probe-__subsys__.c+, where +__subsys__+ is your
5212 .path:{probes/lttng-probe-my-subsys.c}
5214 #include <linux/module.h>
5215 #include "../lttng-tracer.h"
5218 * Build-time verification of mismatch between mainline
5219 * TRACE_EVENT() arguments and the LTTng-modules adaptation
5220 * layer LTTNG_TRACEPOINT_EVENT() arguments.
5222 #include <trace/events/my_subsys.h>
5224 /* Create LTTng tracepoint probes */
5225 #define LTTNG_PACKAGE_BUILD
5226 #define CREATE_TRACE_POINTS
5227 #define TRACE_INCLUDE_PATH ../instrumentation/events/lttng-module
5229 #include "../instrumentation/events/lttng-module/my_subsys.h"
5231 MODULE_LICENSE("GPL and additional rights");
5232 MODULE_AUTHOR("Your name <your-email>");
5233 MODULE_DESCRIPTION("LTTng my_subsys probes");
5234 MODULE_VERSION(__stringify(LTTNG_MODULES_MAJOR_VERSION) "."
5235 __stringify(LTTNG_MODULES_MINOR_VERSION) "."
5236 __stringify(LTTNG_MODULES_PATCHLEVEL_VERSION)
5237 LTTNG_MODULES_EXTRAVERSION);
5241 . Edit path:{probes/Makefile} and add your new kernel module object
5242 next to the existing ones:
5246 .path:{probes/Makefile}
5250 obj-m += lttng-probe-module.o
5251 obj-m += lttng-probe-power.o
5253 obj-m += lttng-probe-my-subsys.o
5259 . Build and install the LTTng kernel modules:
5264 make KERNELDIR=/path/to/linux
5265 sudo make modules_install
5269 Replace `/path/to/linux` with the path to the Linux source tree where
5270 you defined and used tracepoints with ftrace's `TRACE_EVENT()` macro.
5272 Note that you can also use the
5273 <<lttng-tracepoint-event-code,`LTTNG_TRACEPOINT_EVENT_CODE()` macro>>
5274 instead of `LTTNG_TRACEPOINT_EVENT()` to use custom local variables and
5275 C code that need to be executed before the event fields are recorded.
5277 The best way to learn how to use the previous LTTng-modules macros is to
5278 inspect the existing LTTng-modules tracepoint definitions in the
5279 dir:{instrumentation/events/lttng-module} header files. Compare them
5280 with the Linux kernel mainline versions in the
5281 dir:{include/trace/events} directory of the Linux source tree.
5285 [[lttng-tracepoint-event-code]]
5286 ===== Use custom C code to access the data for tracepoint fields
5288 Although we recommended to always use the
5289 <<lttng-adaptation-layer,`LTTNG_TRACEPOINT_EVENT()`>> macro to describe
5290 the arguments and fields of an LTTng-modules tracepoint when possible,
5291 sometimes you need a more complex process to access the data that the
5292 tracer records as event record fields. In other words, you need local
5293 variables and multiple C{nbsp}statements instead of simple
5294 argument-based expressions that you pass to the
5295 <<lttng-modules-tp-fields,`ctf_*()` macros of `TP_FIELDS()`>>.
5297 You can use the `LTTNG_TRACEPOINT_EVENT_CODE()` macro instead of
5298 `LTTNG_TRACEPOINT_EVENT()` to declare custom local variables and define
5299 a block of C{nbsp}code to be executed before LTTng records the fields.
5300 The structure of this macro is:
5303 .`LTTNG_TRACEPOINT_EVENT_CODE()` macro syntax.
5305 LTTNG_TRACEPOINT_EVENT_CODE(
5307 * Format identical to the LTTNG_TRACEPOINT_EVENT()
5308 * version for the following three macro parameters:
5311 TP_PROTO(int my_int, const char *my_string),
5312 TP_ARGS(my_int, my_string),
5314 /* Declarations of custom local variables */
5317 unsigned long b = 0;
5318 const char *name = "(undefined)";
5319 struct my_struct *my_struct;
5323 * Custom code which uses both tracepoint arguments
5324 * (in TP_ARGS()) and local variables (in TP_locvar()).
5326 * Local variables are actually members of a structure pointed
5327 * to by the special variable tp_locvar.
5331 tp_locvar->a = my_int + 17;
5332 tp_locvar->my_struct = get_my_struct_at(tp_locvar->a);
5333 tp_locvar->b = my_struct_compute_b(tp_locvar->my_struct);
5334 tp_locvar->name = my_struct_get_name(tp_locvar->my_struct);
5335 put_my_struct(tp_locvar->my_struct);
5344 * Format identical to the LTTNG_TRACEPOINT_EVENT()
5345 * version for this, except that tp_locvar members can be
5346 * used in the argument expression parameters of
5347 * the ctf_*() macros.
5350 ctf_integer(unsigned long, my_struct_b, tp_locvar->b)
5351 ctf_integer(int, my_struct_a, tp_locvar->a)
5352 ctf_string(my_string_field, my_string)
5353 ctf_string(my_struct_name, tp_locvar->name)
5358 IMPORTANT: The C code defined in `TP_code()` must not have any side
5359 effects when executed. In particular, the code must not allocate
5360 memory or get resources without deallocating this memory or putting
5361 those resources afterwards.
5364 [[instrumenting-linux-kernel-tracing]]
5365 ==== Load and unload a custom probe kernel module
5367 You must load a <<lttng-adaptation-layer,created LTTng-modules probe
5368 kernel module>> in the kernel before it can emit LTTng events.
5370 To load the default probe kernel modules and a custom probe kernel
5373 * Use the opt:lttng-sessiond(8):--extra-kmod-probes option to give extra
5374 probe modules to load when starting a root <<lttng-sessiond,session
5378 .Load the `my_subsys`, `usb`, and the default probe modules.
5382 sudo lttng-sessiond --extra-kmod-probes=my_subsys,usb
5387 You only need to pass the subsystem name, not the whole kernel module
5390 To load _only_ a given custom probe kernel module:
5392 * Use the opt:lttng-sessiond(8):--kmod-probes option to give the probe
5393 modules to load when starting a root session daemon:
5396 .Load only the `my_subsys` and `usb` probe modules.
5400 sudo lttng-sessiond --kmod-probes=my_subsys,usb
5405 To confirm that a probe module is loaded:
5412 lsmod | grep lttng_probe_usb
5416 To unload the loaded probe modules:
5418 * Kill the session daemon with `SIGTERM`:
5423 sudo pkill lttng-sessiond
5427 You can also use man:modprobe(8)'s `--remove` option if the session
5428 daemon terminates abnormally.
5431 [[controlling-tracing]]
5434 Once an application or a Linux kernel is
5435 <<instrumenting,instrumented>> for LTTng tracing,
5438 This section is divided in topics on how to use the various
5439 <<plumbing,components of LTTng>>, in particular the <<lttng-cli,cmd:lttng
5440 command-line tool>>, to _control_ the LTTng daemons and tracers.
5442 NOTE: In the following subsections, we refer to an man:lttng(1) command
5443 using its man page name. For example, instead of _Run the `create`
5444 command to..._, we use _Run the man:lttng-create(1) command to..._.
5448 === Start a session daemon
5450 In some situations, you need to run a <<lttng-sessiond,session daemon>>
5451 (man:lttng-sessiond(8)) _before_ you can use the man:lttng(1)
5454 You will see the following error when you run a command while no session
5458 Error: No session daemon is available
5461 The only command that automatically runs a session daemon is
5462 man:lttng-create(1), which you use to
5463 <<creating-destroying-tracing-sessions,create a tracing session>>. While
5464 this is most of the time the first operation that you do, sometimes it's
5465 not. Some examples are:
5467 * <<list-instrumentation-points,List the available instrumentation points>>.
5468 * <<saving-loading-tracing-session,Load a tracing session configuration>>.
5470 [[tracing-group]] Each Unix user must have its own running session
5471 daemon to trace user applications. The session daemon that the root user
5472 starts is the only one allowed to control the LTTng kernel tracer. Users
5473 that are part of the _tracing group_ can control the root session
5474 daemon. The default tracing group name is `tracing`; you can set it to
5475 something else with the opt:lttng-sessiond(8):--group option when you
5476 start the root session daemon.
5478 To start a user session daemon:
5480 * Run man:lttng-sessiond(8):
5485 lttng-sessiond --daemonize
5489 To start the root session daemon:
5491 * Run man:lttng-sessiond(8) as the root user:
5496 sudo lttng-sessiond --daemonize
5500 In both cases, remove the opt:lttng-sessiond(8):--daemonize option to
5501 start the session daemon in foreground.
5503 To stop a session daemon, use man:kill(1) on its process ID (standard
5506 Note that some Linux distributions could manage the LTTng session daemon
5507 as a service. In this case, you should use the service manager to
5508 start, restart, and stop session daemons.
5511 [[creating-destroying-tracing-sessions]]
5512 === Create and destroy a tracing session
5514 Almost all the LTTng control operations happen in the scope of
5515 a <<tracing-session,tracing session>>, which is the dialogue between the
5516 <<lttng-sessiond,session daemon>> and you.
5518 To create a tracing session with a generated name:
5520 * Use the man:lttng-create(1) command:
5529 The created tracing session's name is `auto` followed by the
5532 To create a tracing session with a specific name:
5534 * Use the optional argument of the man:lttng-create(1) command:
5539 lttng create my-session
5543 Replace `my-session` with the specific tracing session name.
5545 LTTng appends the creation date to the created tracing session's name.
5547 LTTng writes the traces of a tracing session in
5548 +$LTTNG_HOME/lttng-trace/__name__+ by default, where +__name__+ is the
5549 name of the tracing session. Note that the env:LTTNG_HOME environment
5550 variable defaults to `$HOME` if not set.
5552 To output LTTng traces to a non-default location:
5554 * Use the opt:lttng-create(1):--output option of the man:lttng-create(1) command:
5559 lttng create --output=/tmp/some-directory my-session
5563 You may create as many tracing sessions as you wish.
5565 To list all the existing tracing sessions for your Unix user:
5567 * Use the man:lttng-list(1) command:
5576 When you create a tracing session, it is set as the _current tracing
5577 session_. The following man:lttng(1) commands operate on the current
5578 tracing session when you don't specify one:
5580 [role="list-3-cols"]
5596 To change the current tracing session:
5598 * Use the man:lttng-set-session(1) command:
5603 lttng set-session new-session
5607 Replace `new-session` by the name of the new current tracing session.
5609 When you are done tracing in a given tracing session, you can destroy
5610 it. This operation frees the resources taken by the tracing session
5611 to destroy; it does not destroy the trace data that LTTng wrote for
5612 this tracing session.
5614 To destroy the current tracing session:
5616 * Use the man:lttng-destroy(1) command:
5626 [[list-instrumentation-points]]
5627 === List the available instrumentation points
5629 The <<lttng-sessiond,session daemon>> can query the running instrumented
5630 user applications and the Linux kernel to get a list of available
5631 instrumentation points. For the Linux kernel <<domain,tracing domain>>,
5632 they are tracepoints and system calls. For the user space tracing
5633 domain, they are tracepoints. For the other tracing domains, they are
5636 To list the available instrumentation points:
5638 * Use the man:lttng-list(1) command with the requested tracing domain's
5642 * opt:lttng-list(1):--kernel: Linux kernel tracepoints (your Unix user
5643 must be a root user, or it must be a member of the
5644 <<tracing-group,tracing group>>).
5645 * opt:lttng-list(1):--kernel with opt:lttng-list(1):--syscall: Linux
5646 kernel system calls (your Unix user must be a root user, or it must be
5647 a member of the tracing group).
5648 * opt:lttng-list(1):--userspace: user space tracepoints.
5649 * opt:lttng-list(1):--jul: `java.util.logging` loggers.
5650 * opt:lttng-list(1):--log4j: Apache log4j loggers.
5651 * opt:lttng-list(1):--python: Python loggers.
5654 .List the available user space tracepoints.
5658 lttng list --userspace
5662 .List the available Linux kernel system call tracepoints.
5666 lttng list --kernel --syscall
5671 [[enabling-disabling-events]]
5672 === Create and enable an event rule
5674 Once you <<creating-destroying-tracing-sessions,create a tracing
5675 session>>, you can create <<event,event rules>> with the
5676 man:lttng-enable-event(1) command.
5678 You specify each condition with a command-line option. The available
5679 condition options are shown in the following table.
5681 [role="growable",cols="asciidoc,asciidoc,default"]
5682 .Condition command-line options for the man:lttng-enable-event(1) command.
5684 |Option |Description |Applicable tracing domains
5690 . +--probe=__ADDR__+
5691 . +--function=__ADDR__+
5694 Instead of using the default _tracepoint_ instrumentation type, use:
5696 . A Linux system call.
5697 . A Linux https://lwn.net/Articles/132196/[KProbe] (symbol or address).
5698 . The entry and return points of a Linux function (symbol or address).
5702 |First positional argument.
5705 Tracepoint or system call name. In the case of a Linux KProbe or
5706 function, this is a custom name given to the event rule. With the
5707 JUL, log4j, and Python domains, this is a logger name.
5709 With a tracepoint, logger, or system call name, the last character
5710 can be `*` to match anything that remains.
5717 . +--loglevel=__LEVEL__+
5718 . +--loglevel-only=__LEVEL__+
5721 . Match only tracepoints or log statements with a logging level at
5722 least as severe as +__LEVEL__+.
5723 . Match only tracepoints or log statements with a logging level
5724 equal to +__LEVEL__+.
5726 See man:lttng-enable-event(1) for the list of available logging level
5729 |User space, JUL, log4j, and Python.
5731 |+--exclude=__EXCLUSIONS__+
5734 When you use a `*` character at the end of the tracepoint or logger
5735 name (first positional argument), exclude the specific names in the
5736 comma-delimited list +__EXCLUSIONS__+.
5739 User space, JUL, log4j, and Python.
5741 |+--filter=__EXPR__+
5744 Match only events which satisfy the expression +__EXPR__+.
5746 See man:lttng-enable-event(1) to learn more about the syntax of a
5753 You attach an event rule to a <<channel,channel>> on creation. If you do
5754 not specify the channel with the opt:lttng-enable-event(1):--channel
5755 option, and if the event rule to create is the first in its
5756 <<domain,tracing domain>> for a given tracing session, then LTTng
5757 creates a _default channel_ for you. This default channel is reused in
5758 subsequent invocations of the man:lttng-enable-event(1) command for the
5759 same tracing domain.
5761 An event rule is always enabled at creation time.
5763 The following examples show how you can combine the previous
5764 command-line options to create simple to more complex event rules.
5766 .Create an event rule targetting a Linux kernel tracepoint (default channel).
5770 lttng enable-event --kernel sched_switch
5774 .Create an event rule matching four Linux kernel system calls (default channel).
5778 lttng enable-event --kernel --syscall open,write,read,close
5782 .Create event rules matching tracepoints with filter expressions (default channel).
5786 lttng enable-event --kernel sched_switch --filter='prev_comm == "bash"'
5791 lttng enable-event --kernel --all \
5792 --filter='$ctx.tid == 1988 || $ctx.tid == 1534'
5797 lttng enable-event --jul my_logger \
5798 --filter='$app.retriever:cur_msg_id > 3'
5801 IMPORTANT: Make sure to always quote the filter string when you
5802 use man:lttng(1) from a shell.
5805 .Create an event rule matching any user space tracepoint of a given tracepoint provider with a log level range (default channel).
5809 lttng enable-event --userspace my_app:'*' --loglevel=TRACE_INFO
5812 IMPORTANT: Make sure to always quote the wildcard character when you
5813 use man:lttng(1) from a shell.
5816 .Create an event rule matching multiple Python loggers with a wildcard and with exclusions (default channel).
5820 lttng enable-event --python my-app.'*' \
5821 --exclude='my-app.module,my-app.hello'
5825 .Create an event rule matching any Apache log4j logger with a specific log level (default channel).
5829 lttng enable-event --log4j --all --loglevel-only=LOG4J_WARN
5833 .Create an event rule attached to a specific channel matching a specific user space tracepoint provider and tracepoint.
5837 lttng enable-event --userspace my_app:my_tracepoint --channel=my-channel
5841 The event rules of a given channel form a whitelist: as soon as an
5842 emitted event passes one of them, LTTng can record the event. For
5843 example, an event named `my_app:my_tracepoint` emitted from a user space
5844 tracepoint with a `TRACE_ERROR` log level passes both of the following
5849 lttng enable-event --userspace my_app:my_tracepoint
5850 lttng enable-event --userspace my_app:my_tracepoint \
5851 --loglevel=TRACE_INFO
5854 The second event rule is redundant: the first one includes
5858 [[disable-event-rule]]
5859 === Disable an event rule
5861 To disable an event rule that you <<enabling-disabling-events,created>>
5862 previously, use the man:lttng-disable-event(1) command. This command
5863 disables _all_ the event rules (of a given tracing domain and channel)
5864 which match an instrumentation point. The other conditions are not
5865 supported as of LTTng{nbsp}{revision}.
5867 The LTTng tracer does not record an emitted event which passes
5868 a _disabled_ event rule.
5870 .Disable an event rule matching a Python logger (default channel).
5874 lttng disable-event --python my-logger
5878 .Disable an event rule matching all `java.util.logging` loggers (default channel).
5882 lttng disable-event --jul '*'
5886 .Disable _all_ the event rules of the default channel.
5888 The opt:lttng-disable-event(1):--all-events option is not, like the
5889 opt:lttng-enable-event(1):--all option of man:lttng-enable-event(1), the
5890 equivalent of the event name `*` (wildcard): it disables _all_ the event
5891 rules of a given channel.
5895 lttng disable-event --jul --all-events
5899 NOTE: You cannot delete an event rule once you create it.
5903 === Get the status of a tracing session
5905 To get the status of the current tracing session, that is, its
5906 parameters, its channels, event rules, and their attributes:
5908 * Use the man:lttng-status(1) command:
5918 To get the status of any tracing session:
5920 * Use the man:lttng-list(1) command with the tracing session's name:
5925 lttng list my-session
5929 Replace `my-session` with the desired tracing session's name.
5932 [[basic-tracing-session-control]]
5933 === Start and stop a tracing session
5935 Once you <<creating-destroying-tracing-sessions,create a tracing
5937 <<enabling-disabling-events,create one or more event rules>>,
5938 you can start and stop the tracers for this tracing session.
5940 To start tracing in the current tracing session:
5942 * Use the man:lttng-start(1) command:
5951 LTTng is very flexible: you can launch user applications before
5952 or after the you start the tracers. The tracers only record the events
5953 if they pass enabled event rules and if they occur while the tracers are
5956 To stop tracing in the current tracing session:
5958 * Use the man:lttng-stop(1) command:
5967 If there were <<channel-overwrite-mode-vs-discard-mode,lost event
5968 records>> or lost sub-buffers since the last time you ran
5969 man:lttng-start(1), warnings are printed when you run the
5970 man:lttng-stop(1) command.
5973 [[enabling-disabling-channels]]
5974 === Create a channel
5976 Once you create a tracing session, you can create a <<channel,channel>>
5977 with the man:lttng-enable-channel(1) command.
5979 Note that LTTng automatically creates a default channel when, for a
5980 given <<domain,tracing domain>>, no channels exist and you
5981 <<enabling-disabling-events,create>> the first event rule. This default
5982 channel is named `channel0` and its attributes are set to reasonable
5983 values. Therefore, you only need to create a channel when you need
5984 non-default attributes.
5986 You specify each non-default channel attribute with a command-line
5987 option when you use the man:lttng-enable-channel(1) command. The
5988 available command-line options are:
5990 [role="growable",cols="asciidoc,asciidoc"]
5991 .Command-line options for the man:lttng-enable-channel(1) command.
5993 |Option |Description
5999 <<channel-overwrite-mode-vs-discard-mode,event loss mode>> instead of
6000 the default _discard_ mode.
6002 |`--buffers-pid` (user space tracing domain only)
6005 Use the per-process <<channel-buffering-schemes,buffering scheme>>
6006 instead of the default per-user buffering scheme.
6008 |+--subbuf-size=__SIZE__+
6011 Allocate sub-buffers of +__SIZE__+ bytes (power of two), for each CPU,
6012 either for each Unix user (default), or for each instrumented process.
6014 See <<channel-subbuf-size-vs-subbuf-count,Sub-buffer count and size>>.
6016 |+--num-subbuf=__COUNT__+
6019 Allocate +__COUNT__+ sub-buffers (power of two), for each CPU, either
6020 for each Unix user (default), or for each instrumented process.
6022 See <<channel-subbuf-size-vs-subbuf-count,Sub-buffer count and size>>.
6024 |+--tracefile-size=__SIZE__+
6027 Set the maximum size of each trace file that this channel writes within
6028 a stream to +__SIZE__+ bytes instead of no maximum.
6030 See <<tracefile-rotation,Trace file count and size>>.
6032 |+--tracefile-count=__COUNT__+
6035 Limit the number of trace files that this channel creates to
6036 +__COUNT__+ channels instead of no limit.
6038 See <<tracefile-rotation,Trace file count and size>>.
6040 |+--switch-timer=__PERIODUS__+
6043 Set the <<channel-switch-timer,switch timer period>>
6044 to +__PERIODUS__+{nbsp}µs.
6046 |+--read-timer=__PERIODUS__+
6049 Set the <<channel-read-timer,read timer period>>
6050 to +__PERIODUS__+{nbsp}µs.
6052 |+--output=__TYPE__+ (Linux kernel tracing domain only)
6055 Set the channel's output type to +__TYPE__+, either `mmap` or `splice`.
6059 You can only create a channel in the Linux kernel and user space
6060 <<domain,tracing domains>>: other tracing domains have their own channel
6061 created on the fly when <<enabling-disabling-events,creating event
6066 Because of a current LTTng limitation, you must create all channels
6067 _before_ you <<basic-tracing-session-control,start tracing>> in a given
6068 tracing session, that is, before the first time you run
6071 Since LTTng automatically creates a default channel when you use the
6072 man:lttng-enable-event(1) command with a specific tracing domain, you
6073 cannot, for example, create a Linux kernel event rule, start tracing,
6074 and then create a user space event rule, because no user space channel
6075 exists yet and it's too late to create one.
6077 For this reason, make sure to configure your channels properly
6078 before starting the tracers for the first time!
6081 The following examples show how you can combine the previous
6082 command-line options to create simple to more complex channels.
6084 .Create a Linux kernel channel with default attributes.
6088 lttng enable-channel --kernel my-channel
6092 .Create a user space channel with 4 sub-buffers or 1{nbsp}MiB each, per CPU, per instrumented process.
6096 lttng enable-channel --userspace --num-subbuf=4 --subbuf-size=1M \
6097 --buffers-pid my-channel
6101 .Create a Linux kernel channel which rotates 8 trace files of 4{nbsp}MiB each for each stream
6105 lttng enable-channel --kernel --tracefile-count=8 \
6106 --tracefile-size=4194304 my-channel
6110 .Create a user space channel in overwrite (or _flight recorder_) mode.
6114 lttng enable-channel --userspace --overwrite my-channel
6118 You can <<enabling-disabling-events,create>> the same event rule in
6119 two different channels:
6123 lttng enable-event --userspace --channel=my-channel app:tp
6124 lttng enable-event --userspace --channel=other-channel app:tp
6127 If both channels are enabled, when a tracepoint named `app:tp` is
6128 reached, LTTng records two events, one for each channel.
6132 === Disable a channel
6134 To disable a specific channel that you <<enabling-disabling-channels,created>>
6135 previously, use the man:lttng-disable-channel(1) command.
6137 .Disable a specific Linux kernel channel.
6141 lttng disable-channel --kernel my-channel
6145 The state of a channel precedes the individual states of event rules
6146 attached to it: event rules which belong to a disabled channel, even if
6147 they are enabled, are also considered disabled.
6151 === Add context fields to a channel
6153 Event record fields in trace files provide important information about
6154 events that occured previously, but sometimes some external context may
6155 help you solve a problem faster. Examples of context fields are:
6157 * The **process ID**, **thread ID**, **process name**, and
6158 **process priority** of the thread in which the event occurs.
6159 * The **hostname** of the system on which the event occurs.
6160 * The current values of many possible **performance counters** using
6162 ** CPU cycles, stalled cycles, idle cycles, and the other cycle types.
6164 ** Branch instructions, misses, and loads.
6166 * Any context defined at the application level (supported for the
6167 JUL and log4j <<domain,tracing domains>>).
6169 To get the full list of available context fields, see
6170 `lttng add-context --list`. Some context fields are reserved for a
6171 specific <<domain,tracing domain>> (Linux kernel or user space).
6173 You add context fields to <<channel,channels>>. All the events
6174 that a channel with added context fields records contain those fields.
6176 To add context fields to one or all the channels of a given tracing
6179 * Use the man:lttng-add-context(1) command.
6181 .Add context fields to all the channels of the current tracing session.
6183 The following command line adds the virtual process identifier and
6184 the per-thread CPU cycles count fields to all the user space channels
6185 of the current tracing session.
6189 lttng add-context --userspace --type=vpid --type=perf:thread:cpu-cycles
6193 .Add a context field to a specific channel.
6195 The following command line adds the thread identifier context field
6196 to the Linux kernel channel named `my-channel` in the current
6201 lttng add-context --kernel --channel=my-channel --type=tid
6205 .Add an application-specific context field to a specific channel.
6207 The following command line adds the `cur_msg_id` context field of the
6208 `retriever` context retriever for all the instrumented
6209 <<java-application,Java applications>> recording <<event,event records>>
6210 in the channel named `my-channel`:
6214 lttng add-context --kernel --channel=my-channel \
6215 --type='$app:retriever:cur_msg_id'
6218 IMPORTANT: Make sure to always quote the `$` character when you
6219 use man:lttng-add-context(1) from a shell.
6222 NOTE: You cannot remove context fields from a channel once you add it.
6227 === Track process IDs
6229 It's often useful to allow only specific process IDs (PIDs) to emit
6230 events. For example, you may wish to record all the system calls made by
6231 a given process (à la http://linux.die.net/man/1/strace[strace]).
6233 The man:lttng-track(1) and man:lttng-untrack(1) commands serve this
6234 purpose. Both commands operate on a whitelist of process IDs. You _add_
6235 entries to this whitelist with the man:lttng-track(1) command and remove
6236 entries with the man:lttng-untrack(1) command. Any process which has one
6237 of the PIDs in the whitelist is allowed to emit LTTng events which pass
6238 an enabled <<event,event rule>>.
6240 NOTE: The PID tracker tracks the _numeric process IDs_. Should a
6241 process with a given tracked ID exit and another process be given this
6242 ID, then the latter would also be allowed to emit events.
6244 .Track and untrack process IDs.
6246 For the sake of the following example, assume the target system has 16
6250 <<creating-destroying-tracing-sessions,create a tracing session>>,
6251 the whitelist contains all the possible PIDs:
6254 .All PIDs are tracked.
6255 image::track-all.png[]
6257 When the whitelist is full and you use the man:lttng-track(1) command to
6258 specify some PIDs to track, LTTng first clears the whitelist, then it
6259 tracks the specific PIDs. After:
6263 lttng track --pid=3,4,7,10,13
6269 .PIDs 3, 4, 7, 10, and 13 are tracked.
6270 image::track-3-4-7-10-13.png[]
6272 You can add more PIDs to the whitelist afterwards:
6276 lttng track --pid=1,15,16
6282 .PIDs 1, 15, and 16 are added to the whitelist.
6283 image::track-1-3-4-7-10-13-15-16.png[]
6285 The man:lttng-untrack(1) command removes entries from the PID tracker's
6286 whitelist. Given the previous example, the following command:
6290 lttng untrack --pid=3,7,10,13
6293 leads to this whitelist:
6296 .PIDs 3, 7, 10, and 13 are removed from the whitelist.
6297 image::track-1-4-15-16.png[]
6299 LTTng can track all possible PIDs again using the opt:track(1):--all
6304 lttng track --pid --all
6307 The result is, again:
6310 .All PIDs are tracked.
6311 image::track-all.png[]
6314 .Track only specific PIDs
6316 A very typical use case with PID tracking is to start with an empty
6317 whitelist, then <<basic-tracing-session-control,start the tracers>>, and
6318 then add PIDs manually while tracers are active. You can accomplish this
6319 by using the opt:lttng-untrack(1):--all option of the
6320 man:lttng-untrack(1) command to clear the whitelist after you
6321 <<creating-destroying-tracing-sessions,create a tracing session>>:
6325 lttng untrack --pid --all
6331 .No PIDs are tracked.
6332 image::untrack-all.png[]
6334 If you trace with this whitelist configuration, the tracer records no
6335 events for this <<domain,tracing domain>> because no processes are
6336 tracked. You can use the man:lttng-track(1) command as usual to track
6337 specific PIDs, for example:
6341 lttng track --pid=6,11
6347 .PIDs 6 and 11 are tracked.
6348 image::track-6-11.png[]
6353 [[saving-loading-tracing-session]]
6354 === Save and load tracing session configurations
6356 Configuring a <<tracing-session,tracing session>> can be long. Some of
6357 the tasks involved are:
6359 * <<enabling-disabling-channels,Create channels>> with
6360 specific attributes.
6361 * <<adding-context,Add context fields>> to specific channels.
6362 * <<enabling-disabling-events,Create event rules>> with specific log
6363 level and filter conditions.
6365 If you use LTTng to solve real world problems, chances are you have to
6366 record events using the same tracing session setup over and over,
6367 modifying a few variables each time in your instrumented program
6368 or environment. To avoid constant tracing session reconfiguration,
6369 the man:lttng(1) command-line tool can save and load tracing session
6370 configurations to/from XML files.
6372 To save a given tracing session configuration:
6374 * Use the man:lttng-save(1) command:
6379 lttng save my-session
6383 Replace `my-session` with the name of the tracing session to save.
6385 LTTng saves tracing session configurations to
6386 dir:{$LTTNG_HOME/.lttng/sessions} by default. Note that the
6387 env:LTTNG_HOME environment variable defaults to `$HOME` if not set. Use
6388 the opt:lttng-save(1):--output-path option to change this destination
6391 LTTng saves all configuration parameters, for example:
6393 * The tracing session name.
6394 * The trace data output path.
6395 * The channels with their state and all their attributes.
6396 * The context fields you added to channels.
6397 * The event rules with their state, log level and filter conditions.
6399 To load a tracing session:
6401 * Use the man:lttng-load(1) command:
6406 lttng load my-session
6410 Replace `my-session` with the name of the tracing session to load.
6412 When LTTng loads a configuration, it restores your saved tracing session
6413 as if you just configured it manually.
6415 See man:lttng(1) for the complete list of command-line options. You
6416 can also save and load all many sessions at a time, and decide in which
6417 directory to output the XML files.
6420 [[sending-trace-data-over-the-network]]
6421 === Send trace data over the network
6423 LTTng can send the recorded trace data to a remote system over the
6424 network instead of writing it to the local file system.
6426 To send the trace data over the network:
6428 . On the _remote_ system (which can also be the target system),
6429 start an LTTng <<lttng-relayd,relay daemon>> (man:lttng-relayd(8)):
6438 . On the _target_ system, create a tracing session configured to
6439 send trace data over the network:
6444 lttng create my-session --set-url=net://remote-system
6448 Replace `remote-system` by the host name or IP address of the
6449 remote system. See man:lttng-create(1) for the exact URL format.
6451 . On the target system, use the man:lttng(1) command-line tool as usual.
6452 When tracing is active, the target's consumer daemon sends sub-buffers
6453 to the relay daemon running on the remote system intead of flushing
6454 them to the local file system. The relay daemon writes the received
6455 packets to the local file system.
6457 The relay daemon writes trace files to
6458 +$LTTNG_HOME/lttng-traces/__hostname__/__session__+ by default, where
6459 +__hostname__+ is the host name of the target system and +__session__+
6460 is the tracing session name. Note that the env:LTTNG_HOME environment
6461 variable defaults to `$HOME` if not set. Use the
6462 opt:lttng-relayd(8):--output option of man:lttng-relayd(8) to write
6463 trace files to another base directory.
6468 === View events as LTTng emits them (noch:{LTTng} live)
6470 LTTng live is a network protocol implemented by the <<lttng-relayd,relay
6471 daemon>> (man:lttng-relayd(8)) to allow compatible trace viewers to
6472 display events as LTTng emits them on the target system while tracing is
6475 The relay daemon creates a _tee_: it forwards the trace data to both
6476 the local file system and to connected live viewers:
6479 .The relay daemon creates a _tee_, forwarding the trace data to both trace files and a connected live viewer.
6484 . On the _target system_, create a <<tracing-session,tracing session>>
6490 lttng create --live my-session
6494 This spawns a local relay daemon.
6496 . Start the live viewer and configure it to connect to the relay
6497 daemon. For example, with http://diamon.org/babeltrace[Babeltrace]:
6502 babeltrace --input-format=lttng-live net://localhost/host/hostname/my-session
6509 * `hostname` with the host name of the target system.
6510 * `my-session` with the name of the tracing session to view.
6513 . Configure the tracing session as usual with the man:lttng(1)
6514 command-line tool, and <<basic-tracing-session-control,start tracing>>.
6516 You can list the available live tracing sessions with Babeltrace:
6520 babeltrace --input-format=lttng-live net://localhost
6523 You can start the relay daemon on another system. In this case, you need
6524 to specify the relay daemon's URL when you create the tracing session
6525 with the opt:lttng-create(1):--set-url option. You also need to replace
6526 `localhost` in the procedure above with the host name of the system on
6527 which the relay daemon is running.
6529 See man:lttng-create(1) and man:lttng-relayd(8) for the complete list of
6530 command-line options.
6534 [[taking-a-snapshot]]
6535 === Take a snapshot of the current sub-buffers of a tracing session
6537 The normal behavior of LTTng is to append full sub-buffers to growing
6538 trace data files. This is ideal to keep a full history of the events
6539 that occurred on the target system, but it can
6540 represent too much data in some situations. For example, you may wish
6541 to trace your application continuously until some critical situation
6542 happens, in which case you only need the latest few recorded
6543 events to perform the desired analysis, not multi-gigabyte trace files.
6545 With the man:lttng-snapshot(1) command, you can take a snapshot of the
6546 current sub-buffers of a given <<tracing-session,tracing session>>.
6547 LTTng can write the snapshot to the local file system or send it over
6552 . Create a tracing session in _snapshot mode_:
6557 lttng create --snapshot my-session
6561 The <<channel-overwrite-mode-vs-discard-mode,event loss mode>> of
6562 <<channel,channels>> created in this mode is automatically set to
6563 _overwrite_ (flight recorder mode).
6565 . Configure the tracing session as usual with the man:lttng(1)
6566 command-line tool, and <<basic-tracing-session-control,start tracing>>.
6568 . **Optional**: When you need to take a snapshot,
6569 <<basic-tracing-session-control,stop tracing>>.
6571 You can take a snapshot when the tracers are active, but if you stop
6572 them first, you are sure that the data in the sub-buffers does not
6573 change before you actually take the snapshot.
6580 lttng snapshot record --name=my-first-snapshot
6584 LTTng writes the current sub-buffers of all the current tracing
6585 session's channels to trace files on the local file system. Those trace
6586 files have `my-first-snapshot` in their name.
6588 There is no difference between the format of a normal trace file and the
6589 format of a snapshot: viewers of LTTng traces also support LTTng
6592 By default, LTTng writes snapshot files to the path shown by
6593 `lttng snapshot list-output`. You can change this path or decide to send
6594 snapshots over the network using either:
6596 . An output path or URL that you specify when you create the
6598 . An snapshot output path or URL that you add using
6599 `lttng snapshot add-output`
6600 . An output path or URL that you provide directly to the
6601 `lttng snapshot record` command.
6603 Method 3 overrides method 2, which overrides method 1. When you
6604 specify a URL, a relay daemon must listen on a remote system (see
6605 <<sending-trace-data-over-the-network,Send trace data over the network>>).
6610 === Use the machine interface
6612 With any command of the man:lttng(1) command-line tool, you can set the
6613 opt:lttng(1):--mi option to `xml` (before the command name) to get an
6614 XML machine interface output, for example:
6618 lttng --mi=xml enable-event --kernel --syscall open
6621 A schema definition (XSD) is
6622 https://github.com/lttng/lttng-tools/blob/stable-2.8/src/common/mi-lttng-3.0.xsd[available]
6623 to ease the integration with external tools as much as possible.
6627 [[metadata-regenerate]]
6628 === Regenerate the metadata of an LTTng trace
6630 An LTTng trace, which is a http://diamon.org/ctf[CTF] trace, has both
6631 data stream files and a metadata file. This metadata file contains,
6632 amongst other things, information about the offset of the clock sources
6633 used to timestamp <<event,event records>> when tracing.
6635 If, once a <<tracing-session,tracing session>> is
6636 <<basic-tracing-session-control,started>>, a major
6637 https://en.wikipedia.org/wiki/Network_Time_Protocol[NTP] correction
6638 happens, the trace's clock offset also needs to be updated. You
6639 can use the man:lttng-metadata(1) command to do so.
6641 The main use case of this command is to allow a system to boot with
6642 an incorrect wall time and trace it with LTTng before its wall time
6643 is corrected. Once the system is known to be in a state where its
6644 wall time is correct, it can run `lttng metadata regenerate`.
6646 To regenerate the metadata of an LTTng trace:
6648 * Use the `regenerate` action of the man:lttng-metadata(1) command:
6653 lttng metadata regenerate
6659 `lttng metadata regenerate` has the following limitations:
6661 * Tracing session <<creating-destroying-tracing-sessions,created>>
6663 * User space <<channel,channels>>, if any, using
6664 <<channel-buffering-schemes,per-user buffering>>.
6669 [[persistent-memory-file-systems]]
6670 === Record trace data on persistent memory file systems
6672 https://en.wikipedia.org/wiki/Non-volatile_random-access_memory[Non-volatile random-access memory]
6673 (NVRAM) is random-access memory that retains its information when power
6674 is turned off (non-volatile). Systems with such memory can store data
6675 structures in RAM and retrieve them after a reboot, without flushing
6676 to typical _storage_.
6678 Linux supports NVRAM file systems thanks to either
6679 http://pramfs.sourceforge.net/[PRAMFS] or
6680 https://www.kernel.org/doc/Documentation/filesystems/dax.txt[DAX]{nbsp}+{nbsp}http://lkml.iu.edu/hypermail/linux/kernel/1504.1/03463.html[pmem]
6681 (requires Linux 4.1+).
6683 This section does not describe how to operate such file systems;
6684 we assume that you have a working persistent memory file system.
6686 When you create a <<tracing-session,tracing session>>, you can specify
6687 the path of the shared memory holding the sub-buffers. If you specify a
6688 location on an NVRAM file system, then you can retrieve the latest
6689 recorded trace data when the system reboots after a crash.
6691 To record trace data on a persistent memory file system and retrieve the
6692 trace data after a system crash:
6694 . Create a tracing session with a sub-buffer shared memory path located
6695 on an NVRAM file system:
6700 lttng create --shm-path=/path/to/shm
6704 . Configure the tracing session as usual with the man:lttng(1)
6705 command-line tool, and <<basic-tracing-session-control,start tracing>>.
6707 . After a system crash, use the man:lttng-crash(1) command-line tool to
6708 view the trace data recorded on the NVRAM file system:
6713 lttng-crash /path/to/shm
6717 The binary layout of the ring buffer files is not exactly the same as
6718 the trace files layout. This is why you need to use man:lttng-crash(1)
6719 instead of your preferred trace viewer directly.
6721 To convert the ring buffer files to LTTng trace files:
6723 * Use the opt:lttng-crash(1):--extract option of man:lttng-crash(1):
6728 lttng-crash --extract=/path/to/trace /path/to/shm
6736 [[lttng-modules-ref]]
6737 === noch:{LTTng-modules}
6740 [[lttng-modules-tp-fields]]
6741 ==== Tracepoint fields macros (for `TP_FIELDS()`)
6743 [[tp-fast-assign]][[tp-struct-entry]]The available macros to define
6744 tracepoint fields, which must be listed within `TP_FIELDS()` in
6745 `LTTNG_TRACEPOINT_EVENT()`, are:
6747 [role="func-desc growable",cols="asciidoc,asciidoc"]
6748 .Available macros to define LTTng-modules tracepoint fields
6750 |Macro |Description and parameters
6753 +ctf_integer(__t__, __n__, __e__)+
6755 +ctf_integer_nowrite(__t__, __n__, __e__)+
6757 +ctf_user_integer(__t__, __n__, __e__)+
6759 +ctf_user_integer_nowrite(__t__, __n__, __e__)+
6761 Standard integer, displayed in base 10.
6764 Integer C type (`int`, `long`, `size_t`, ...).
6770 Argument expression.
6773 +ctf_integer_hex(__t__, __n__, __e__)+
6775 +ctf_user_integer_hex(__t__, __n__, __e__)+
6777 Standard integer, displayed in base 16.
6786 Argument expression.
6788 |+ctf_integer_oct(__t__, __n__, __e__)+
6790 Standard integer, displayed in base 8.
6799 Argument expression.
6802 +ctf_integer_network(__t__, __n__, __e__)+
6804 +ctf_user_integer_network(__t__, __n__, __e__)+
6806 Integer in network byte order (big-endian), displayed in base 10.
6815 Argument expression.
6818 +ctf_integer_network_hex(__t__, __n__, __e__)+
6820 +ctf_user_integer_network_hex(__t__, __n__, __e__)+
6822 Integer in network byte order, displayed in base 16.
6831 Argument expression.
6834 +ctf_string(__n__, __e__)+
6836 +ctf_string_nowrite(__n__, __e__)+
6838 +ctf_user_string(__n__, __e__)+
6840 +ctf_user_string_nowrite(__n__, __e__)+
6842 Null-terminated string; undefined behavior if +__e__+ is `NULL`.
6848 Argument expression.
6851 +ctf_array(__t__, __n__, __e__, __s__)+
6853 +ctf_array_nowrite(__t__, __n__, __e__, __s__)+
6855 +ctf_user_array(__t__, __n__, __e__, __s__)+
6857 +ctf_user_array_nowrite(__t__, __n__, __e__, __s__)+
6859 Statically-sized array of integers.
6862 Array element C type.
6868 Argument expression.
6874 +ctf_array_bitfield(__t__, __n__, __e__, __s__)+
6876 +ctf_array_bitfield_nowrite(__t__, __n__, __e__, __s__)+
6878 +ctf_user_array_bitfield(__t__, __n__, __e__, __s__)+
6880 +ctf_user_array_bitfield_nowrite(__t__, __n__, __e__, __s__)+
6882 Statically-sized array of bits.
6884 The type of +__e__+ must be an integer type. +__s__+ is the number
6885 of elements of such type in +__e__+, not the number of bits.
6888 Array element C type.
6894 Argument expression.
6900 +ctf_array_text(__t__, __n__, __e__, __s__)+
6902 +ctf_array_text_nowrite(__t__, __n__, __e__, __s__)+
6904 +ctf_user_array_text(__t__, __n__, __e__, __s__)+
6906 +ctf_user_array_text_nowrite(__t__, __n__, __e__, __s__)+
6908 Statically-sized array, printed as text.
6910 The string does not need to be null-terminated.
6913 Array element C type (always `char`).
6919 Argument expression.
6925 +ctf_sequence(__t__, __n__, __e__, __T__, __E__)+
6927 +ctf_sequence_nowrite(__t__, __n__, __e__, __T__, __E__)+
6929 +ctf_user_sequence(__t__, __n__, __e__, __T__, __E__)+
6931 +ctf_user_sequence_nowrite(__t__, __n__, __e__, __T__, __E__)+
6933 Dynamically-sized array of integers.
6935 The type of +__E__+ must be unsigned.
6938 Array element C type.
6944 Argument expression.
6947 Length expression C type.
6953 +ctf_sequence_hex(__t__, __n__, __e__, __T__, __E__)+
6955 +ctf_user_sequence_hex(__t__, __n__, __e__, __T__, __E__)+
6957 Dynamically-sized array of integers, displayed in base 16.
6959 The type of +__E__+ must be unsigned.
6962 Array element C type.
6968 Argument expression.
6971 Length expression C type.
6976 |+ctf_sequence_network(__t__, __n__, __e__, __T__, __E__)+
6978 Dynamically-sized array of integers in network byte order (big-endian),
6979 displayed in base 10.
6981 The type of +__E__+ must be unsigned.
6984 Array element C type.
6990 Argument expression.
6993 Length expression C type.
6999 +ctf_sequence_bitfield(__t__, __n__, __e__, __T__, __E__)+
7001 +ctf_sequence_bitfield_nowrite(__t__, __n__, __e__, __T__, __E__)+
7003 +ctf_user_sequence_bitfield(__t__, __n__, __e__, __T__, __E__)+
7005 +ctf_user_sequence_bitfield_nowrite(__t__, __n__, __e__, __T__, __E__)+
7007 Dynamically-sized array of bits.
7009 The type of +__e__+ must be an integer type. +__s__+ is the number
7010 of elements of such type in +__e__+, not the number of bits.
7012 The type of +__E__+ must be unsigned.
7015 Array element C type.
7021 Argument expression.
7024 Length expression C type.
7030 +ctf_sequence_text(__t__, __n__, __e__, __T__, __E__)+
7032 +ctf_sequence_text_nowrite(__t__, __n__, __e__, __T__, __E__)+
7034 +ctf_user_sequence_text(__t__, __n__, __e__, __T__, __E__)+
7036 +ctf_user_sequence_text_nowrite(__t__, __n__, __e__, __T__, __E__)+
7038 Dynamically-sized array, displayed as text.
7040 The string does not need to be null-terminated.
7042 The type of +__E__+ must be unsigned.
7044 The behaviour is undefined if +__e__+ is `NULL`.
7047 Sequence element C type (always `char`).
7053 Argument expression.
7056 Length expression C type.
7062 Use the `_user` versions when the argument expression, `e`, is
7063 a user space address. In the cases of `ctf_user_integer*()` and
7064 `ctf_user_float*()`, `&e` must be a user space address, thus `e` must
7067 The `_nowrite` versions omit themselves from the session trace, but are
7068 otherwise identical. This means the `_nowrite` fields won't be written
7069 in the recorded trace. Their primary purpose is to make some
7070 of the event context available to the
7071 <<enabling-disabling-events,event filters>> without having to
7072 commit the data to sub-buffers.
7078 Terms related to LTTng and to tracing in general:
7081 The http://diamon.org/babeltrace[Babeltrace] project, which includes
7082 the cmd:babeltrace command, some libraries, and Python bindings.
7084 <<channel-buffering-schemes,buffering scheme>>::
7085 A layout of sub-buffers applied to a given channel.
7087 <<channel,channel>>::
7088 An entity which is responsible for a set of ring buffers.
7090 <<event,Event rules>> are always attached to a specific channel.
7093 A reference of time for a tracer.
7095 <<lttng-consumerd,consumer daemon>>::
7096 A process which is responsible for consuming the full sub-buffers
7097 and write them to a file system or send them over the network.
7099 <<channel-overwrite-mode-vs-discard-mode,discard mode>>:: The event loss
7100 mode in which the tracer _discards_ new event records when there's no
7101 sub-buffer space left to store them.
7104 The consequence of the execution of an instrumentation
7105 point, like a tracepoint that you manually place in some source code,
7106 or a Linux kernel KProbe.
7108 An event is said to _occur_ at a specific time. Different actions can
7109 be taken upon the occurance of an event, like record the event's payload
7112 <<channel-overwrite-mode-vs-discard-mode,event loss mode>>::
7113 The mechanism by which event records of a given channel are lost
7114 (not recorded) when there is no sub-buffer space left to store them.
7116 [[def-event-name]]event name::
7117 The name of an event, which is also the name of the event record.
7118 This is also called the _instrumentation point name_.
7121 A record, in a trace, of the payload of an event which occured.
7123 <<event,event rule>>::
7124 Set of conditions which must be satisfied for one or more occuring
7125 events to be recorded.
7127 `java.util.logging`::
7129 https://docs.oracle.com/javase/7/docs/api/java/util/logging/package-summary.html[core logging facilities].
7131 <<instrumenting,instrumentation>>::
7132 The use of LTTng probes to make a piece of software traceable.
7134 instrumentation point::
7135 A point in the execution path of a piece of software that, when
7136 reached by this execution, can emit an event.
7138 instrumentation point name::
7139 See _<<def-event-name,event name>>_.
7142 A http://logging.apache.org/log4j/1.2/[logging library] for Java
7143 developed by the Apache Software Foundation.
7146 Level of severity of a log statement or user space
7147 instrumentation point.
7150 The _Linux Trace Toolkit: next generation_ project.
7152 <<lttng-cli,cmd:lttng>>::
7153 A command-line tool provided by the LTTng-tools project which you
7154 can use to send and receive control messages to and from a
7158 The https://github.com/lttng/lttng-analyses[LTTng analyses] project,
7159 which is a set of analyzing programs that are used to obtain a
7160 higher level view of an LTTng trace.
7162 cmd:lttng-consumerd::
7163 The name of the consumer daemon program.
7166 A utility provided by the LTTng-tools project which can convert
7167 ring buffer files (usually
7168 <<persistent-memory-file-systems,saved on a persistent memory file system>>)
7171 LTTng Documentation::
7174 <<lttng-live,LTTng live>>::
7175 A communication protocol between the relay daemon and live viewers
7176 which makes it possible to see events "live", as they are received by
7179 <<lttng-modules,LTTng-modules>>::
7180 The https://github.com/lttng/lttng-modules[LTTng-modules] project,
7181 which contains the Linux kernel modules to make the Linux kernel
7182 instrumentation points available for LTTng tracing.
7185 The name of the relay daemon program.
7187 cmd:lttng-sessiond::
7188 The name of the session daemon program.
7191 The https://github.com/lttng/lttng-tools[LTTng-tools] project, which
7192 contains the various programs and libraries used to
7193 <<controlling-tracing,control tracing>>.
7195 <<lttng-ust,LTTng-UST>>::
7196 The https://github.com/lttng/lttng-ust[LTTng-UST] project, which
7197 contains libraries to instrument user applications.
7199 <<lttng-ust-agents,LTTng-UST Java agent>>::
7200 A Java package provided by the LTTng-UST project to allow the
7201 LTTng instrumentation of `java.util.logging` and Apache log4j 1.2
7204 <<lttng-ust-agents,LTTng-UST Python agent>>::
7205 A Python package provided by the LTTng-UST project to allow the
7206 LTTng instrumentation of Python logging statements.
7208 <<channel-overwrite-mode-vs-discard-mode,overwrite mode>>::
7209 The event loss mode in which new event records overwrite older
7210 event records when there's no sub-buffer space left to store them.
7212 <<channel-buffering-schemes,per-process buffering>>::
7213 A buffering scheme in which each instrumented process has its own
7214 sub-buffers for a given user space channel.
7216 <<channel-buffering-schemes,per-user buffering>>::
7217 A buffering scheme in which all the processes of a Unix user share the
7218 same sub-buffer for a given user space channel.
7220 <<lttng-relayd,relay daemon>>::
7221 A process which is responsible for receiving the trace data sent by
7222 a distant consumer daemon.
7225 A set of sub-buffers.
7227 <<lttng-sessiond,session daemon>>::
7228 A process which receives control commands from you and orchestrates
7229 the tracers and various LTTng daemons.
7231 <<taking-a-snapshot,snapshot>>::
7232 A copy of the current data of all the sub-buffers of a given tracing
7233 session, saved as trace files.
7236 One part of an LTTng ring buffer which contains event records.
7239 The time information attached to an event when it is emitted.
7242 A set of files which are the concatenations of one or more
7243 flushed sub-buffers.
7246 The action of recording the events emitted by an application
7247 or by a system, or to initiate such recording by controlling
7251 The http://tracecompass.org[Trace Compass] project and application.
7254 An instrumentation point using the tracepoint mechanism of the Linux
7255 kernel or of LTTng-UST.
7257 tracepoint definition::
7258 The definition of a single tracepoint.
7261 The name of a tracepoint.
7263 tracepoint provider::
7264 A set of functions providing tracepoints to an instrumented user
7267 Not to be confused with a _tracepoint provider package_: many tracepoint
7268 providers can exist within a tracepoint provider package.
7270 tracepoint provider package::
7271 One or more tracepoint providers compiled as an object file or as
7275 A software which records emitted events.
7277 <<domain,tracing domain>>::
7278 A namespace for event sources.
7281 The Unix group in which a Unix user can be to be allowed to trace the
7284 <<tracing-session,tracing session>>::
7285 A stateful dialogue between you and a <<lttng-sessiond,session
7289 An application running in user space, as opposed to a Linux kernel
7290 module, for example.