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 regenerate the metadata
116 file of an LTTng trace at any moment. This command is meant to be
117 used to resample the wall time following a major
118 https://en.wikipedia.org/wiki/Network_Time_Protocol[NTP] correction
119 so that a system which boots with an incorrect wall time can be
120 traced before its wall time is NTP-corrected.
122 See man:lttng-metadata(1).
124 ** New command-line interface warnings when <<event,event records>> or
125 whole sub-buffers are
126 <<channel-overwrite-mode-vs-discard-mode,lost>>. The warning messages
127 are printed when a <<tracing-session,tracing session>> is
128 <<basic-tracing-session-control,stopped>> (man:lttng-stop(1)
131 * **User space tracing**:
132 ** Shared object base address dump in order to map <<event,event
133 records>> to original source location (file and line number).
135 If you attach the `ip` and `vpid` <<adding-context,context fields>> to a
136 user space <<channel,channel>> and if you use the
137 <<liblttng-ust-dl,path:{liblttng-ust-dl.so} helper>>, you can retrieve
138 the source location where a given event record was generated.
140 The http://diamon.org/babeltrace/[Babeltrace] trace viewer supports this
141 state dump and those context fields since version 1.4 to print the
142 source location of a given event record. http://tracecompass.org/[Trace
143 Compass] also supports this since version 2.0.
145 ** A <<java-application,Java application>> which uses
146 `java.util.logging` now adds an LTTng-UST log handler to the desired
149 The previous workflow was to initialize the LTTng-UST Java agent
150 by calling `LTTngAgent.getLTTngAgent()`. This had the effect of adding
151 an LTTng-UST log handler to the root loggers.
153 ** A <<java-application,Java application>> which uses Apache log4j now
154 adds an LTTng-UST log appender to the desired log4j loggers.
156 The previous workflow was to initialize the LTTng-UST Java agent
157 by calling `LTTngAgent.getLTTngAgent()`. This had the effect of adding
158 an LTTng-UST appender to the root loggers.
160 ** Any <<java-application,Java application>> can provide
161 <<java-application-context,dynamic context fields>> while running
162 thanks to a new API provided by the <<lttng-ust-agents,LTTng-UST Java
163 agent>>. You can require LTTng to record specific context fields in
164 event records, and you can use them in the filter expression of
165 <<event,event rules>>.
167 * **Linux kernel tracing**:
168 ** The LTTng kernel modules can now be built into a Linux kernel image,
169 that is, not as loadable modules.
172 https://github.com/lttng/lttng-modules/blob/stable-{revision}/README.md#kernel-built-in-support[`README.md`]
175 ** New instrumentation:
176 *** ARM64 architecture support.
178 *** x86 `irq_vectors`.
179 ** New <<adding-context,context fields>>:
182 *** `need_reschedule`
183 *** `migratable` (specific to RT-Preempt)
184 ** Clock source plugin support for advanced cases where a custom source
185 of time is needed to timestamp LTTng event records.
187 See https://github.com/lttng/lttng-modules/blob/stable-{revision}/lttng-clock.h[`lttng-clock.h`]
188 for an overview of the small API.
191 ** The link:/man[man pages] of the man:lttng(1) command-line tool are
192 split into one man page per command (à la Git), for example:
197 man lttng-enable-event
201 You can also use the `--help` option of any man:lttng(1) command to
204 The content and formatting of all the LTTng man pages has improved
211 What is LTTng? As its name suggests, the _Linux Trace Toolkit: next
212 generation_ is a modern toolkit for tracing Linux systems and
213 applications. So your first question might be:
220 As the history of software engineering progressed and led to what
221 we now take for granted--complex, numerous and
222 interdependent software applications running in parallel on
223 sophisticated operating systems like Linux--the authors of such
224 components, software developers, began feeling a natural
225 urge to have tools that would ensure the robustness and good performance
226 of their masterpieces.
228 One major achievement in this field is, inarguably, the
229 https://www.gnu.org/software/gdb/[GNU debugger (GDB)],
230 an essential tool for developers to find and fix bugs. But even the best
231 debugger won't help make your software run faster, and nowadays, faster
232 software means either more work done by the same hardware, or cheaper
233 hardware for the same work.
235 A _profiler_ is often the tool of choice to identify performance
236 bottlenecks. Profiling is suitable to identify _where_ performance is
237 lost in a given software. The profiler outputs a profile, a statistical
238 summary of observed events, which you may use to discover which
239 functions took the most time to execute. However, a profiler won't
240 report _why_ some identified functions are the bottleneck. Bottlenecks
241 might only occur when specific conditions are met, conditions that are
242 sometimes impossible to capture by a statistical profiler, or impossible
243 to reproduce with an application altered by the overhead of an
244 event-based profiler. For a thorough investigation of software
245 performance issues, a history of execution is essential, with the
246 recorded values of variables and context fields you choose, and
247 with as little influence as possible on the instrumented software. This
248 is where tracing comes in handy.
250 _Tracing_ is a technique used to understand what goes on in a running
251 software system. The software used for tracing is called a _tracer_,
252 which is conceptually similar to a tape recorder. When recording,
253 specific instrumentation points placed in the software source code
254 generate events that are saved on a giant tape: a _trace_ file. You
255 can trace user applications and the operating system at the same time,
256 opening the possibility of resolving a wide range of problems that would
257 otherwise be extremely challenging.
259 Tracing is often compared to _logging_. However, tracers and loggers are
260 two different tools, serving two different purposes. Tracers are
261 designed to record much lower-level events that occur much more
262 frequently than log messages, often in the range of thousands per
263 second, with very little execution overhead. Logging is more appropriate
264 for a very high-level analysis of less frequent events: user accesses,
265 exceptional conditions (errors and warnings, for example), database
266 transactions, instant messaging communications, and such. Simply put,
267 logging is one of the many use cases that can be satisfied with tracing.
269 The list of recorded events inside a trace file can be read manually
270 like a log file for the maximum level of detail, but it is generally
271 much more interesting to perform application-specific analyses to
272 produce reduced statistics and graphs that are useful to resolve a
273 given problem. Trace viewers and analyzers are specialized tools
276 In the end, this is what LTTng is: a powerful, open source set of
277 tools to trace the Linux kernel and user applications at the same time.
278 LTTng is composed of several components actively maintained and
279 developed by its link:/community/#where[community].
282 [[lttng-alternatives]]
283 === Alternatives to noch:{LTTng}
285 Excluding proprietary solutions, a few competing software tracers
288 * https://github.com/dtrace4linux/linux[dtrace4linux] is a port of
289 Sun Microsystems's DTrace to Linux. The cmd:dtrace tool interprets
290 user scripts and is responsible for loading code into the
291 Linux kernel for further execution and collecting the outputted data.
292 * https://en.wikipedia.org/wiki/Berkeley_Packet_Filter[eBPF] is a
293 subsystem in the Linux kernel in which a virtual machine can execute
294 programs passed from the user space to the kernel. You can attach
295 such programs to tracepoints and KProbes thanks to a system call, and
296 they can output data to the user space when executed thanks to
297 different mechanisms (pipe, VM register values, and eBPF maps, to name
299 * https://www.kernel.org/doc/Documentation/trace/ftrace.txt[ftrace]
300 is the de facto function tracer of the Linux kernel. Its user
301 interface is a set of special files in sysfs.
302 * https://perf.wiki.kernel.org/[perf] is
303 a performance analyzing tool for Linux which supports hardware
304 performance counters, tracepoints, as well as other counters and
305 types of probes. perf's controlling utility is the cmd:perf command
307 * http://linux.die.net/man/1/strace[strace]
308 is a command-line utility which records system calls made by a
309 user process, as well as signal deliveries and changes of process
310 state. strace makes use of https://en.wikipedia.org/wiki/Ptrace[ptrace]
311 to fulfill its function.
312 * http://www.sysdig.org/[sysdig], like SystemTap, uses scripts to
313 analyze Linux kernel events. You write scripts, or _chisels_ in
314 sysdig's jargon, in Lua and sysdig executes them while the system is
315 being traced or afterwards. sysdig's interface is the cmd:sysdig
316 command-line tool as well as the curses-based cmd:csysdig tool.
317 * https://sourceware.org/systemtap/[SystemTap] is a Linux kernel and
318 user space tracer which uses custom user scripts to produce plain text
319 traces. SystemTap converts the scripts to the C language, and then
320 compiles them as Linux kernel modules which are loaded to produce
321 trace data. SystemTap's primary user interface is the cmd:stap
324 The main distinctive features of LTTng is that it produces correlated
325 kernel and user space traces, as well as doing so with the lowest
326 overhead amongst other solutions. It produces trace files in the
327 http://diamon.org/ctf[CTF] format, a file format optimized
328 for the production and analyses of multi-gigabyte data.
330 LTTng is the result of more than 10 years of active open source
331 development by a community of passionate developers.
332 LTTng{nbsp}{revision} is currently available on major desktop and server
335 The main interface for tracing control is a single command-line tool
336 named cmd:lttng. The latter can create several tracing sessions, enable
337 and disable events on the fly, filter events efficiently with custom
338 user expressions, start and stop tracing, and much more. LTTng can
339 record the traces on the file system or send them over the network, and
340 keep them totally or partially. You can view the traces once tracing
341 becomes inactive or in real-time.
343 <<installing-lttng,Install LTTng now>> and
344 <<getting-started,start tracing>>!
350 **LTTng** is a set of software <<plumbing,components>> which interact to
351 <<instrumenting,instrument>> the Linux kernel and user applications, and
352 to <<controlling-tracing,control tracing>> (start and stop
353 tracing, enable and disable event rules, and the rest). Those
354 components are bundled into the following packages:
356 * **LTTng-tools**: Libraries and command-line interface to
358 * **LTTng-modules**: Linux kernel modules to instrument and
360 * **LTTng-UST**: Libraries and Java/Python packages to instrument and
361 trace user applications.
363 Most distributions mark the LTTng-modules and LTTng-UST packages as
364 optional when installing LTTng-tools (which is always required). In the
365 following sections, we always provide the steps to install all three,
368 * You only need to install LTTng-modules if you intend to trace the
370 * You only need to install LTTng-UST if you intend to trace user
374 .Availability of LTTng{nbsp}{revision} for major Linux distributions.
376 |Distribution |Available in releases |Alternatives
379 |<<ubuntu,Ubuntu{nbsp}16.10 _Yakkety Yak_>>
380 |LTTng{nbsp}{revision} for Ubuntu{nbsp}14.04 _Trusty Tahr_
381 and Ubuntu{nbsp}16.04 _Xenial Xerus_:
382 <<ubuntu-ppa,use the LTTng Stable{nbsp}{revision} PPA>>.
384 <<building-from-source,Build LTTng{nbsp}{revision} from source>> for
385 other Ubuntu releases.
389 |LTTng{nbsp}{revision} for Fedora{nbsp}25 and Fedora{nbsp}26 (not
392 <<building-from-source,Build LTTng{nbsp}{revision} from source>> for
393 other Fedora releases.
396 |<<debian,Debian "stretch" (testing)>>
398 <<building-from-source,Build LTTng{nbsp}{revision} from source>> for
399 previous Debian releases.
403 |LTTng{nbsp}2.7 for openSUSE Leap{nbsp}42.1.
405 <<building-from-source,Build LTTng{nbsp}{revision} from source>> for
406 other openSUSE releases.
409 |Latest AUR packages.
413 |<<alpine-linux,Alpine Linux "edge">>
414 |LTTng{nbsp}{revision} for Alpine Linux{nbsp}3.5 (not released yet).
416 <<building-from-source,Build LTTng{nbsp}{revision} from source>> for
417 other Alpine Linux releases.
420 |See http://packages.efficios.com/[EfficiOS Enterprise Packages].
425 |LTTng{nbsp}{revision} for Buildroot{nbsp}2016.11 (not released yet).
427 LTTng{nbsp}2.7 for Buildroot{nbsp}2016.02, Buildroot{nbsp}2016.05,
428 and Buildroot{nbsp}2016.08.
430 <<building-from-source,Build LTTng{nbsp}{revision} from source>> for
431 other Buildroot releases.
433 |OpenEmbedded and Yocto
434 |<<oe-yocto,`openembedded-core` layer since 3{nbsp}September 2016>>
435 |LTTng{nbsp}2.7 for OpenEmbedded from 1{nbsp}December 2016 until
436 3{nbsp}September 2016.
438 <<building-from-source,Build LTTng{nbsp}{revision} from source>> for
439 other OpenEmbedded releases.
444 === [[ubuntu-official-repositories]]Ubuntu
446 LTTng{nbsp}{revision} is available on Ubuntu{nbsp}16.10 _Yakkety Yak_.
447 For previous releases of Ubuntu, <<ubuntu-ppa,use the LTTng
448 Stable{nbsp}{revision} PPA>>.
450 To install LTTng{nbsp}{revision} on Ubuntu 16.10{nbsp}_Yakkety Yak_:
452 . Install the main LTTng{nbsp}{revision} packages:
457 sudo apt-get install lttng-tools
458 sudo apt-get install lttng-modules-dkms
459 sudo apt-get install liblttng-ust-dev
463 . **If you need to instrument and trace
464 <<java-application,Java applications>>**, install the LTTng-UST
470 sudo apt-get install liblttng-ust-agent-java
474 . **If you need to instrument and trace
475 <<python-application,Python applications>>**, install the
476 LTTng-UST Python agent:
481 sudo apt-get install python3-lttngust
487 ==== noch:{LTTng} Stable {revision} PPA
489 The https://launchpad.net/~lttng/+archive/ubuntu/stable-{revision}[LTTng
490 Stable{nbsp}{revision} PPA] offers the latest stable
491 LTTng{nbsp}{revision} packages for:
493 * Ubuntu{nbsp}14.04 _Trusty Tahr_
494 * Ubuntu{nbsp}16.04 _Xenial Xerus_
496 To install LTTng{nbsp}{revision} from the LTTng Stable{nbsp}{revision} PPA:
498 . Add the LTTng Stable{nbsp}{revision} PPA repository and update the
504 sudo apt-add-repository ppa:lttng/stable-2.8
509 . Install the main LTTng{nbsp}{revision} packages:
514 sudo apt-get install lttng-tools
515 sudo apt-get install lttng-modules-dkms
516 sudo apt-get install liblttng-ust-dev
520 . **If you need to instrument and trace
521 <<java-application,Java applications>>**, install the LTTng-UST
527 sudo apt-get install liblttng-ust-agent-java
531 . **If you need to instrument and trace
532 <<python-application,Python applications>>**, install the
533 LTTng-UST Python agent:
538 sudo apt-get install python3-lttngust
546 To install LTTng{nbsp}{revision} on Debian "stretch" (testing):
548 . Install the main LTTng{nbsp}{revision} packages:
553 sudo apt-get install lttng-modules-dkms
554 sudo apt-get install liblttng-ust-dev
555 sudo apt-get install lttng-tools
559 . **If you need to instrument and trace <<java-application,Java
560 applications>>**, install the LTTng-UST Java agent:
565 sudo apt-get install liblttng-ust-agent-java
569 . **If you need to instrument and trace <<python-application,Python
570 applications>>**, install the LTTng-UST Python agent:
575 sudo apt-get install python3-lttngust
583 To install LTTng{nbsp}{revision} (tracing control and user space
584 tracing) on Alpine Linux "edge":
586 . Make sure your system is
587 https://wiki.alpinelinux.org/wiki/Edge[configured for "edge"].
588 . Enable the _testing_ repository by uncommenting the corresponding
589 line in path:{/etc/apk/repositories}.
590 . Add the LTTng packages:
595 sudo apk add lttng-tools
596 sudo apk add lttng-ust-dev
600 To install LTTng-modules{nbsp}{revision} (Linux kernel tracing support)
601 on Alpine Linux "edge":
603 . Add the vanilla Linux kernel:
608 apk add linux-vanilla linux-vanilla-dev
612 . Reboot with the vanilla Linux kernel.
613 . Download, build, and install the latest LTTng-modules{nbsp}{revision}:
619 wget http://lttng.org/files/lttng-modules/lttng-modules-latest-2.8.tar.bz2 &&
620 tar -xf lttng-modules-latest-2.8.tar.bz2 &&
621 cd lttng-modules-2.8.* &&
623 sudo make modules_install &&
630 === OpenEmbedded and Yocto
632 LTTng{nbsp}{revision} recipes are available in the
633 http://layers.openembedded.org/layerindex/branch/master/layer/openembedded-core/[`openembedded-core`]
634 layer of OpenEmbedded since 3{nbsp}September 2016
635 under the following names:
641 With BitBake, the simplest way to include LTTng recipes in your target
642 image is to add them to `IMAGE_INSTALL_append` in path:{conf/local.conf}:
645 IMAGE_INSTALL_append = " lttng-tools lttng-modules lttng-ust"
650 . Select a machine and an image recipe.
651 . Click **Edit image recipe**.
652 . Under the **All recipes** tab, search for **lttng**.
653 . Check the desired LTTng recipes.
656 .Java and Python application instrumentation and tracing
658 If you need to instrument and trace <<java-application,Java
659 applications>> on openSUSE, you need to build and install
660 LTTng-UST{nbsp}{revision} <<building-from-source,from source>> and pass
661 the `--enable-java-agent-jul`, `--enable-java-agent-log4j`, or
662 `--enable-java-agent-all` options to the `configure` script, depending
663 on which Java logging framework you use.
665 If you need to instrument and trace <<python-application,Python
666 applications>> on openSUSE, you need to build and install
667 LTTng-UST{nbsp}{revision} from source and pass the
668 `--enable-python-agent` option to the `configure` script.
672 [[enterprise-distributions]]
673 === RHEL, SUSE, and other enterprise distributions
675 To install LTTng on enterprise Linux distributions, such as Red Hat
676 Enterprise Linux (RHEL) and SUSE Linux Enterprise Server (SUSE), please
677 see http://packages.efficios.com/[EfficiOS Enterprise Packages].
680 [[building-from-source]]
681 === Build from source
683 To build and install LTTng{nbsp}{revision} from source:
685 . Using your distribution's package manager, or from source, install
686 the following dependencies of LTTng-tools and LTTng-UST:
689 * https://sourceforge.net/projects/libuuid/[libuuid]
690 * http://directory.fsf.org/wiki/Popt[popt]
691 * http://liburcu.org/[Userspace RCU]
692 * http://www.xmlsoft.org/[libxml2]
695 . Download, build, and install the latest LTTng-modules{nbsp}{revision}:
701 wget http://lttng.org/files/lttng-modules/lttng-modules-latest-2.8.tar.bz2 &&
702 tar -xf lttng-modules-latest-2.8.tar.bz2 &&
703 cd lttng-modules-2.8.* &&
705 sudo make modules_install &&
710 . Download, build, and install the latest LTTng-UST{nbsp}{revision}:
716 wget http://lttng.org/files/lttng-ust/lttng-ust-latest-2.8.tar.bz2 &&
717 tar -xf lttng-ust-latest-2.8.tar.bz2 &&
718 cd lttng-ust-2.8.* &&
728 .Java and Python application tracing
730 If you need to instrument and trace <<java-application,Java
731 applications>>, pass the `--enable-java-agent-jul`,
732 `--enable-java-agent-log4j`, or `--enable-java-agent-all` options to the
733 `configure` script, depending on which Java logging framework you use.
735 If you need to instrument and trace <<python-application,Python
736 applications>>, pass the `--enable-python-agent` option to the
737 `configure` script. You can set the `PYTHON` environment variable to the
738 path to the Python interpreter for which to install the LTTng-UST Python
746 By default, LTTng-UST libraries are installed to
747 dir:{/usr/local/lib}, which is the de facto directory in which to
748 keep self-compiled and third-party libraries.
750 When <<building-tracepoint-providers-and-user-application,linking an
751 instrumented user application with `liblttng-ust`>>:
753 * Append `/usr/local/lib` to the env:LD_LIBRARY_PATH environment
755 * Pass the `-L/usr/local/lib` and `-Wl,-rpath,/usr/local/lib` options to
756 man:gcc(1), man:g++(1), or man:clang(1).
760 . Download, build, and install the latest LTTng-tools{nbsp}{revision}:
766 wget http://lttng.org/files/lttng-tools/lttng-tools-latest-2.8.tar.bz2 &&
767 tar -xf lttng-tools-latest-2.8.tar.bz2 &&
768 cd lttng-tools-2.8.* &&
776 TIP: The https://github.com/eepp/vlttng[vlttng tool] can do all the
777 previous steps automatically for a given version of LTTng and confine
778 the installed files in a specific directory. This can be useful to test
779 LTTng without installing it on your system.
785 This is a short guide to get started quickly with LTTng kernel and user
788 Before you follow this guide, make sure to <<installing-lttng,install>>
791 This tutorial walks you through the steps to:
793 . <<tracing-the-linux-kernel,Trace the Linux kernel>>.
794 . <<tracing-your-own-user-application,Trace a user application>> written
796 . <<viewing-and-analyzing-your-traces,View and analyze the
800 [[tracing-the-linux-kernel]]
801 === Trace the Linux kernel
803 The following command lines start with cmd:sudo because you need root
804 privileges to trace the Linux kernel. You can avoid using cmd:sudo if
805 your Unix user is a member of the <<lttng-sessiond,tracing group>>.
807 . Create a <<tracing-session,tracing session>>:
812 sudo lttng create my-kernel-session
816 . List the available kernel tracepoints and system calls:
825 . Create an <<event,event rule>> which matches the desired event names,
826 for example `sched_switch` and `sched_process_fork`:
831 sudo lttng enable-event --kernel sched_switch,sched_process_fork
835 You can also create an event rule which _matches_ all the Linux kernel
836 tracepoints (this will generate a lot of data when tracing):
841 sudo lttng enable-event --kernel --all
854 . Do some operation on your system for a few seconds. For example,
855 load a website, or list the files of a directory.
856 . Stop tracing and destroy the tracing session:
866 The man:lttng-destroy(1) command does not destroy the trace data; it
867 only destroys the state of the tracing session.
869 By default, LTTng saves the traces in
870 +$LTTNG_HOME/lttng-traces/__name__-__date__-__time__+,
871 where +__name__+ is the tracing session name. Note that the
872 env:LTTNG_HOME environment variable defaults to `$HOME` if not set.
874 See <<viewing-and-analyzing-your-traces,View and analyze the
875 recorded events>> to view the recorded events.
878 [[tracing-your-own-user-application]]
879 === Trace a user application
881 This section steps you through a simple example to trace a
882 _Hello world_ program written in C.
884 To create the traceable user application:
886 . Create the tracepoint provider header file, which defines the
887 tracepoints and the events they can generate:
893 #undef TRACEPOINT_PROVIDER
894 #define TRACEPOINT_PROVIDER hello_world
896 #undef TRACEPOINT_INCLUDE
897 #define TRACEPOINT_INCLUDE "./hello-tp.h"
899 #if !defined(_HELLO_TP_H) || defined(TRACEPOINT_HEADER_MULTI_READ)
902 #include <lttng/tracepoint.h>
912 ctf_string(my_string_field, my_string_arg)
913 ctf_integer(int, my_integer_field, my_integer_arg)
917 #endif /* _HELLO_TP_H */
919 #include <lttng/tracepoint-event.h>
923 . Create the tracepoint provider package source file:
929 #define TRACEPOINT_CREATE_PROBES
930 #define TRACEPOINT_DEFINE
932 #include "hello-tp.h"
936 . Build the tracepoint provider package:
941 gcc -c -I. hello-tp.c
945 . Create the _Hello World_ application source file:
952 #include "hello-tp.h"
954 int main(int argc, char *argv[])
958 puts("Hello, World!\nPress Enter to continue...");
961 * The following getchar() call is only placed here for the purpose
962 * of this demonstration, to pause the application in order for
963 * you to have time to list its tracepoints. It is not
969 * A tracepoint() call.
971 * Arguments, as defined in hello-tp.h:
973 * 1. Tracepoint provider name (required)
974 * 2. Tracepoint name (required)
975 * 3. my_integer_arg (first user-defined argument)
976 * 4. my_string_arg (second user-defined argument)
978 * Notice the tracepoint provider and tracepoint names are
979 * NOT strings: they are in fact parts of variables that the
980 * macros in hello-tp.h create.
982 tracepoint(hello_world, my_first_tracepoint, 23, "hi there!");
984 for (x = 0; x < argc; ++x) {
985 tracepoint(hello_world, my_first_tracepoint, x, argv[x]);
988 puts("Quitting now!");
989 tracepoint(hello_world, my_first_tracepoint, x * x, "x^2");
996 . Build the application:
1005 . Link the application with the tracepoint provider package,
1006 `liblttng-ust`, and `libdl`:
1011 gcc -o hello hello.o hello-tp.o -llttng-ust -ldl
1015 Here's the whole build process:
1018 .User space tracing tutorial's build steps.
1019 image::ust-flow.png[]
1021 To trace the user application:
1023 . Run the application with a few arguments:
1028 ./hello world and beyond
1037 Press Enter to continue...
1041 . Start an LTTng <<lttng-sessiond,session daemon>>:
1046 lttng-sessiond --daemonize
1050 Note that a session daemon might already be running, for example as
1051 a service that the distribution's service manager started.
1053 . List the available user space tracepoints:
1058 lttng list --userspace
1062 You see the `hello_world:my_first_tracepoint` tracepoint listed
1063 under the `./hello` process.
1065 . Create a <<tracing-session,tracing session>>:
1070 lttng create my-user-space-session
1074 . Create an <<event,event rule>> which matches the
1075 `hello_world:my_first_tracepoint` event name:
1080 lttng enable-event --userspace hello_world:my_first_tracepoint
1093 . Go back to the running `hello` application and press Enter. The
1094 program executes all `tracepoint()` instrumentation points and exits.
1095 . Stop tracing and destroy the tracing session:
1105 The man:lttng-destroy(1) command does not destroy the trace data; it
1106 only destroys the state of the tracing session.
1108 By default, LTTng saves the traces in
1109 +$LTTNG_HOME/lttng-traces/__name__-__date__-__time__+,
1110 where +__name__+ is the tracing session name. Note that the
1111 env:LTTNG_HOME environment variable defaults to `$HOME` if not set.
1113 See <<viewing-and-analyzing-your-traces,View and analyze the
1114 recorded events>> to view the recorded events.
1117 [[viewing-and-analyzing-your-traces]]
1118 === View and analyze the recorded events
1120 Once you have completed the <<tracing-the-linux-kernel,Trace the Linux
1121 kernel>> and <<tracing-your-own-user-application,Trace a user
1122 application>> tutorials, you can inspect the recorded events.
1124 Many tools are available to read LTTng traces:
1126 * **cmd:babeltrace** is a command-line utility which converts trace
1127 formats; it supports the format that LTTng produces, CTF, as well as a
1128 basic text output which can be ++grep++ed. The cmd:babeltrace command
1129 is part of the http://diamon.org/babeltrace[Babeltrace] project.
1130 * Babeltrace also includes
1131 **https://www.python.org/[Python] bindings** so
1132 that you can easily open and read an LTTng trace with your own script,
1133 benefiting from the power of Python.
1134 * http://tracecompass.org/[**Trace Compass**]
1135 is a graphical user interface for viewing and analyzing any type of
1136 logs or traces, including LTTng's.
1137 * https://github.com/lttng/lttng-analyses[**LTTng analyses**] is a
1138 project which includes many high-level analyses of LTTng kernel
1139 traces, like scheduling statistics, interrupt frequency distribution,
1140 top CPU usage, and more.
1142 NOTE: This section assumes that the traces recorded during the previous
1143 tutorials were saved to their default location, in the
1144 dir:{$LTTNG_HOME/lttng-traces} directory. Note that the env:LTTNG_HOME
1145 environment variable defaults to `$HOME` if not set.
1148 [[viewing-and-analyzing-your-traces-bt]]
1149 ==== Use the cmd:babeltrace command-line tool
1151 The simplest way to list all the recorded events of a trace is to pass
1152 its path to cmd:babeltrace with no options:
1156 babeltrace ~/lttng-traces/my-user-space-session*
1159 cmd:babeltrace finds all traces recursively within the given path and
1160 prints all their events, merging them in chronological order.
1162 You can pipe the output of cmd:babeltrace into a tool like man:grep(1) for
1167 babeltrace ~/lttng-traces/my-kernel-session* | grep sys_
1170 You can pipe the output of cmd:babeltrace into a tool like man:wc(1) to
1171 count the recorded events:
1175 babeltrace ~/lttng-traces/my-kernel-session* | grep sys_read | wc --lines
1179 [[viewing-and-analyzing-your-traces-bt-python]]
1180 ==== Use the Babeltrace Python bindings
1182 The <<viewing-and-analyzing-your-traces-bt,text output of cmd:babeltrace>>
1183 is useful to isolate events by simple matching using man:grep(1) and
1184 similar utilities. However, more elaborate filters, such as keeping only
1185 event records with a field value falling within a specific range, are
1186 not trivial to write using a shell. Moreover, reductions and even the
1187 most basic computations involving multiple event records are virtually
1188 impossible to implement.
1190 Fortunately, Babeltrace ships with Python 3 bindings which makes it easy
1191 to read the event records of an LTTng trace sequentially and compute the
1192 desired information.
1194 The following script accepts an LTTng Linux kernel trace path as its
1195 first argument and prints the short names of the top 5 running processes
1196 on CPU 0 during the whole trace:
1201 from collections import Counter
1207 if len(sys.argv) != 2:
1208 msg = 'Usage: python3 {} TRACEPATH'.format(sys.argv[0])
1209 print(msg, file=sys.stderr)
1212 # A trace collection contains one or more traces
1213 col = babeltrace.TraceCollection()
1215 # Add the trace provided by the user (LTTng traces always have
1217 if col.add_trace(sys.argv[1], 'ctf') is None:
1218 raise RuntimeError('Cannot add trace')
1220 # This counter dict contains execution times:
1222 # task command name -> total execution time (ns)
1223 exec_times = Counter()
1225 # This contains the last `sched_switch` timestamp
1229 for event in col.events:
1230 # Keep only `sched_switch` events
1231 if event.name != 'sched_switch':
1234 # Keep only events which happened on CPU 0
1235 if event['cpu_id'] != 0:
1239 cur_ts = event.timestamp
1245 # Previous task command (short) name
1246 prev_comm = event['prev_comm']
1248 # Initialize entry in our dict if not yet done
1249 if prev_comm not in exec_times:
1250 exec_times[prev_comm] = 0
1252 # Compute previous command execution time
1253 diff = cur_ts - last_ts
1255 # Update execution time of this command
1256 exec_times[prev_comm] += diff
1258 # Update last timestamp
1262 for name, ns in exec_times.most_common(5):
1264 print('{:20}{} s'.format(name, s))
1269 if __name__ == '__main__':
1270 sys.exit(0 if top5proc() else 1)
1277 python3 top5proc.py ~/lttng-traces/my-kernel-session*/kernel
1283 swapper/0 48.607245889 s
1284 chromium 7.192738188 s
1285 pavucontrol 0.709894415 s
1286 Compositor 0.660867933 s
1287 Xorg.bin 0.616753786 s
1290 Note that `swapper/0` is the "idle" process of CPU 0 on Linux; since we
1291 weren't using the CPU that much when tracing, its first position in the
1296 == [[understanding-lttng]]Core concepts
1298 From a user's perspective, the LTTng system is built on a few concepts,
1299 or objects, on which the <<lttng-cli,cmd:lttng command-line tool>>
1300 operates by sending commands to the <<lttng-sessiond,session daemon>>.
1301 Understanding how those objects relate to eachother is key in mastering
1304 The core concepts are:
1306 * <<tracing-session,Tracing session>>
1307 * <<domain,Tracing domain>>
1308 * <<channel,Channel and ring buffer>>
1309 * <<"event","Instrumentation point, event rule, event, and event record">>
1315 A _tracing session_ is a stateful dialogue between you and
1316 a <<lttng-sessiond,session daemon>>. You can
1317 <<creating-destroying-tracing-sessions,create a new tracing
1318 session>> with the `lttng create` command.
1320 Anything that you do when you control LTTng tracers happens within a
1321 tracing session. In particular, a tracing session:
1324 * Has its own set of trace files.
1325 * Has its own state of activity (started or stopped).
1326 * Has its own <<tracing-session-mode,mode>> (local, network streaming,
1328 * Has its own <<channel,channels>> which have their own
1329 <<event,event rules>>.
1332 .A _tracing session_ contains <<channel,channels>> that are members of <<domain,tracing domains>> and contain <<event,event rules>>.
1333 image::concepts.png[]
1335 Those attributes and objects are completely isolated between different
1338 A tracing session is analogous to a cash machine session:
1339 the operations you do on the banking system through the cash machine do
1340 not alter the data of other users of the same system. In the case of
1341 the cash machine, a session lasts as long as your bank card is inside.
1342 In the case of LTTng, a tracing session lasts from the `lttng create`
1343 command to the `lttng destroy` command.
1346 .Each Unix user has its own set of tracing sessions.
1347 image::many-sessions.png[]
1350 [[tracing-session-mode]]
1351 ==== Tracing session mode
1353 LTTng can send the generated trace data to different locations. The
1354 _tracing session mode_ dictates where to send it. The following modes
1355 are available in LTTng{nbsp}{revision}:
1358 LTTng writes the traces to the file system of the machine being traced
1361 Network streaming mode::
1362 LTTng sends the traces over the network to a
1363 <<lttng-relayd,relay daemon>> running on a remote system.
1366 LTTng does not write the traces by default. Instead, you can request
1367 LTTng to <<taking-a-snapshot,take a snapshot>>, that is, a copy of the
1368 current tracing buffers, and to write it to the target's file system
1369 or to send it over the network to a <<lttng-relayd,relay daemon>>
1370 running on a remote system.
1373 This mode is similar to the network streaming mode, but a live
1374 trace viewer can connect to the distant relay daemon to
1375 <<lttng-live,view event records as LTTng generates them>> by
1382 A _tracing domain_ is a namespace for event sources. A tracing domain
1383 has its own properties and features.
1385 There are currently five available tracing domains:
1389 * `java.util.logging` (JUL)
1393 You must specify a tracing domain when using some commands to avoid
1394 ambiguity. For example, since all the domains support named tracepoints
1395 as event sources (instrumentation points that you manually insert in the
1396 source code), you need to specify a tracing domain when
1397 <<enabling-disabling-events,creating an event rule>> because all the
1398 tracing domains could have tracepoints with the same names.
1400 Some features are reserved to specific tracing domains. Dynamic function
1401 entry and return instrumentation points, for example, are currently only
1402 supported in the Linux kernel tracing domain, but support for other
1403 tracing domains could be added in the future.
1405 You can create <<channel,channels>> in the Linux kernel and user space
1406 tracing domains. The other tracing domains have a single default
1411 === Channel and ring buffer
1413 A _channel_ is an object which is responsible for a set of ring buffers.
1414 Each ring buffer is divided into multiple sub-buffers. When an LTTng
1415 tracer emits an event, it can record it to one or more
1416 sub-buffers. The attributes of a channel determine what to do when
1417 there's no space left for a new event record because all sub-buffers
1418 are full, where to send a full sub-buffer, and other behaviours.
1420 A channel is always associated to a <<domain,tracing domain>>. The
1421 `java.util.logging` (JUL), log4j, and Python tracing domains each have
1422 a default channel which you cannot configure.
1424 A channel also owns <<event,event rules>>. When an LTTng tracer emits
1425 an event, it records it to the sub-buffers of all
1426 the enabled channels with a satisfied event rule, as long as those
1427 channels are part of active <<tracing-session,tracing sessions>>.
1430 [[channel-buffering-schemes]]
1431 ==== Per-user vs. per-process buffering schemes
1433 A channel has at least one ring buffer _per CPU_. LTTng always
1434 records an event to the ring buffer associated to the CPU on which it
1437 Two _buffering schemes_ are available when you
1438 <<enabling-disabling-channels,create a channel>> in the
1439 user space <<domain,tracing domain>>:
1441 Per-user buffering::
1442 Allocate one set of ring buffers--one per CPU--shared by all the
1443 instrumented processes of each Unix user.
1447 .Per-user buffering scheme.
1448 image::per-user-buffering.png[]
1451 Per-process buffering::
1452 Allocate one set of ring buffers--one per CPU--for each
1453 instrumented process.
1457 .Per-process buffering scheme.
1458 image::per-process-buffering.png[]
1461 The per-process buffering scheme tends to consume more memory than the
1462 per-user option because systems generally have more instrumented
1463 processes than Unix users running instrumented processes. However, the
1464 per-process buffering scheme ensures that one process having a high
1465 event throughput won't fill all the shared sub-buffers of the same
1468 The Linux kernel tracing domain has only one available buffering scheme
1469 which is to allocate a single set of ring buffers for the whole system.
1470 This scheme is similar to the per-user option, but with a single, global
1471 user "running" the kernel.
1474 [[channel-overwrite-mode-vs-discard-mode]]
1475 ==== Overwrite vs. discard event loss modes
1477 When an event occurs, LTTng records it to a specific sub-buffer (yellow
1478 arc in the following animation) of a specific channel's ring buffer.
1479 When there's no space left in a sub-buffer, the tracer marks it as
1480 consumable (red) and another, empty sub-buffer starts receiving the
1481 following event records. A <<lttng-consumerd,consumer daemon>>
1482 eventually consumes the marked sub-buffer (returns to white).
1485 [role="docsvg-channel-subbuf-anim"]
1490 In an ideal world, sub-buffers are consumed faster than they are filled,
1491 as is the case in the previous animation. In the real world,
1492 however, all sub-buffers can be full at some point, leaving no space to
1493 record the following events.
1495 By design, LTTng is a _non-blocking_ tracer: when no empty sub-buffer is
1496 available, it is acceptable to lose event records when the alternative
1497 would be to cause substantial delays in the instrumented application's
1498 execution. LTTng privileges performance over integrity; it aims at
1499 perturbing the traced system as little as possible in order to make
1500 tracing of subtle race conditions and rare interrupt cascades possible.
1502 When it comes to losing event records because no empty sub-buffer is
1503 available, the channel's _event loss mode_ determines what to do. The
1504 available event loss modes are:
1507 Drop the newest event records until a the tracer
1508 releases a sub-buffer.
1511 Clear the sub-buffer containing the oldest event records and start
1512 writing the newest event records there.
1514 This mode is sometimes called _flight recorder mode_ because it's
1516 https://en.wikipedia.org/wiki/Flight_recorder[flight recorder]:
1517 always keep a fixed amount of the latest data.
1519 Which mechanism you should choose depends on your context: prioritize
1520 the newest or the oldest event records in the ring buffer?
1522 Beware that, in overwrite mode, the tracer abandons a whole sub-buffer
1523 as soon as a there's no space left for a new event record, whereas in
1524 discard mode, the tracer only discards the event record that doesn't
1527 In discard mode, LTTng increments a count of lost event records when
1528 an event record is lost and saves this count to the trace. In
1529 overwrite mode, LTTng keeps no information when it overwrites a
1530 sub-buffer before consuming it.
1532 There are a few ways to decrease your probability of losing event
1534 <<channel-subbuf-size-vs-subbuf-count,Sub-buffer count and size>> shows
1535 how you can fine-une the sub-buffer count and size of a channel to
1536 virtually stop losing event records, though at the cost of greater
1540 [[channel-subbuf-size-vs-subbuf-count]]
1541 ==== Sub-buffer count and size
1543 When you <<enabling-disabling-channels,create a channel>>, you can
1544 set its number of sub-buffers and their size.
1546 Note that there is noticeable CPU overhead introduced when
1547 switching sub-buffers (marking a full one as consumable and switching
1548 to an empty one for the following events to be recorded). Knowing this,
1549 the following list presents a few practical situations along with how
1550 to configure the sub-buffer count and size for them:
1552 * **High event throughput**: In general, prefer bigger sub-buffers to
1553 lower the risk of losing event records.
1555 Having bigger sub-buffers also ensures a lower sub-buffer switching
1558 The number of sub-buffers is only meaningful if you create the channel
1559 in overwrite mode: in this case, if a sub-buffer overwrite happens, the
1560 other sub-buffers are left unaltered.
1562 * **Low event throughput**: In general, prefer smaller sub-buffers
1563 since the risk of losing event records is low.
1565 Because events occur less frequently, the sub-buffer switching frequency
1566 should remain low and thus the tracer's overhead should not be a
1569 * **Low memory system**: If your target system has a low memory
1570 limit, prefer fewer first, then smaller sub-buffers.
1572 Even if the system is limited in memory, you want to keep the
1573 sub-buffers as big as possible to avoid a high sub-buffer switching
1576 Note that LTTng uses http://diamon.org/ctf/[CTF] as its trace format,
1577 which means event data is very compact. For example, the average
1578 LTTng kernel event record weights about 32{nbsp}bytes. Thus, a
1579 sub-buffer size of 1{nbsp}MiB is considered big.
1581 The previous situations highlight the major trade-off between a few big
1582 sub-buffers and more, smaller sub-buffers: sub-buffer switching
1583 frequency vs. how much data is lost in overwrite mode. Assuming a
1584 constant event throughput and using the overwrite mode, the two
1585 following configurations have the same ring buffer total size:
1588 [role="docsvg-channel-subbuf-size-vs-count-anim"]
1593 * **2 sub-buffers of 4{nbsp}MiB each**: Expect a very low sub-buffer
1594 switching frequency, but if a sub-buffer overwrite happens, half of
1595 the event records so far (4{nbsp}MiB) are definitely lost.
1596 * **8 sub-buffers of 1{nbsp}MiB each**: Expect 4{nbsp}times the tracer's
1597 overhead as the previous configuration, but if a sub-buffer
1598 overwrite happens, only the eighth of event records so far are
1601 In discard mode, the sub-buffers count parameter is pointless: use two
1602 sub-buffers and set their size according to the requirements of your
1606 [[channel-switch-timer]]
1607 ==== Switch timer period
1609 The _switch timer period_ is an important configurable attribute of
1610 a channel to ensure periodic sub-buffer flushing.
1612 When the _switch timer_ expires, a sub-buffer switch happens. You can
1613 set the switch timer period attribute when you
1614 <<enabling-disabling-channels,create a channel>> to ensure that event
1615 data is consumed and committed to trace files or to a distant relay
1616 daemon periodically in case of a low event throughput.
1619 [role="docsvg-channel-switch-timer"]
1624 This attribute is also convenient when you use big sub-buffers to cope
1625 with a sporadic high event throughput, even if the throughput is
1629 [[channel-read-timer]]
1630 ==== Read timer period
1632 By default, the LTTng tracers use a notification mechanism to signal a
1633 full sub-buffer so that a consumer daemon can consume it. When such
1634 notifications must be avoided, for example in real-time applications,
1635 you can use the channel's _read timer_ instead. When the read timer
1636 fires, the <<lttng-consumerd,consumer daemon>> checks for full,
1637 consumable sub-buffers.
1640 [[tracefile-rotation]]
1641 ==== Trace file count and size
1643 By default, trace files can grow as large as needed. You can set the
1644 maximum size of each trace file that a channel writes when you
1645 <<enabling-disabling-channels,create a channel>>. When the size of
1646 a trace file reaches the channel's fixed maximum size, LTTng creates
1647 another file to contain the next event records. LTTng appends a file
1648 count to each trace file name in this case.
1650 If you set the trace file size attribute when you create a channel, the
1651 maximum number of trace files that LTTng creates is _unlimited_ by
1652 default. To limit them, you can also set a maximum number of trace
1653 files. When the number of trace files reaches the channel's fixed
1654 maximum count, the oldest trace file is overwritten. This mechanism is
1655 called _trace file rotation_.
1659 === Instrumentation point, event rule, event, and event record
1661 An _event rule_ is a set of conditions which must be **all** satisfied
1662 for LTTng to record an occuring event.
1664 You set the conditions when you <<enabling-disabling-events,create
1667 You always attach an event rule to <<channel,channel>> when you create
1670 When an event passes the conditions of an event rule, LTTng records it
1671 in one of the attached channel's sub-buffers.
1673 The available conditions, as of LTTng{nbsp}{revision}, are:
1675 * The event rule _is enabled_.
1676 * The instrumentation point's type _is{nbsp}T_.
1677 * The instrumentation point's name (sometimes called _event name_)
1678 _matches{nbsp}N_, but _is not{nbsp}E_.
1679 * The instrumentation point's log level _is as severe as{nbsp}L_, or
1680 _is exactly{nbsp}L_.
1681 * The fields of the event's payload _satisfy_ a filter
1682 expression{nbsp}__F__.
1684 As you can see, all the conditions but the dynamic filter are related to
1685 the event rule's status or to the instrumentation point, not to the
1686 occurring events. This is why, without a filter, checking if an event
1687 passes an event rule is not a dynamic task: when you create or modify an
1688 event rule, all the tracers of its tracing domain enable or disable the
1689 instrumentation points themselves once. This is possible because the
1690 attributes of an instrumentation point (type, name, and log level) are
1691 defined statically. In other words, without a dynamic filter, the tracer
1692 _does not evaluate_ the arguments of an instrumentation point unless it
1693 matches an enabled event rule.
1695 Note that, for LTTng to record an event, the <<channel,channel>> to
1696 which a matching event rule is attached must also be enabled, and the
1697 tracing session owning this channel must be active.
1700 .Logical path from an instrumentation point to an event record.
1701 image::event-rule.png[]
1703 .Event, event record, or event rule?
1705 With so many similar terms, it's easy to get confused.
1707 An **event** is the consequence of the execution of an _instrumentation
1708 point_, like a tracepoint that you manually place in some source code,
1709 or a Linux kernel KProbe. An event is said to _occur_ at a specific
1710 time. Different actions can be taken upon the occurance of an event,
1711 like record the event's payload to a buffer.
1713 An **event record** is the representation of an event in a sub-buffer. A
1714 tracer is responsible for capturing the payload of an event, current
1715 context variables, the event's ID, and the event's timestamp. LTTng
1716 can append this sub-buffer to a trace file.
1718 An **event rule** is a set of conditions which must all be satisfied for
1719 LTTng to record an occuring event. Events still occur without
1720 satisfying event rules, but LTTng does not record them.
1725 == Components of noch:{LTTng}
1727 The second _T_ in _LTTng_ stands for _toolkit_: it would be wrong
1728 to call LTTng a simple _tool_ since it is composed of multiple
1729 interacting components. This section describes those components,
1730 explains their respective roles, and shows how they connect together to
1731 form the LTTng ecosystem.
1733 The following diagram shows how the most important components of LTTng
1734 interact with user applications, the Linux kernel, and you:
1737 .Control and trace data paths between LTTng components.
1738 image::plumbing.png[]
1740 The LTTng project incorporates:
1742 * **LTTng-tools**: Libraries and command-line interface to
1743 control tracing sessions.
1744 ** <<lttng-sessiond,Session daemon>> (man:lttng-sessiond(8)).
1745 ** <<lttng-consumerd,Consumer daemon>> (man:lttng-consumerd(8)).
1746 ** <<lttng-relayd,Relay daemon>> (man:lttng-relayd(8)).
1747 ** <<liblttng-ctl-lttng,Tracing control library>> (`liblttng-ctl`).
1748 ** <<lttng-cli,Tracing control command-line tool>> (man:lttng(1)).
1749 * **LTTng-UST**: Libraries and Java/Python packages to trace user
1751 ** <<lttng-ust,User space tracing library>> (`liblttng-ust`) and its
1752 headers to instrument and trace any native user application.
1753 ** <<prebuilt-ust-helpers,Preloadable user space tracing helpers>>:
1754 *** `liblttng-ust-libc-wrapper`
1755 *** `liblttng-ust-pthread-wrapper`
1756 *** `liblttng-ust-cyg-profile`
1757 *** `liblttng-ust-cyg-profile-fast`
1758 *** `liblttng-ust-dl`
1759 ** User space tracepoint provider source files generator command-line
1760 tool (man:lttng-gen-tp(1)).
1761 ** <<lttng-ust-agents,LTTng-UST Java agent>> to instrument and trace
1762 Java applications using `java.util.logging` or
1763 Apache log4j 1.2 logging.
1764 ** <<lttng-ust-agents,LTTng-UST Python agent>> to instrument
1765 Python applications using the standard `logging` package.
1766 * **LTTng-modules**: <<lttng-modules,Linux kernel modules>> to trace
1768 ** LTTng kernel tracer module.
1769 ** Tracing ring buffer kernel modules.
1770 ** Probe kernel modules.
1771 ** LTTng logger kernel module.
1775 === Tracing control command-line interface
1778 .The tracing control command-line interface.
1779 image::plumbing-lttng-cli.png[]
1781 The _man:lttng(1) command-line tool_ is the standard user interface to
1782 control LTTng <<tracing-session,tracing sessions>>. The cmd:lttng tool
1783 is part of LTTng-tools.
1785 The cmd:lttng tool is linked with
1786 <<liblttng-ctl-lttng,`liblttng-ctl`>> to communicate with
1787 one or more <<lttng-sessiond,session daemons>> behind the scenes.
1789 The cmd:lttng tool has a Git-like interface:
1793 lttng <general options> <command> <command options>
1796 The <<controlling-tracing,Tracing control>> section explores the
1797 available features of LTTng using the cmd:lttng tool.
1800 [[liblttng-ctl-lttng]]
1801 === Tracing control library
1804 .The tracing control library.
1805 image::plumbing-liblttng-ctl.png[]
1807 The _LTTng control library_, `liblttng-ctl`, is used to communicate
1808 with a <<lttng-sessiond,session daemon>> using a C API that hides the
1809 underlying protocol's details. `liblttng-ctl` is part of LTTng-tools.
1811 The <<lttng-cli,cmd:lttng command-line tool>>
1812 is linked with `liblttng-ctl`.
1814 You can use `liblttng-ctl` in C or $$C++$$ source code by including its
1819 #include <lttng/lttng.h>
1822 Some objects are referenced by name (C string), such as tracing
1823 sessions, but most of them require to create a handle first using
1824 `lttng_create_handle()`.
1826 The best available developer documentation for `liblttng-ctl` is, as of
1827 LTTng{nbsp}{revision}, its installed header files. Every function and
1828 structure is thoroughly documented.
1832 === User space tracing library
1835 .The user space tracing library.
1836 image::plumbing-liblttng-ust.png[]
1838 The _user space tracing library_, `liblttng-ust` (see man:lttng-ust(3)),
1839 is the LTTng user space tracer. It receives commands from a
1840 <<lttng-sessiond,session daemon>>, for example to
1841 enable and disable specific instrumentation points, and writes event
1842 records to ring buffers shared with a
1843 <<lttng-consumerd,consumer daemon>>.
1844 `liblttng-ust` is part of LTTng-UST.
1846 Public C header files are installed beside `liblttng-ust` to
1847 instrument any <<c-application,C or $$C++$$ application>>.
1849 <<lttng-ust-agents,LTTng-UST agents>>, which are regular Java and Python
1850 packages, use their own library providing tracepoints which is
1851 linked with `liblttng-ust`.
1853 An application or library does not have to initialize `liblttng-ust`
1854 manually: its constructor does the necessary tasks to properly register
1855 to a session daemon. The initialization phase also enables the
1856 instrumentation points matching the <<event,event rules>> that you
1860 [[lttng-ust-agents]]
1861 === User space tracing agents
1864 .The user space tracing agents.
1865 image::plumbing-lttng-ust-agents.png[]
1867 The _LTTng-UST Java and Python agents_ are regular Java and Python
1868 packages which add LTTng tracing capabilities to the
1869 native logging frameworks. The LTTng-UST agents are part of LTTng-UST.
1871 In the case of Java, the
1872 https://docs.oracle.com/javase/7/docs/api/java/util/logging/package-summary.html[`java.util.logging`
1873 core logging facilities] and
1874 https://logging.apache.org/log4j/1.2/[Apache log4j 1.2] are supported.
1875 Note that Apache Log4{nbsp}2 is not supported.
1877 In the case of Python, the standard
1878 https://docs.python.org/3/library/logging.html[`logging`] package
1879 is supported. Both Python 2 and Python 3 modules can import the
1880 LTTng-UST Python agent package.
1882 The applications using the LTTng-UST agents are in the
1883 `java.util.logging` (JUL),
1884 log4j, and Python <<domain,tracing domains>>.
1886 Both agents use the same mechanism to trace the log statements. When an
1887 agent is initialized, it creates a log handler that attaches to the root
1888 logger. The agent also registers to a <<lttng-sessiond,session daemon>>.
1889 When the application executes a log statement, it is passed to the
1890 agent's log handler by the root logger. The agent's log handler calls a
1891 native function in a tracepoint provider package shared library linked
1892 with <<lttng-ust,`liblttng-ust`>>, passing the formatted log message and
1893 other fields, like its logger name and its log level. This native
1894 function contains a user space instrumentation point, hence tracing the
1897 The log level condition of an
1898 <<event,event rule>> is considered when tracing
1899 a Java or a Python application, and it's compatible with the standard
1900 JUL, log4j, and Python log levels.
1904 === LTTng kernel modules
1907 .The LTTng kernel modules.
1908 image::plumbing-lttng-modules.png[]
1910 The _LTTng kernel modules_ are a set of Linux kernel modules
1911 which implement the kernel tracer of the LTTng project. The LTTng
1912 kernel modules are part of LTTng-modules.
1914 The LTTng kernel modules include:
1916 * A set of _probe_ modules.
1918 Each module attaches to a specific subsystem
1919 of the Linux kernel using its tracepoint instrument points. There are
1920 also modules to attach to the entry and return points of the Linux
1921 system call functions.
1923 * _Ring buffer_ modules.
1925 A ring buffer implementation is provided as kernel modules. The LTTng
1926 kernel tracer writes to the ring buffer; a
1927 <<lttng-consumerd,consumer daemon>> reads from the ring buffer.
1929 * The _LTTng kernel tracer_ module.
1930 * The _LTTng logger_ module.
1932 The LTTng logger module implements the special path:{/proc/lttng-logger}
1933 file so that any executable can generate LTTng events by opening and
1934 writing to this file.
1936 See <<proc-lttng-logger-abi,LTTng logger>>.
1938 Generally, you do not have to load the LTTng kernel modules manually
1939 (using man:modprobe(8), for example): a root <<lttng-sessiond,session
1940 daemon>> loads the necessary modules when starting. If you have extra
1941 probe modules, you can specify to load them to the session daemon on
1944 The LTTng kernel modules are installed in
1945 +/usr/lib/modules/__release__/extra+ by default, where +__release__+ is
1946 the kernel release (see `uname --kernel-release`).
1953 .The session daemon.
1954 image::plumbing-sessiond.png[]
1956 The _session daemon_, man:lttng-sessiond(8), is a daemon responsible for
1957 managing tracing sessions and for controlling the various components of
1958 LTTng. The session daemon is part of LTTng-tools.
1960 The session daemon sends control requests to and receives control
1963 * The <<lttng-ust,user space tracing library>>.
1965 Any instance of the user space tracing library first registers to
1966 a session daemon. Then, the session daemon can send requests to
1967 this instance, such as:
1970 ** Get the list of tracepoints.
1971 ** Share an <<event,event rule>> so that the user space tracing library
1972 can enable or disable tracepoints. Amongst the possible conditions
1973 of an event rule is a filter expression which `liblttng-ust` evalutes
1974 when an event occurs.
1975 ** Share <<channel,channel>> attributes and ring buffer locations.
1978 The session daemon and the user space tracing library use a Unix
1979 domain socket for their communication.
1981 * The <<lttng-ust-agents,user space tracing agents>>.
1983 Any instance of a user space tracing agent first registers to
1984 a session daemon. Then, the session daemon can send requests to
1985 this instance, such as:
1988 ** Get the list of loggers.
1989 ** Enable or disable a specific logger.
1992 The session daemon and the user space tracing agent use a TCP connection
1993 for their communication.
1995 * The <<lttng-modules,LTTng kernel tracer>>.
1996 * The <<lttng-consumerd,consumer daemon>>.
1998 The session daemon sends requests to the consumer daemon to instruct
1999 it where to send the trace data streams, amongst other information.
2001 * The <<lttng-relayd,relay daemon>>.
2003 The session daemon receives commands from the
2004 <<liblttng-ctl-lttng,tracing control library>>.
2006 The root session daemon loads the appropriate
2007 <<lttng-modules,LTTng kernel modules>> on startup. It also spawns
2008 a <<lttng-consumerd,consumer daemon>> as soon as you create
2009 an <<event,event rule>>.
2011 The session daemon does not send and receive trace data: this is the
2012 role of the <<lttng-consumerd,consumer daemon>> and
2013 <<lttng-relayd,relay daemon>>. It does, however, generate the
2014 http://diamon.org/ctf/[CTF] metadata stream.
2016 Each Unix user can have its own session daemon instance. The
2017 tracing sessions managed by different session daemons are completely
2020 The root user's session daemon is the only one which is
2021 allowed to control the LTTng kernel tracer, and its spawned consumer
2022 daemon is the only one which is allowed to consume trace data from the
2023 LTTng kernel tracer. Note, however, that any Unix user which is a member
2024 of the <<tracing-group,tracing group>> is allowed
2025 to create <<channel,channels>> in the
2026 Linux kernel <<domain,tracing domain>>, and thus to trace the Linux
2029 The <<lttng-cli,cmd:lttng command-line tool>> automatically starts a
2030 session daemon when using its `create` command if none is currently
2031 running. You can also start the session daemon manually.
2038 .The consumer daemon.
2039 image::plumbing-consumerd.png[]
2041 The _consumer daemon_, man:lttng-consumerd(8), is a daemon which shares
2042 ring buffers with user applications or with the LTTng kernel modules to
2043 collect trace data and send it to some location (on disk or to a
2044 <<lttng-relayd,relay daemon>> over the network). The consumer daemon
2045 is part of LTTng-tools.
2047 You do not start a consumer daemon manually: a consumer daemon is always
2048 spawned by a <<lttng-sessiond,session daemon>> as soon as you create an
2049 <<event,event rule>>, that is, before you start tracing. When you kill
2050 its owner session daemon, the consumer daemon also exits because it is
2051 the session daemon's child process. Command-line options of
2052 man:lttng-sessiond(8) target the consumer daemon process.
2054 There are up to two running consumer daemons per Unix user, whereas only
2055 one session daemon can run per user. This is because each process can be
2056 either 32-bit or 64-bit: if the target system runs a mixture of 32-bit
2057 and 64-bit processes, it is more efficient to have separate
2058 corresponding 32-bit and 64-bit consumer daemons. The root user is an
2059 exception: it can have up to _three_ running consumer daemons: 32-bit
2060 and 64-bit instances for its user applications, and one more
2061 reserved for collecting kernel trace data.
2069 image::plumbing-relayd.png[]
2071 The _relay daemon_, man:lttng-relayd(8), is a daemon acting as a bridge
2072 between remote session and consumer daemons, local trace files, and a
2073 remote live trace viewer. The relay daemon is part of LTTng-tools.
2075 The main purpose of the relay daemon is to implement a receiver of
2076 <<sending-trace-data-over-the-network,trace data over the network>>.
2077 This is useful when the target system does not have much file system
2078 space to record trace files locally.
2080 The relay daemon is also a server to which a
2081 <<lttng-live,live trace viewer>> can
2082 connect. The live trace viewer sends requests to the relay daemon to
2083 receive trace data as the target system emits events. The
2084 communication protocol is named _LTTng live_; it is used over TCP
2087 Note that you can start the relay daemon on the target system directly.
2088 This is the setup of choice when the use case is to view events as
2089 the target system emits them without the need of a remote system.
2093 == [[using-lttng]]Instrumentation
2095 There are many examples of tracing and monitoring in our everyday life:
2097 * You have access to real-time and historical weather reports and
2098 forecasts thanks to weather stations installed around the country.
2099 * You know your heart is safe thanks to an electrocardiogram.
2100 * You make sure not to drive your car too fast and to have enough fuel
2101 to reach your destination thanks to gauges visible on your dashboard.
2103 All the previous examples have something in common: they rely on
2104 **instruments**. Without the electrodes attached to the surface of your
2105 body's skin, cardiac monitoring is futile.
2107 LTTng, as a tracer, is no different from those real life examples. If
2108 you're about to trace a software system or, in other words, record its
2109 history of execution, you better have **instrumentation points** in the
2110 subject you're tracing, that is, the actual software.
2112 Various ways were developed to instrument a piece of software for LTTng
2113 tracing. The most straightforward one is to manually place
2114 instrumentation points, called _tracepoints_, in the software's source
2115 code. It is also possible to add instrumentation points dynamically in
2116 the Linux kernel <<domain,tracing domain>>.
2118 If you're only interested in tracing the Linux kernel, your
2119 instrumentation needs are probably already covered by LTTng's built-in
2120 <<lttng-modules,Linux kernel tracepoints>>. You may also wish to trace a
2121 user application which is already instrumented for LTTng tracing.
2122 In such cases, you can skip this whole section and read the topics of
2123 the <<controlling-tracing,Tracing control>> section.
2125 Many methods are available to instrument a piece of software for LTTng
2128 * <<c-application,User space instrumentation for C and $$C++$$
2130 * <<prebuilt-ust-helpers,Prebuilt user space tracing helpers>>.
2131 * <<java-application,User space Java agent>>.
2132 * <<python-application,User space Python agent>>.
2133 * <<proc-lttng-logger-abi,LTTng logger>>.
2134 * <<instrumenting-linux-kernel,LTTng kernel tracepoints>>.
2138 === [[cxx-application]]User space instrumentation for C and $$C++$$ applications
2140 The procedure to instrument a C or $$C++$$ user application with
2141 the <<lttng-ust,LTTng user space tracing library>>, `liblttng-ust`, is:
2143 . <<tracepoint-provider,Create the source files of a tracepoint provider
2145 . <<probing-the-application-source-code,Add tracepoints to
2146 the application's source code>>.
2147 . <<building-tracepoint-providers-and-user-application,Build and link
2148 a tracepoint provider package and the user application>>.
2150 If you need quick, man:printf(3)-like instrumentation, you can skip
2151 those steps and use <<tracef,`tracef()`>> or <<tracelog,`tracelog()`>>
2154 IMPORTANT: You need to <<installing-lttng,install>> LTTng-UST to
2155 instrument a user application with `liblttng-ust`.
2158 [[tracepoint-provider]]
2159 ==== Create the source files of a tracepoint provider package
2161 A _tracepoint provider_ is a set of compiled functions which provide
2162 **tracepoints** to an application, the type of instrumentation point
2163 supported by LTTng-UST. Those functions can emit events with
2164 user-defined fields and serialize those events as event records to one
2165 or more LTTng-UST <<channel,channel>> sub-buffers. The `tracepoint()`
2166 macro, which you <<probing-the-application-source-code,insert in a user
2167 application's source code>>, calls those functions.
2169 A _tracepoint provider package_ is an object file (`.o`) or a shared
2170 library (`.so`) which contains one or more tracepoint providers.
2171 Its source files are:
2173 * One or more <<tpp-header,tracepoint provider header>> (`.h`).
2174 * A <<tpp-source,tracepoint provider package source>> (`.c`).
2176 A tracepoint provider package is dynamically linked with `liblttng-ust`,
2177 the LTTng user space tracer, at run time.
2180 .User application linked with `liblttng-ust` and containing a tracepoint provider.
2181 image::ust-app.png[]
2183 NOTE: If you need quick, man:printf(3)-like instrumentation, you can
2184 skip creating and using a tracepoint provider and use
2185 <<tracef,`tracef()`>> or <<tracelog,`tracelog()`>> instead.
2189 ===== Create a tracepoint provider header file template
2191 A _tracepoint provider header file_ contains the tracepoint
2192 definitions of a tracepoint provider.
2194 To create a tracepoint provider header file:
2196 . Start from this template:
2200 .Tracepoint provider header file template (`.h` file extension).
2202 #undef TRACEPOINT_PROVIDER
2203 #define TRACEPOINT_PROVIDER provider_name
2205 #undef TRACEPOINT_INCLUDE
2206 #define TRACEPOINT_INCLUDE "./tp.h"
2208 #if !defined(_TP_H) || defined(TRACEPOINT_HEADER_MULTI_READ)
2211 #include <lttng/tracepoint.h>
2214 * Use TRACEPOINT_EVENT(), TRACEPOINT_EVENT_CLASS(),
2215 * TRACEPOINT_EVENT_INSTANCE(), and TRACEPOINT_LOGLEVEL() here.
2220 #include <lttng/tracepoint-event.h>
2226 * `provider_name` with the name of your tracepoint provider.
2227 * `"tp.h"` with the name of your tracepoint provider header file.
2229 . Below the `#include <lttng/tracepoint.h>` line, put your
2230 <<defining-tracepoints,tracepoint definitions>>.
2232 Your tracepoint provider name must be unique amongst all the possible
2233 tracepoint provider names used on the same target system. We
2234 suggest to include the name of your project or company in the name,
2235 for example, `org_lttng_my_project_tpp`.
2237 TIP: [[lttng-gen-tp]]You can use the man:lttng-gen-tp(1) tool to create
2238 this boilerplate for you. When using cmd:lttng-gen-tp, all you need to
2239 write are the <<defining-tracepoints,tracepoint definitions>>.
2242 [[defining-tracepoints]]
2243 ===== Create a tracepoint definition
2245 A _tracepoint definition_ defines, for a given tracepoint:
2247 * Its **input arguments**. They are the macro parameters that the
2248 `tracepoint()` macro accepts for this particular tracepoint
2249 in the user application's source code.
2250 * Its **output event fields**. They are the sources of event fields
2251 that form the payload of any event that the execution of the
2252 `tracepoint()` macro emits for this particular tracepoint.
2254 You can create a tracepoint definition by using the
2255 `TRACEPOINT_EVENT()` macro below the `#include <lttng/tracepoint.h>`
2257 <<tpp-header,tracepoint provider header file template>>.
2259 The syntax of the `TRACEPOINT_EVENT()` macro is:
2262 .`TRACEPOINT_EVENT()` macro syntax.
2265 /* Tracepoint provider name */
2268 /* Tracepoint name */
2271 /* Input arguments */
2276 /* Output event fields */
2285 * `provider_name` with your tracepoint provider name.
2286 * `tracepoint_name` with your tracepoint name.
2287 * `arguments` with the <<tpp-def-input-args,input arguments>>.
2288 * `fields` with the <<tpp-def-output-fields,output event field>>
2291 This tracepoint emits events named `provider_name:tracepoint_name`.
2294 .Event name's length limitation
2296 The concatenation of the tracepoint provider name and the
2297 tracepoint name must not exceed **254 characters**. If it does, the
2298 instrumented application compiles and runs, but LTTng throws multiple
2299 warnings and you could experience serious issues.
2302 [[tpp-def-input-args]]The syntax of the `TP_ARGS()` macro is:
2305 .`TP_ARGS()` macro syntax.
2314 * `type` with the C type of the argument.
2315 * `arg_name` with the argument name.
2317 You can repeat `type` and `arg_name` up to 10 times to have
2318 more than one argument.
2320 .`TP_ARGS()` usage with three arguments.
2332 The `TP_ARGS()` and `TP_ARGS(void)` forms are valid to create a
2333 tracepoint definition with no input arguments.
2335 [[tpp-def-output-fields]]The `TP_FIELDS()` macro contains a list of
2336 `ctf_*()` macros. Each `ctf_*()` macro defines one event field. See
2337 man:lttng-ust(3) for a complete description of the available `ctf_*()`
2338 macros. A `ctf_*()` macro specifies the type, size, and byte order of
2341 Each `ctf_*()` macro takes an _argument expression_ parameter. This is a
2342 C expression that the tracer evalutes at the `tracepoint()` macro site
2343 in the application's source code. This expression provides a field's
2344 source of data. The argument expression can include input argument names
2345 listed in the `TP_ARGS()` macro.
2347 Each `ctf_*()` macro also takes a _field name_ parameter. Field names
2348 must be unique within a given tracepoint definition.
2350 Here's a complete tracepoint definition example:
2352 .Tracepoint definition.
2354 The following tracepoint definition defines a tracepoint which takes
2355 three input arguments and has four output event fields.
2359 #include "my-custom-structure.h"
2365 const struct my_custom_structure*, my_custom_structure,
2370 ctf_string(query_field, query)
2371 ctf_float(double, ratio_field, ratio)
2372 ctf_integer(int, recv_size, my_custom_structure->recv_size)
2373 ctf_integer(int, send_size, my_custom_structure->send_size)
2378 You can refer to this tracepoint definition with the `tracepoint()`
2379 macro in your application's source code like this:
2383 tracepoint(my_provider, my_tracepoint,
2384 my_structure, some_ratio, the_query);
2388 NOTE: The LTTng tracer only evaluates tracepoint arguments at run time
2389 if they satisfy an enabled <<event,event rule>>.
2392 [[using-tracepoint-classes]]
2393 ===== Use a tracepoint class
2395 A _tracepoint class_ is a class of tracepoints which share the same
2396 output event field definitions. A _tracepoint instance_ is one
2397 instance of such a defined tracepoint class, with its own tracepoint
2400 The <<defining-tracepoints,`TRACEPOINT_EVENT()` macro>> is actually a
2401 shorthand which defines both a tracepoint class and a tracepoint
2402 instance at the same time.
2404 When you build a tracepoint provider package, the C or $$C++$$ compiler
2405 creates one serialization function for each **tracepoint class**. A
2406 serialization function is responsible for serializing the event fields
2407 of a tracepoint to a sub-buffer when tracing.
2409 For various performance reasons, when your situation requires multiple
2410 tracepoint definitions with different names, but with the same event
2411 fields, we recommend that you manually create a tracepoint class
2412 and instantiate as many tracepoint instances as needed. One positive
2413 effect of such a design, amongst other advantages, is that all
2414 tracepoint instances of the same tracepoint class reuse the same
2415 serialization function, thus reducing
2416 https://en.wikipedia.org/wiki/Cache_pollution[cache pollution].
2418 .Use a tracepoint class and tracepoint instances.
2420 Consider the following three tracepoint definitions:
2432 ctf_integer(int, userid, userid)
2433 ctf_integer(size_t, len, len)
2445 ctf_integer(int, userid, userid)
2446 ctf_integer(size_t, len, len)
2458 ctf_integer(int, userid, userid)
2459 ctf_integer(size_t, len, len)
2464 In this case, we create three tracepoint classes, with one implicit
2465 tracepoint instance for each of them: `get_account`, `get_settings`, and
2466 `get_transaction`. However, they all share the same event field names
2467 and types. Hence three identical, yet independent serialization
2468 functions are created when you build the tracepoint provider package.
2470 A better design choice is to define a single tracepoint class and three
2471 tracepoint instances:
2475 /* The tracepoint class */
2476 TRACEPOINT_EVENT_CLASS(
2477 /* Tracepoint provider name */
2480 /* Tracepoint class name */
2483 /* Input arguments */
2489 /* Output event fields */
2491 ctf_integer(int, userid, userid)
2492 ctf_integer(size_t, len, len)
2496 /* The tracepoint instances */
2497 TRACEPOINT_EVENT_INSTANCE(
2498 /* Tracepoint provider name */
2501 /* Tracepoint class name */
2504 /* Tracepoint name */
2507 /* Input arguments */
2513 TRACEPOINT_EVENT_INSTANCE(
2522 TRACEPOINT_EVENT_INSTANCE(
2535 [[assigning-log-levels]]
2536 ===== Assign a log level to a tracepoint definition
2538 You can assign an optional _log level_ to a
2539 <<defining-tracepoints,tracepoint definition>>.
2541 Assigning different levels of severity to tracepoint definitions can
2542 be useful: when you <<enabling-disabling-events,create an event rule>>,
2543 you can target tracepoints having a log level as severe as a specific
2546 The concept of LTTng-UST log levels is similar to the levels found
2547 in typical logging frameworks:
2549 * In a logging framework, the log level is given by the function
2550 or method name you use at the log statement site: `debug()`,
2551 `info()`, `warn()`, `error()`, and so on.
2552 * In LTTng-UST, you statically assign the log level to a tracepoint
2553 definition; any `tracepoint()` macro invocation which refers to
2554 this definition has this log level.
2556 You can assign a log level to a tracepoint definition with the
2557 `TRACEPOINT_LOGLEVEL()` macro. You must use this macro _after_ the
2558 <<defining-tracepoints,`TRACEPOINT_EVENT()`>> or
2559 <<using-tracepoint-classes,`TRACEPOINT_INSTANCE()`>> macro for a given
2562 The syntax of the `TRACEPOINT_LOGLEVEL()` macro is:
2565 .`TRACEPOINT_LOGLEVEL()` macro syntax.
2567 TRACEPOINT_LOGLEVEL(provider_name, tracepoint_name, log_level)
2572 * `provider_name` with the tracepoint provider name.
2573 * `tracepoint_name` with the tracepoint name.
2574 * `log_level` with the log level to assign to the tracepoint
2575 definition named `tracepoint_name` in the `provider_name`
2576 tracepoint provider.
2578 See man:lttng-ust(3) for a list of available log level names.
2580 .Assign the `TRACE_DEBUG_UNIT` log level to a tracepoint definition.
2584 /* Tracepoint definition */
2593 ctf_integer(int, userid, userid)
2594 ctf_integer(size_t, len, len)
2598 /* Log level assignment */
2599 TRACEPOINT_LOGLEVEL(my_app, get_transaction, TRACE_DEBUG_UNIT)
2605 ===== Create a tracepoint provider package source file
2607 A _tracepoint provider package source file_ is a C source file which
2608 includes a <<tpp-header,tracepoint provider header file>> to expand its
2609 macros into event serialization and other functions.
2611 You can always use the following tracepoint provider package source
2615 .Tracepoint provider package source file template.
2617 #define TRACEPOINT_CREATE_PROBES
2622 Replace `tp.h` with the name of your <<tpp-header,tracepoint provider
2623 header file>> name. You may also include more than one tracepoint
2624 provider header file here to create a tracepoint provider package
2625 holding more than one tracepoint providers.
2628 [[probing-the-application-source-code]]
2629 ==== Add tracepoints to an application's source code
2631 Once you <<tpp-header,create a tracepoint provider header file>>, you
2632 can use the `tracepoint()` macro in your application's
2633 source code to insert the tracepoints that this header
2634 <<defining-tracepoints,defined>> defines.
2636 The `tracepoint()` macro takes at least two parameters: the tracepoint
2637 provider name and the tracepoint name. The corresponding tracepoint
2638 definition defines the other parameters.
2640 .`tracepoint()` usage.
2642 The following <<defining-tracepoints,tracepoint definition>> defines a
2643 tracepoint which takes two input arguments and has two output event
2647 .Tracepoint provider header file.
2649 #include "my-custom-structure.h"
2656 const char*, cmd_name
2659 ctf_string(cmd_name, cmd_name)
2660 ctf_integer(int, number_of_args, argc)
2665 You can refer to this tracepoint definition with the `tracepoint()`
2666 macro in your application's source code like this:
2669 .Application's source file.
2673 int main(int argc, char* argv[])
2675 tracepoint(my_provider, my_tracepoint, argc, argv[0]);
2681 Note how the application's source code includes
2682 the tracepoint provider header file containing the tracepoint
2683 definitions to use, path:{tp.h}.
2686 .`tracepoint()` usage with a complex tracepoint definition.
2688 Consider this complex tracepoint definition, where multiple event
2689 fields refer to the same input arguments in their argument expression
2693 .Tracepoint provider header file.
2695 /* For `struct stat` */
2696 #include <sys/types.h>
2697 #include <sys/stat.h>
2709 ctf_integer(int, my_constant_field, 23 + 17)
2710 ctf_integer(int, my_int_arg_field, my_int_arg)
2711 ctf_integer(int, my_int_arg_field2, my_int_arg * my_int_arg)
2712 ctf_integer(int, sum4_field, my_str_arg[0] + my_str_arg[1] +
2713 my_str_arg[2] + my_str_arg[3])
2714 ctf_string(my_str_arg_field, my_str_arg)
2715 ctf_integer_hex(off_t, size_field, st->st_size)
2716 ctf_float(double, size_dbl_field, (double) st->st_size)
2717 ctf_sequence_text(char, half_my_str_arg_field, my_str_arg,
2718 size_t, strlen(my_str_arg) / 2)
2723 You can refer to this tracepoint definition with the `tracepoint()`
2724 macro in your application's source code like this:
2727 .Application's source file.
2729 #define TRACEPOINT_DEFINE
2736 stat("/etc/fstab", &s);
2737 tracepoint(my_provider, my_tracepoint, 23, "Hello, World!", &s);
2743 If you look at the event record that LTTng writes when tracing this
2744 program, assuming the file size of path:{/etc/fstab} is 301{nbsp}bytes,
2745 it should look like this:
2747 .Event record fields
2749 |Field's name |Field's value
2750 |`my_constant_field` |40
2751 |`my_int_arg_field` |23
2752 |`my_int_arg_field2` |529
2754 |`my_str_arg_field` |`Hello, World!`
2755 |`size_field` |0x12d
2756 |`size_dbl_field` |301.0
2757 |`half_my_str_arg_field` |`Hello,`
2761 Sometimes, the arguments you pass to `tracepoint()` are expensive to
2762 compute--they use the call stack, for example. To avoid this
2763 computation when the tracepoint is disabled, you can use the
2764 `tracepoint_enabled()` and `do_tracepoint()` macros.
2766 The syntax of the `tracepoint_enabled()` and `do_tracepoint()` macros
2770 .`tracepoint_enabled()` and `do_tracepoint()` macros syntax.
2772 tracepoint_enabled(provider_name, tracepoint_name)
2773 do_tracepoint(provider_name, tracepoint_name, ...)
2778 * `provider_name` with the tracepoint provider name.
2779 * `tracepoint_name` with the tracepoint name.
2781 `tracepoint_enabled()` returns a non-zero value if the tracepoint named
2782 `tracepoint_name` from the provider named `provider_name` is enabled
2785 `do_tracepoint()` is like `tracepoint()`, except that it doesn't check
2786 if the tracepoint is enabled. Using `tracepoint()` with
2787 `tracepoint_enabled()` is dangerous since `tracepoint()` also contains
2788 the `tracepoint_enabled()` check, thus a race condition is
2789 possible in this situation:
2792 .Possible race condition when using `tracepoint_enabled()` with `tracepoint()`.
2794 if (tracepoint_enabled(my_provider, my_tracepoint)) {
2795 stuff = prepare_stuff();
2798 tracepoint(my_provider, my_tracepoint, stuff);
2801 If the tracepoint is enabled after the condition, then `stuff` is not
2802 prepared: the emitted event will either contain wrong data, or the whole
2803 application could crash (segmentation fault, for example).
2805 NOTE: Neither `tracepoint_enabled()` nor `do_tracepoint()` have an
2806 `STAP_PROBEV()` call. If you need it, you must emit
2810 [[building-tracepoint-providers-and-user-application]]
2811 ==== Build and link a tracepoint provider package and an application
2813 Once you have one or more <<tpp-header,tracepoint provider header
2814 files>> and a <<tpp-source,tracepoint provider package source file>>,
2815 you can create the tracepoint provider package by compiling its source
2816 file. From here, multiple build and run scenarios are possible. The
2817 following table shows common application and library configurations
2818 along with the required command lines to achieve them.
2820 In the following diagrams, we use the following file names:
2823 Executable application.
2826 Application's object file.
2829 Tracepoint provider package object file.
2832 Tracepoint provider package archive file.
2835 Tracepoint provider package shared object file.
2838 User library object file.
2841 User library shared object file.
2843 The red star indicates that this object file is instrumented
2844 (contains code which uses the `tracepoint()` macro). The spring
2845 symbol between the application and a library means the application is
2846 linked with the library at build time.
2848 We assume that path:{.} is part of the env:LD_LIBRARY_PATH environment
2849 variable in the following instructions.
2851 [role="growable ust-scenarios",cols="asciidoc,asciidoc"]
2852 .Common tracepoint provider package scenarios.
2854 |Scenario |Instructions
2857 The instrumented application is statically linked with
2858 the tracepoint provider package object.
2860 image::ust-sit+app-linked-with-tp-o+app-instrumented.png[]
2863 include::../common/ust-sit-step-tp-o.txt[]
2865 To build the instrumented application:
2867 . In path:{app.c}, before including path:{tpp.h}, add the following line:
2872 #define TRACEPOINT_DEFINE
2876 . Compile the application source file:
2885 . Build the application:
2890 gcc -o app app.o tpp.o -llttng-ust -ldl
2894 To run the instrumented application:
2896 * Start the application:
2906 The instrumented application is statically linked with the
2907 tracepoint provider package archive file.
2909 image::ust-sit+app-linked-with-tp-a+app-instrumented.png[]
2912 To create the tracepoint provider package archive file:
2914 . Compile the <<tpp-source,tracepoint provider package source file>>:
2923 . Create the tracepoint provider package archive file:
2932 To build the instrumented application:
2934 . In path:{app.c}, before including path:{tpp.h}, add the following line:
2939 #define TRACEPOINT_DEFINE
2943 . Compile the application source file:
2952 . Build the application:
2957 gcc -o app app.o tpp.a -llttng-ust -ldl
2961 To run the instrumented application:
2963 * Start the application:
2973 The instrumented application is linked with the tracepoint provider
2974 package shared object.
2976 image::ust-sit+app-linked-with-tp-so+app-instrumented.png[]
2979 include::../common/ust-sit-step-tp-so.txt[]
2981 To build the instrumented application:
2983 . In path:{app.c}, before including path:{tpp.h}, add the following line:
2988 #define TRACEPOINT_DEFINE
2992 . Compile the application source file:
3001 . Build the application:
3006 gcc -o app app.o -ldl -L. -ltpp
3010 To run the instrumented application:
3012 * Start the application:
3022 The tracepoint provider package shared object is preloaded before the
3023 instrumented application starts.
3025 image::ust-sit+tp-so-preloaded+app-instrumented.png[]
3028 include::../common/ust-sit-step-tp-so.txt[]
3030 To build the instrumented application:
3032 . In path:{app.c}, before including path:{tpp.h}, add the
3038 #define TRACEPOINT_DEFINE
3039 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3043 . Compile the application source file:
3052 . Build the application:
3057 gcc -o app app.o -ldl
3061 To run the instrumented application with tracing support:
3063 * Preload the tracepoint provider package shared object and
3064 start the application:
3069 LD_PRELOAD=./libtpp.so ./app
3073 To run the instrumented application without tracing support:
3075 * Start the application:
3085 The instrumented application dynamically loads the tracepoint provider
3086 package shared object.
3088 See the <<dlclose-warning,warning about `dlclose()`>>.
3090 image::ust-sit+app-dlopens-tp-so+app-instrumented.png[]
3093 include::../common/ust-sit-step-tp-so.txt[]
3095 To build the instrumented application:
3097 . In path:{app.c}, before including path:{tpp.h}, add the
3103 #define TRACEPOINT_DEFINE
3104 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3108 . Compile the application source file:
3117 . Build the application:
3122 gcc -o app app.o -ldl
3126 To run the instrumented application:
3128 * Start the application:
3138 The application is linked with the instrumented user library.
3140 The instrumented user library is statically linked with the tracepoint
3141 provider package object file.
3143 image::ust-sit+app-linked-with-lib+lib-linked-with-tp-o+lib-instrumented.png[]
3146 include::../common/ust-sit-step-tp-o-fpic.txt[]
3148 To build the instrumented user library:
3150 . In path:{emon.c}, before including path:{tpp.h}, add the
3156 #define TRACEPOINT_DEFINE
3160 . Compile the user library source file:
3165 gcc -I. -fpic -c emon.c
3169 . Build the user library shared object:
3174 gcc -shared -o libemon.so emon.o tpp.o -llttng-ust -ldl
3178 To build the application:
3180 . Compile the application source file:
3189 . Build the application:
3194 gcc -o app app.o -L. -lemon
3198 To run the application:
3200 * Start the application:
3210 The application is linked with the instrumented user library.
3212 The instrumented user library is linked with the tracepoint provider
3213 package shared object.
3215 image::ust-sit+app-linked-with-lib+lib-linked-with-tp-so+lib-instrumented.png[]
3218 include::../common/ust-sit-step-tp-so.txt[]
3220 To build the instrumented user library:
3222 . In path:{emon.c}, before including path:{tpp.h}, add the
3228 #define TRACEPOINT_DEFINE
3232 . Compile the user library source file:
3237 gcc -I. -fpic -c emon.c
3241 . Build the user library shared object:
3246 gcc -shared -o libemon.so emon.o -ldl -L. -ltpp
3250 To build the application:
3252 . Compile the application source file:
3261 . Build the application:
3266 gcc -o app app.o -L. -lemon
3270 To run the application:
3272 * Start the application:
3282 The tracepoint provider package shared object is preloaded before the
3285 The application is linked with the instrumented user library.
3287 image::ust-sit+tp-so-preloaded+app-linked-with-lib+lib-instrumented.png[]
3290 include::../common/ust-sit-step-tp-so.txt[]
3292 To build the instrumented user library:
3294 . In path:{emon.c}, before including path:{tpp.h}, add the
3300 #define TRACEPOINT_DEFINE
3301 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3305 . Compile the user library source file:
3310 gcc -I. -fpic -c emon.c
3314 . Build the user library shared object:
3319 gcc -shared -o libemon.so emon.o -ldl
3323 To build the application:
3325 . Compile the application source file:
3334 . Build the application:
3339 gcc -o app app.o -L. -lemon
3343 To run the application with tracing support:
3345 * Preload the tracepoint provider package shared object and
3346 start the application:
3351 LD_PRELOAD=./libtpp.so ./app
3355 To run the application without tracing support:
3357 * Start the application:
3367 The application is linked with the instrumented user library.
3369 The instrumented user library dynamically loads the tracepoint provider
3370 package shared object.
3372 See the <<dlclose-warning,warning about `dlclose()`>>.
3374 image::ust-sit+app-linked-with-lib+lib-dlopens-tp-so+lib-instrumented.png[]
3377 include::../common/ust-sit-step-tp-so.txt[]
3379 To build the instrumented user library:
3381 . In path:{emon.c}, before including path:{tpp.h}, add the
3387 #define TRACEPOINT_DEFINE
3388 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3392 . Compile the user library source file:
3397 gcc -I. -fpic -c emon.c
3401 . Build the user library shared object:
3406 gcc -shared -o libemon.so emon.o -ldl
3410 To build the application:
3412 . Compile the application source file:
3421 . Build the application:
3426 gcc -o app app.o -L. -lemon
3430 To run the application:
3432 * Start the application:
3442 The application dynamically loads the instrumented user library.
3444 The instrumented user library is linked with the tracepoint provider
3445 package shared object.
3447 See the <<dlclose-warning,warning about `dlclose()`>>.
3449 image::ust-sit+app-dlopens-lib+lib-linked-with-tp-so+lib-instrumented.png[]
3452 include::../common/ust-sit-step-tp-so.txt[]
3454 To build the instrumented user library:
3456 . In path:{emon.c}, before including path:{tpp.h}, add the
3462 #define TRACEPOINT_DEFINE
3466 . Compile the user library source file:
3471 gcc -I. -fpic -c emon.c
3475 . Build the user library shared object:
3480 gcc -shared -o libemon.so emon.o -ldl -L. -ltpp
3484 To build the application:
3486 . Compile the application source file:
3495 . Build the application:
3500 gcc -o app app.o -ldl -L. -lemon
3504 To run the application:
3506 * Start the application:
3516 The application dynamically loads the instrumented user library.
3518 The instrumented user library dynamically loads the tracepoint provider
3519 package shared object.
3521 See the <<dlclose-warning,warning about `dlclose()`>>.
3523 image::ust-sit+app-dlopens-lib+lib-dlopens-tp-so+lib-instrumented.png[]
3526 include::../common/ust-sit-step-tp-so.txt[]
3528 To build the instrumented user library:
3530 . In path:{emon.c}, before including path:{tpp.h}, add the
3536 #define TRACEPOINT_DEFINE
3537 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3541 . Compile the user library source file:
3546 gcc -I. -fpic -c emon.c
3550 . Build the user library shared object:
3555 gcc -shared -o libemon.so emon.o -ldl
3559 To build the application:
3561 . Compile the application source file:
3570 . Build the application:
3575 gcc -o app app.o -ldl -L. -lemon
3579 To run the application:
3581 * Start the application:
3591 The tracepoint provider package shared object is preloaded before the
3594 The application dynamically loads the instrumented user library.
3596 image::ust-sit+tp-so-preloaded+app-dlopens-lib+lib-instrumented.png[]
3599 include::../common/ust-sit-step-tp-so.txt[]
3601 To build the instrumented user library:
3603 . In path:{emon.c}, before including path:{tpp.h}, add the
3609 #define TRACEPOINT_DEFINE
3610 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3614 . Compile the user library source file:
3619 gcc -I. -fpic -c emon.c
3623 . Build the user library shared object:
3628 gcc -shared -o libemon.so emon.o -ldl
3632 To build the application:
3634 . Compile the application source file:
3643 . Build the application:
3648 gcc -o app app.o -L. -lemon
3652 To run the application with tracing support:
3654 * Preload the tracepoint provider package shared object and
3655 start the application:
3660 LD_PRELOAD=./libtpp.so ./app
3664 To run the application without tracing support:
3666 * Start the application:
3676 The application is statically linked with the tracepoint provider
3677 package object file.
3679 The application is linked with the instrumented user library.
3681 image::ust-sit+app-linked-with-tp-o+app-linked-with-lib+lib-instrumented.png[]
3684 include::../common/ust-sit-step-tp-o.txt[]
3686 To build the instrumented user library:
3688 . In path:{emon.c}, before including path:{tpp.h}, add the
3694 #define TRACEPOINT_DEFINE
3698 . Compile the user library source file:
3703 gcc -I. -fpic -c emon.c
3707 . Build the user library shared object:
3712 gcc -shared -o libemon.so emon.o
3716 To build the application:
3718 . Compile the application source file:
3727 . Build the application:
3732 gcc -o app app.o tpp.o -llttng-ust -ldl -L. -lemon
3736 To run the instrumented application:
3738 * Start the application:
3748 The application is statically linked with the tracepoint provider
3749 package object file.
3751 The application dynamically loads the instrumented user library.
3753 image::ust-sit+app-linked-with-tp-o+app-dlopens-lib+lib-instrumented.png[]
3756 include::../common/ust-sit-step-tp-o.txt[]
3758 To build the application:
3760 . In path:{app.c}, before including path:{tpp.h}, add the following line:
3765 #define TRACEPOINT_DEFINE
3769 . Compile the application source file:
3778 . Build the application:
3783 gcc -Wl,--export-dynamic -o app app.o tpp.o \
3788 The `--export-dynamic` option passed to the linker is necessary for the
3789 dynamically loaded library to ``see'' the tracepoint symbols defined in
3792 To build the instrumented user library:
3794 . Compile the user library source file:
3799 gcc -I. -fpic -c emon.c
3803 . Build the user library shared object:
3808 gcc -shared -o libemon.so emon.o
3812 To run the application:
3814 * Start the application:
3826 .Do not use man:dlclose(3) on a tracepoint provider package
3828 Never use man:dlclose(3) on any shared object which:
3830 * Is linked with, statically or dynamically, a tracepoint provider
3832 * Calls man:dlopen(3) itself to dynamically open a tracepoint provider
3833 package shared object.
3835 This is currently considered **unsafe** due to a lack of reference
3836 counting from LTTng-UST to the shared object.
3838 A known workaround (available since glibc 2.2) is to use the
3839 `RTLD_NODELETE` flag when calling man:dlopen(3) initially. This has the
3840 effect of not unloading the loaded shared object, even if man:dlclose(3)
3843 You can also preload the tracepoint provider package shared object with
3844 the env:LD_PRELOAD environment variable to overcome this limitation.
3848 [[using-lttng-ust-with-daemons]]
3849 ===== Use noch:{LTTng-UST} with daemons
3851 If your instrumented application calls man:fork(2), man:clone(2),
3852 or BSD's man:rfork(2), without a following man:exec(3)-family
3853 system call, you must preload the path:{liblttng-ust-fork.so} shared
3854 object when starting the application.
3858 LD_PRELOAD=liblttng-ust-fork.so ./my-app
3861 If your tracepoint provider package is
3862 a shared library which you also preload, you must put both
3863 shared objects in env:LD_PRELOAD:
3867 LD_PRELOAD=liblttng-ust-fork.so:/path/to/tp.so ./my-app
3871 [[lttng-ust-pkg-config]]
3872 ===== Use noch:{pkg-config}
3874 On some distributions, LTTng-UST ships with a
3875 https://www.freedesktop.org/wiki/Software/pkg-config/[pkg-config]
3876 metadata file. If this is your case, then you can use cmd:pkg-config to
3877 build an application on the command line:
3881 gcc -o my-app my-app.o tp.o $(pkg-config --cflags --libs lttng-ust)
3885 [[instrumenting-32-bit-app-on-64-bit-system]]
3886 ===== [[advanced-instrumenting-techniques]]Build a 32-bit instrumented application for a 64-bit target system
3888 In order to trace a 32-bit application running on a 64-bit system,
3889 LTTng must use a dedicated 32-bit
3890 <<lttng-consumerd,consumer daemon>>.
3892 The following steps show how to build and install a 32-bit consumer
3893 daemon, which is _not_ part of the default 64-bit LTTng build, how to
3894 build and install the 32-bit LTTng-UST libraries, and how to build and
3895 link an instrumented 32-bit application in that context.
3897 To build a 32-bit instrumented application for a 64-bit target system,
3898 assuming you have a fresh target system with no installed Userspace RCU
3901 . Download, build, and install a 32-bit version of Userspace RCU:
3907 wget http://lttng.org/files/urcu/userspace-rcu-latest-0.9.tar.bz2 &&
3908 tar -xf userspace-rcu-latest-0.9.tar.bz2 &&
3909 cd userspace-rcu-0.9.* &&
3910 ./configure --libdir=/usr/local/lib32 CFLAGS=-m32 &&
3912 sudo make install &&
3917 . Using your distribution's package manager, or from source, install
3918 the following 32-bit versions of the following dependencies of
3919 LTTng-tools and LTTng-UST:
3922 * https://sourceforge.net/projects/libuuid/[libuuid]
3923 * http://directory.fsf.org/wiki/Popt[popt]
3924 * http://www.xmlsoft.org/[libxml2]
3927 . Download, build, and install a 32-bit version of the latest
3928 LTTng-UST{nbsp}{revision}:
3934 wget http://lttng.org/files/lttng-ust/lttng-ust-latest-2.8.tar.bz2 &&
3935 tar -xf lttng-ust-latest-2.8.tar.bz2 &&
3936 cd lttng-ust-2.8.* &&
3937 ./configure --libdir=/usr/local/lib32 \
3938 CFLAGS=-m32 CXXFLAGS=-m32 \
3939 LDFLAGS='-L/usr/local/lib32 -L/usr/lib32' &&
3941 sudo make install &&
3948 Depending on your distribution,
3949 32-bit libraries could be installed at a different location than
3950 `/usr/lib32`. For example, Debian is known to install
3951 some 32-bit libraries in `/usr/lib/i386-linux-gnu`.
3953 In this case, make sure to set `LDFLAGS` to all the
3954 relevant 32-bit library paths, for example:
3958 LDFLAGS='-L/usr/lib/i386-linux-gnu -L/usr/lib32'
3962 . Download the latest LTTng-tools{nbsp}{revision}, build, and install
3963 the 32-bit consumer daemon:
3969 wget http://lttng.org/files/lttng-tools/lttng-tools-latest-2.8.tar.bz2 &&
3970 tar -xf lttng-tools-latest-2.8.tar.bz2 &&
3971 cd lttng-tools-2.8.* &&
3972 ./configure --libdir=/usr/local/lib32 CFLAGS=-m32 CXXFLAGS=-m32 \
3973 LDFLAGS='-L/usr/local/lib32 -L/usr/lib32' &&
3975 cd src/bin/lttng-consumerd &&
3976 sudo make install &&
3981 . From your distribution or from source,
3982 <<installing-lttng,install>> the 64-bit versions of
3983 LTTng-UST and Userspace RCU.
3984 . Download, build, and install the 64-bit version of the
3985 latest LTTng-tools{nbsp}{revision}:
3991 wget http://lttng.org/files/lttng-tools/lttng-tools-latest-2.8.tar.bz2 &&
3992 tar -xf lttng-tools-latest-2.8.tar.bz2 &&
3993 cd lttng-tools-2.8.* &&
3994 ./configure --with-consumerd32-libdir=/usr/local/lib32 \
3995 --with-consumerd32-bin=/usr/local/lib32/lttng/libexec/lttng-consumerd &&
3997 sudo make install &&
4002 . Pass the following options to man:gcc(1), man:g++(1), or man:clang(1)
4003 when linking your 32-bit application:
4006 -m32 -L/usr/lib32 -L/usr/local/lib32 \
4007 -Wl,-rpath,/usr/lib32,-rpath,/usr/local/lib32
4010 For example, let's rebuild the quick start example in
4011 <<tracing-your-own-user-application,Trace a user application>> as an
4012 instrumented 32-bit application:
4017 gcc -m32 -c -I. hello-tp.c
4019 gcc -m32 -o hello hello.o hello-tp.o \
4020 -L/usr/lib32 -L/usr/local/lib32 \
4021 -Wl,-rpath,/usr/lib32,-rpath,/usr/local/lib32 \
4026 No special action is required to execute the 32-bit application and
4027 to trace it: use the command-line man:lttng(1) tool as usual.
4034 man:tracef(3) is a small LTTng-UST API designed for quick,
4035 man:printf(3)-like instrumentation without the burden of
4036 <<tracepoint-provider,creating>> and
4037 <<building-tracepoint-providers-and-user-application,building>>
4038 a tracepoint provider package.
4040 To use `tracef()` in your application:
4042 . In the C or C++ source files where you need to use `tracef()`,
4043 include `<lttng/tracef.h>`:
4048 #include <lttng/tracef.h>
4052 . In the application's source code, use `tracef()` like you would use
4060 tracef("my message: %d (%s)", my_integer, my_string);
4066 . Link your application with `liblttng-ust`:
4071 gcc -o app app.c -llttng-ust
4075 To trace the events that `tracef()` calls emit:
4077 * <<enabling-disabling-events,Create an event rule>> which matches the
4078 `lttng_ust_tracef:*` event name:
4083 lttng enable-event --userspace 'lttng_ust_tracef:*'
4088 .Limitations of `tracef()`
4090 The `tracef()` utility function was developed to make user space tracing
4091 super simple, albeit with notable disadvantages compared to
4092 <<defining-tracepoints,user-defined tracepoints>>:
4094 * All the emitted events have the same tracepoint provider and
4095 tracepoint names, respectively `lttng_ust_tracef` and `event`.
4096 * There is no static type checking.
4097 * The only event record field you actually get, named `msg`, is a string
4098 potentially containing the values you passed to `tracef()`
4099 using your own format string. This also means that you cannot filter
4100 events with a custom expression at run time because there are no
4102 * Since `tracef()` uses the C standard library's man:vasprintf(3)
4103 function behind the scenes to format the strings at run time, its
4104 expected performance is lower than with user-defined tracepoints,
4105 which do not require a conversion to a string.
4107 Taking this into consideration, `tracef()` is useful for some quick
4108 prototyping and debugging, but you should not consider it for any
4109 permanent and serious applicative instrumentation.
4115 ==== Use `tracelog()`
4117 The man:tracelog(3) API is very similar to <<tracef,`tracef()`>>, with
4118 the difference that it accepts an additional log level parameter.
4120 The goal of `tracelog()` is to ease the migration from logging to
4123 To use `tracelog()` in your application:
4125 . In the C or C++ source files where you need to use `tracelog()`,
4126 include `<lttng/tracelog.h>`:
4131 #include <lttng/tracelog.h>
4135 . In the application's source code, use `tracelog()` like you would use
4136 man:printf(3), except for the first parameter which is the log
4144 tracelog(TRACE_WARNING, "my message: %d (%s)",
4145 my_integer, my_string);
4151 See man:lttng-ust(3) for a list of available log level names.
4153 . Link your application with `liblttng-ust`:
4158 gcc -o app app.c -llttng-ust
4162 To trace the events that `tracelog()` calls emit with a log level
4163 _as severe as_ a specific log level:
4165 * <<enabling-disabling-events,Create an event rule>> which matches the
4166 `lttng_ust_tracelog:*` event name and a minimum level
4172 lttng enable-event --userspace 'lttng_ust_tracelog:*'
4173 --loglevel=TRACE_WARNING
4177 To trace the events that `tracelog()` calls emit with a
4178 _specific log level_:
4180 * Create an event rule which matches the `lttng_ust_tracelog:*`
4181 event name and a specific log level:
4186 lttng enable-event --userspace 'lttng_ust_tracelog:*'
4187 --loglevel-only=TRACE_INFO
4192 [[prebuilt-ust-helpers]]
4193 === Prebuilt user space tracing helpers
4195 The LTTng-UST package provides a few helpers in the form or preloadable
4196 shared objects which automatically instrument system functions and
4199 The helper shared objects are normally found in dir:{/usr/lib}. If you
4200 built LTTng-UST <<building-from-source,from source>>, they are probably
4201 located in dir:{/usr/local/lib}.
4203 The installed user space tracing helpers in LTTng-UST{nbsp}{revision}
4206 path:{liblttng-ust-libc-wrapper.so}::
4207 path:{liblttng-ust-pthread-wrapper.so}::
4208 <<liblttng-ust-libc-pthread-wrapper,C{nbsp}standard library
4209 memory and POSIX threads function tracing>>.
4211 path:{liblttng-ust-cyg-profile.so}::
4212 path:{liblttng-ust-cyg-profile-fast.so}::
4213 <<liblttng-ust-cyg-profile,Function entry and exit tracing>>.
4215 path:{liblttng-ust-dl.so}::
4216 <<liblttng-ust-dl,Dynamic linker tracing>>.
4218 To use a user space tracing helper with any user application:
4220 * Preload the helper shared object when you start the application:
4225 LD_PRELOAD=liblttng-ust-libc-wrapper.so my-app
4229 You can preload more than one helper:
4234 LD_PRELOAD=liblttng-ust-libc-wrapper.so:liblttng-ust-dl.so my-app
4240 [[liblttng-ust-libc-pthread-wrapper]]
4241 ==== Instrument C standard library memory and POSIX threads functions
4243 The path:{liblttng-ust-libc-wrapper.so} and
4244 path:{liblttng-ust-pthread-wrapper.so} helpers
4245 add instrumentation to some C standard library and POSIX
4249 .Functions instrumented by preloading path:{liblttng-ust-libc-wrapper.so}.
4251 |TP provider name |TP name |Instrumented function
4253 .6+|`lttng_ust_libc` |`malloc` |man:malloc(3)
4254 |`calloc` |man:calloc(3)
4255 |`realloc` |man:realloc(3)
4256 |`free` |man:free(3)
4257 |`memalign` |man:memalign(3)
4258 |`posix_memalign` |man:posix_memalign(3)
4262 .Functions instrumented by preloading path:{liblttng-ust-pthread-wrapper.so}.
4264 |TP provider name |TP name |Instrumented function
4266 .4+|`lttng_ust_pthread` |`pthread_mutex_lock_req` |man:pthread_mutex_lock(3p) (request time)
4267 |`pthread_mutex_lock_acq` |man:pthread_mutex_lock(3p) (acquire time)
4268 |`pthread_mutex_trylock` |man:pthread_mutex_trylock(3p)
4269 |`pthread_mutex_unlock` |man:pthread_mutex_unlock(3p)
4272 When you preload the shared object, it replaces the functions listed
4273 in the previous tables by wrappers which contain tracepoints and call
4274 the replaced functions.
4277 [[liblttng-ust-cyg-profile]]
4278 ==== Instrument function entry and exit
4280 The path:{liblttng-ust-cyg-profile*.so} helpers can add instrumentation
4281 to the entry and exit points of functions.
4283 man:gcc(1) and man:clang(1) have an option named
4284 https://gcc.gnu.org/onlinedocs/gcc/Code-Gen-Options.html[`-finstrument-functions`]
4285 which generates instrumentation calls for entry and exit to functions.
4286 The LTTng-UST function tracing helpers,
4287 path:{liblttng-ust-cyg-profile.so} and
4288 path:{liblttng-ust-cyg-profile-fast.so}, take advantage of this feature
4289 to add tracepoints to the two generated functions (which contain
4290 `cyg_profile` in their names, hence the helper's name).
4292 To use the LTTng-UST function tracing helper, the source files to
4293 instrument must be built using the `-finstrument-functions` compiler
4296 There are two versions of the LTTng-UST function tracing helper:
4298 * **path:{liblttng-ust-cyg-profile-fast.so}** is a lightweight variant
4299 that you should only use when it can be _guaranteed_ that the
4300 complete event stream is recorded without any lost event record.
4301 Any kind of duplicate information is left out.
4303 Assuming no event record is lost, having only the function addresses on
4304 entry is enough to create a call graph, since an event record always
4305 contains the ID of the CPU that generated it.
4307 You can use a tool like
4308 https://sourceware.org/binutils/docs/binutils/addr2line.html[cmd:addr2line]
4309 to convert function addresses back to source file names and
4312 * **path:{liblttng-ust-cyg-profile.so}** is a more robust variant
4313 which also works in use cases where event records might get discarded or
4314 not recorded from application startup.
4315 In these cases, the trace analyzer needs more information to be
4316 able to reconstruct the program flow.
4318 See man:lttng-ust-cyg-profile(3) to learn more about the instrumentation
4319 points of this helper.
4321 All the tracepoints that this helper provides have the
4322 log level `TRACE_DEBUG_FUNCTION` (see man:lttng-ust(3)).
4324 TIP: It's sometimes a good idea to limit the number of source files that
4325 you compile with the `-finstrument-functions` option to prevent LTTng
4326 from writing an excessive amount of trace data at run time. When using
4327 man:gcc(1), you can use the
4328 `-finstrument-functions-exclude-function-list` option to avoid
4329 instrument entries and exits of specific function names.
4334 ==== Instrument the dynamic linker
4336 The path:{liblttng-ust-dl.so} helper adds instrumentation to the
4337 man:dlopen(3) and man:dlclose(3) function calls.
4339 See man:lttng-ust-dl(3) to learn more about the instrumentation points
4344 [[java-application]]
4345 === User space Java agent
4347 You can instrument any Java application which uses one of the following
4350 * The https://docs.oracle.com/javase/7/docs/api/java/util/logging/package-summary.html[**`java.util.logging`**]
4351 (JUL) core logging facilities.
4352 * http://logging.apache.org/log4j/1.2/[**Apache log4j 1.2**], since
4353 LTTng 2.6. Note that Apache Log4j{nbsp}2 is not supported.
4356 .LTTng-UST Java agent imported by a Java application.
4357 image::java-app.png[]
4359 Note that the methods described below are new in LTTng{nbsp}{revision}.
4360 Previous LTTng versions use another technique.
4362 NOTE: We use http://openjdk.java.net/[OpenJDK]{nbsp}8 for development
4363 and https://ci.lttng.org/[continuous integration], thus this version is
4364 directly supported. However, the LTTng-UST Java agent is also tested
4365 with OpenJDK{nbsp}7.
4370 ==== Use the LTTng-UST Java agent for `java.util.logging`
4372 To use the LTTng-UST Java agent in a Java application which uses
4373 `java.util.logging` (JUL):
4375 . In the Java application's source code, import the LTTng-UST
4376 log handler package for `java.util.logging`:
4381 import org.lttng.ust.agent.jul.LttngLogHandler;
4385 . Create an LTTng-UST JUL log handler:
4390 Handler lttngUstLogHandler = new LttngLogHandler();
4394 . Add this handler to the JUL loggers which should emit LTTng events:
4399 Logger myLogger = Logger.getLogger("some-logger");
4401 myLogger.addHandler(lttngUstLogHandler);
4405 . Use `java.util.logging` log statements and configuration as usual.
4406 The loggers with an attached LTTng-UST log handler can emit
4409 . Before exiting the application, remove the LTTng-UST log handler from
4410 the loggers attached to it and call its `close()` method:
4415 myLogger.removeHandler(lttngUstLogHandler);
4416 lttngUstLogHandler.close();
4420 This is not strictly necessary, but it is recommended for a clean
4421 disposal of the handler's resources.
4423 . Include the LTTng-UST Java agent's common and JUL-specific JAR files,
4424 path:{lttng-ust-agent-common.jar} and path:{lttng-ust-agent-jul.jar},
4426 https://docs.oracle.com/javase/tutorial/essential/environment/paths.html[class
4427 path] when you build the Java application.
4429 The JAR files are typically located in dir:{/usr/share/java}.
4431 IMPORTANT: The LTTng-UST Java agent must be
4432 <<installing-lttng,installed>> for the logging framework your
4435 .Use the LTTng-UST Java agent for `java.util.logging`.
4440 import java.io.IOException;
4441 import java.util.logging.Handler;
4442 import java.util.logging.Logger;
4443 import org.lttng.ust.agent.jul.LttngLogHandler;
4447 private static final int answer = 42;
4449 public static void main(String[] argv) throws Exception
4452 Logger logger = Logger.getLogger("jello");
4454 // Create an LTTng-UST log handler
4455 Handler lttngUstLogHandler = new LttngLogHandler();
4457 // Add the LTTng-UST log handler to our logger
4458 logger.addHandler(lttngUstLogHandler);
4461 logger.info("some info");
4462 logger.warning("some warning");
4464 logger.finer("finer information; the answer is " + answer);
4466 logger.severe("error!");
4468 // Not mandatory, but cleaner
4469 logger.removeHandler(lttngUstLogHandler);
4470 lttngUstLogHandler.close();
4479 javac -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-jul.jar Test.java
4482 <<creating-destroying-tracing-sessions,Create a tracing session>>,
4483 <<enabling-disabling-events,create an event rule>> matching the
4484 `jello` JUL logger, and <<basic-tracing-session-control,start tracing>>:
4489 lttng enable-event --jul jello
4493 Run the compiled class:
4497 java -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-jul.jar:. Test
4500 <<basic-tracing-session-control,Stop tracing>> and inspect the
4510 You can use the opt:lttng-enable-event(1):--loglevel or
4511 opt:lttng-enable-event(1):--loglevel-only option of the
4512 man:lttng-enable-event(1) command to target a range of JUL log levels
4513 or a specific JUL log level.
4518 ==== Use the LTTng-UST Java agent for Apache log4j
4520 To use the LTTng-UST Java agent in a Java application which uses
4523 . In the Java application's source code, import the LTTng-UST
4524 log appender package for Apache log4j:
4529 import org.lttng.ust.agent.log4j.LttngLogAppender;
4533 . Create an LTTng-UST log4j log appender:
4538 Appender lttngUstLogAppender = new LttngLogAppender();
4542 . Add this appender to the log4j loggers which should emit LTTng events:
4547 Logger myLogger = Logger.getLogger("some-logger");
4549 myLogger.addAppender(lttngUstLogAppender);
4553 . Use Apache log4j log statements and configuration as usual. The
4554 loggers with an attached LTTng-UST log appender can emit LTTng events.
4556 . Before exiting the application, remove the LTTng-UST log appender from
4557 the loggers attached to it and call its `close()` method:
4562 myLogger.removeAppender(lttngUstLogAppender);
4563 lttngUstLogAppender.close();
4567 This is not strictly necessary, but it is recommended for a clean
4568 disposal of the appender's resources.
4570 . Include the LTTng-UST Java agent's common and log4j-specific JAR
4571 files, path:{lttng-ust-agent-common.jar} and
4572 path:{lttng-ust-agent-log4j.jar}, in the
4573 https://docs.oracle.com/javase/tutorial/essential/environment/paths.html[class
4574 path] when you build the Java application.
4576 The JAR files are typically located in dir:{/usr/share/java}.
4578 IMPORTANT: The LTTng-UST Java agent must be
4579 <<installing-lttng,installed>> for the logging framework your
4582 .Use the LTTng-UST Java agent for Apache log4j.
4587 import org.apache.log4j.Appender;
4588 import org.apache.log4j.Logger;
4589 import org.lttng.ust.agent.log4j.LttngLogAppender;
4593 private static final int answer = 42;
4595 public static void main(String[] argv) throws Exception
4598 Logger logger = Logger.getLogger("jello");
4600 // Create an LTTng-UST log appender
4601 Appender lttngUstLogAppender = new LttngLogAppender();
4603 // Add the LTTng-UST log appender to our logger
4604 logger.addAppender(lttngUstLogAppender);
4607 logger.info("some info");
4608 logger.warn("some warning");
4610 logger.debug("debug information; the answer is " + answer);
4612 logger.fatal("error!");
4614 // Not mandatory, but cleaner
4615 logger.removeAppender(lttngUstLogAppender);
4616 lttngUstLogAppender.close();
4622 Build this example (`$LOG4JPATH` is the path to the Apache log4j JAR
4627 javac -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-log4j.jar:$LOG4JPATH Test.java
4630 <<creating-destroying-tracing-sessions,Create a tracing session>>,
4631 <<enabling-disabling-events,create an event rule>> matching the
4632 `jello` log4j logger, and <<basic-tracing-session-control,start tracing>>:
4637 lttng enable-event --log4j jello
4641 Run the compiled class:
4645 java -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-log4j.jar:$LOG4JPATH:. Test
4648 <<basic-tracing-session-control,Stop tracing>> and inspect the
4658 You can use the opt:lttng-enable-event(1):--loglevel or
4659 opt:lttng-enable-event(1):--loglevel-only option of the
4660 man:lttng-enable-event(1) command to target a range of Apache log4j log levels
4661 or a specific log4j log level.
4665 [[java-application-context]]
4666 ==== Provide application-specific context fields in a Java application
4668 A Java application-specific context field is a piece of state provided
4669 by the application which <<adding-context,you can add>>, using the
4670 man:lttng-add-context(1) command, to each <<event,event record>>
4671 produced by the log statements of this application.
4673 For example, a given object might have a current request ID variable.
4674 You can create a context information retriever for this object and
4675 assign a name to this current request ID. You can then, using the
4676 man:lttng-add-context(1) command, add this context field by name to
4677 the JUL or log4j <<channel,channel>>.
4679 To provide application-specific context fields in a Java application:
4681 . In the Java application's source code, import the LTTng-UST
4682 Java agent context classes and interfaces:
4687 import org.lttng.ust.agent.context.ContextInfoManager;
4688 import org.lttng.ust.agent.context.IContextInfoRetriever;
4692 . Create a context information retriever class, that is, a class which
4693 implements the `IContextInfoRetriever` interface:
4698 class MyContextInfoRetriever implements IContextInfoRetriever
4701 public Object retrieveContextInfo(String key)
4703 if (key.equals("intCtx")) {
4705 } else if (key.equals("strContext")) {
4706 return "context value!";
4715 This `retrieveContextInfo()` method is the only member of the
4716 `IContextInfoRetriever` interface. Its role is to return the current
4717 value of a state by name to create a context field. The names of the
4718 context fields and which state variables they return depends on your
4721 All primitive types and objects are supported as context fields.
4722 When `retrieveContextInfo()` returns an object, the context field
4723 serializer calls its `toString()` method to add a string field to
4724 event records. The method can also return `null`, which means that
4725 no context field is available for the required name.
4727 . Register an instance of your context information retriever class to
4728 the context information manager singleton:
4733 IContextInfoRetriever cir = new MyContextInfoRetriever();
4734 ContextInfoManager cim = ContextInfoManager.getInstance();
4735 cim.registerContextInfoRetriever("retrieverName", cir);
4739 . Before exiting the application, remove your context information
4740 retriever from the context information manager singleton:
4745 ContextInfoManager cim = ContextInfoManager.getInstance();
4746 cim.unregisterContextInfoRetriever("retrieverName");
4750 This is not strictly necessary, but it is recommended for a clean
4751 disposal of some manager's resources.
4753 . Build your Java application with LTTng-UST Java agent support as
4754 usual, following the procedure for either the <<jul,JUL>> or
4755 <<log4j,Apache log4j>> framework.
4758 .Provide application-specific context fields in a Java application.
4763 import java.util.logging.Handler;
4764 import java.util.logging.Logger;
4765 import org.lttng.ust.agent.jul.LttngLogHandler;
4766 import org.lttng.ust.agent.context.ContextInfoManager;
4767 import org.lttng.ust.agent.context.IContextInfoRetriever;
4771 // Our context information retriever class
4772 private static class MyContextInfoRetriever
4773 implements IContextInfoRetriever
4776 public Object retrieveContextInfo(String key) {
4777 if (key.equals("intCtx")) {
4779 } else if (key.equals("strContext")) {
4780 return "context value!";
4787 private static final int answer = 42;
4789 public static void main(String args[]) throws Exception
4791 // Get the context information manager instance
4792 ContextInfoManager cim = ContextInfoManager.getInstance();
4794 // Create and register our context information retriever
4795 IContextInfoRetriever cir = new MyContextInfoRetriever();
4796 cim.registerContextInfoRetriever("myRetriever", cir);
4799 Logger logger = Logger.getLogger("jello");
4801 // Create an LTTng-UST log handler
4802 Handler lttngUstLogHandler = new LttngLogHandler();
4804 // Add the LTTng-UST log handler to our logger
4805 logger.addHandler(lttngUstLogHandler);
4808 logger.info("some info");
4809 logger.warning("some warning");
4811 logger.finer("finer information; the answer is " + answer);
4813 logger.severe("error!");
4815 // Not mandatory, but cleaner
4816 logger.removeHandler(lttngUstLogHandler);
4817 lttngUstLogHandler.close();
4818 cim.unregisterContextInfoRetriever("myRetriever");
4827 javac -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-jul.jar Test.java
4830 <<creating-destroying-tracing-sessions,Create a tracing session>>
4831 and <<enabling-disabling-events,create an event rule>> matching the
4837 lttng enable-event --jul jello
4840 <<adding-context,Add the application-specific context fields>> to the
4845 lttng add-context --jul --type='$app.myRetriever:intCtx'
4846 lttng add-context --jul --type='$app.myRetriever:strContext'
4849 <<basic-tracing-session-control,Start tracing>>:
4856 Run the compiled class:
4860 java -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-jul.jar:. Test
4863 <<basic-tracing-session-control,Stop tracing>> and inspect the
4875 [[python-application]]
4876 === User space Python agent
4878 You can instrument a Python 2 or Python 3 application which uses the
4879 standard https://docs.python.org/3/library/logging.html[`logging`]
4882 Each log statement emits an LTTng event once the
4883 application module imports the
4884 <<lttng-ust-agents,LTTng-UST Python agent>> package.
4887 .A Python application importing the LTTng-UST Python agent.
4888 image::python-app.png[]
4890 To use the LTTng-UST Python agent:
4892 . In the Python application's source code, import the LTTng-UST Python
4902 The LTTng-UST Python agent automatically adds its logging handler to the
4903 root logger at import time.
4905 Any log statement that the application executes before this import does
4906 not emit an LTTng event.
4908 IMPORTANT: The LTTng-UST Python agent must be
4909 <<installing-lttng,installed>>.
4911 . Use log statements and logging configuration as usual.
4912 Since the LTTng-UST Python agent adds a handler to the _root_
4913 logger, you can trace any log statement from any logger.
4915 .Use the LTTng-UST Python agent.
4926 logging.basicConfig()
4927 logger = logging.getLogger('my-logger')
4930 logger.debug('debug message')
4931 logger.info('info message')
4932 logger.warn('warn message')
4933 logger.error('error message')
4934 logger.critical('critical message')
4938 if __name__ == '__main__':
4942 NOTE: `logging.basicConfig()`, which adds to the root logger a basic
4943 logging handler which prints to the standard error stream, is not
4944 strictly required for LTTng-UST tracing to work, but in versions of
4945 Python preceding 3.2, you could see a warning message which indicates
4946 that no handler exists for the logger `my-logger`.
4948 <<creating-destroying-tracing-sessions,Create a tracing session>>,
4949 <<enabling-disabling-events,create an event rule>> matching the
4950 `my-logger` Python logger, and <<basic-tracing-session-control,start
4956 lttng enable-event --python my-logger
4960 Run the Python script:
4967 <<basic-tracing-session-control,Stop tracing>> and inspect the recorded
4977 You can use the opt:lttng-enable-event(1):--loglevel or
4978 opt:lttng-enable-event(1):--loglevel-only option of the
4979 man:lttng-enable-event(1) command to target a range of Python log levels
4980 or a specific Python log level.
4982 When an application imports the LTTng-UST Python agent, the agent tries
4983 to register to a <<lttng-sessiond,session daemon>>. Note that you must
4984 <<start-sessiond,start the session daemon>> _before_ you run the Python
4985 application. If a session daemon is found, the agent tries to register
4986 to it during 5{nbsp}seconds, after which the application continues
4987 without LTTng tracing support. You can override this timeout value with
4988 the env:LTTNG_UST_PYTHON_REGISTER_TIMEOUT environment variable
4991 If the session daemon stops while a Python application with an imported
4992 LTTng-UST Python agent runs, the agent retries to connect and to
4993 register to a session daemon every 3{nbsp}seconds. You can override this
4994 delay with the env:LTTNG_UST_PYTHON_REGISTER_RETRY_DELAY environment
4999 [[proc-lttng-logger-abi]]
5002 The `lttng-tracer` Linux kernel module, part of
5003 <<lttng-modules,LTTng-modules>>, creates the special LTTng logger file
5004 path:{/proc/lttng-logger} when it's loaded. Any application can write
5005 text data to this file to emit an LTTng event.
5008 .An application writes to the LTTng logger file to emit an LTTng event.
5009 image::lttng-logger.png[]
5011 The LTTng logger is the quickest method--not the most efficient,
5012 however--to add instrumentation to an application. It is designed
5013 mostly to instrument shell scripts:
5017 echo "Some message, some $variable" > /proc/lttng-logger
5020 Any event that the LTTng logger emits is named `lttng_logger` and
5021 belongs to the Linux kernel <<domain,tracing domain>>. However, unlike
5022 other instrumentation points in the kernel tracing domain, **any Unix
5023 user** can <<enabling-disabling-events,create an event rule>> which
5024 matches its event name, not only the root user or users in the tracing
5027 To use the LTTng logger:
5029 * From any application, write text data to the path:{/proc/lttng-logger}
5032 The `msg` field of `lttng_logger` event records contains the
5035 NOTE: The maximum message length of an LTTng logger event is
5036 1024{nbsp}bytes. Writing more than this makes the LTTng logger emit more
5037 than one event to contain the remaining data.
5039 You should not use the LTTng logger to trace a user application which
5040 can be instrumented in a more efficient way, namely:
5042 * <<c-application,C and $$C++$$ applications>>.
5043 * <<java-application,Java applications>>.
5044 * <<python-application,Python applications>>.
5046 .Use the LTTng logger.
5051 echo 'Hello, World!' > /proc/lttng-logger
5053 df --human-readable --print-type / > /proc/lttng-logger
5056 <<creating-destroying-tracing-sessions,Create a tracing session>>,
5057 <<enabling-disabling-events,create an event rule>> matching the
5058 `lttng_logger` Linux kernel tracepoint, and
5059 <<basic-tracing-session-control,start tracing>>:
5064 lttng enable-event --kernel lttng_logger
5068 Run the Bash script:
5075 <<basic-tracing-session-control,Stop tracing>> and inspect the recorded
5086 [[instrumenting-linux-kernel]]
5087 === LTTng kernel tracepoints
5089 NOTE: This section shows how to _add_ instrumentation points to the
5090 Linux kernel. The kernel's subsystems are already thoroughly
5091 instrumented at strategic places for LTTng when you
5092 <<installing-lttng,install>> the <<lttng-modules,LTTng-modules>>
5096 There are two methods to instrument the Linux kernel:
5098 . <<linux-add-lttng-layer,Add an LTTng layer>> over an existing ftrace
5099 tracepoint which uses the `TRACE_EVENT()` API.
5101 Choose this if you want to instrumentation a Linux kernel tree with an
5102 instrumentation point compatible with ftrace, perf, and SystemTap.
5104 . Use an <<linux-lttng-tracepoint-event,LTTng-only approach>> to
5105 instrument an out-of-tree kernel module.
5107 Choose this if you don't need ftrace, perf, or SystemTap support.
5111 [[linux-add-lttng-layer]]
5112 ==== [[instrumenting-linux-kernel-itself]][[mainline-trace-event]][[lttng-adaptation-layer]]Add an LTTng layer to an existing ftrace tracepoint
5114 This section shows how to add an LTTng layer to existing ftrace
5115 instrumentation using the `TRACE_EVENT()` API.
5117 This section does not document the `TRACE_EVENT()` macro. You can
5118 read the following articles to learn more about this API:
5120 * http://lwn.net/Articles/379903/[Using the TRACE_EVENT() macro (Part 1)]
5121 * http://lwn.net/Articles/381064/[Using the TRACE_EVENT() macro (Part 2)]
5122 * http://lwn.net/Articles/383362/[Using the TRACE_EVENT() macro (Part 3)]
5124 The following procedure assumes that your ftrace tracepoints are
5125 correctly defined in their own header and that they are created in
5126 one source file using the `CREATE_TRACE_POINTS` definition.
5128 To add an LTTng layer over an existing ftrace tracepoint:
5130 . Make sure the following kernel configuration options are
5136 * `CONFIG_HIGH_RES_TIMERS`
5137 * `CONFIG_TRACEPOINTS`
5140 . Build the Linux source tree with your custom ftrace tracepoints.
5141 . Boot the resulting Linux image on your target system.
5143 Confirm that the tracepoints exist by looking for their names in the
5144 dir:{/sys/kernel/debug/tracing/events/subsys} directory, where `subsys`
5145 is your subsystem's name.
5147 . Get a copy of the latest LTTng-modules{nbsp}{revision}:
5153 wget http://lttng.org/files/lttng-modules/lttng-modules-latest-2.8.tar.bz2 &&
5154 tar -xf lttng-modules-latest-2.8.tar.bz2 &&
5155 cd lttng-modules-2.8.*
5159 . In dir:{instrumentation/events/lttng-module}, relative to the root
5160 of the LTTng-modules source tree, create a header file named
5161 +__subsys__.h+ for your custom subsystem +__subsys__+ and write your
5162 LTTng-modules tracepoint definitions using the LTTng-modules
5165 Start with this template:
5169 .path:{instrumentation/events/lttng-module/my_subsys.h}
5172 #define TRACE_SYSTEM my_subsys
5174 #if !defined(_LTTNG_MY_SUBSYS_H) || defined(TRACE_HEADER_MULTI_READ)
5175 #define _LTTNG_MY_SUBSYS_H
5177 #include "../../../probes/lttng-tracepoint-event.h"
5178 #include <linux/tracepoint.h>
5180 LTTNG_TRACEPOINT_EVENT(
5182 * Format is identical to TRACE_EVENT()'s version for the three
5183 * following macro parameters:
5186 TP_PROTO(int my_int, const char *my_string),
5187 TP_ARGS(my_int, my_string),
5189 /* LTTng-modules specific macros */
5191 ctf_integer(int, my_int_field, my_int)
5192 ctf_string(my_bar_field, my_bar)
5196 #endif /* !defined(_LTTNG_MY_SUBSYS_H) || defined(TRACE_HEADER_MULTI_READ) */
5198 #include "../../../probes/define_trace.h"
5202 The entries in the `TP_FIELDS()` section are the list of fields for the
5203 LTTng tracepoint. This is similar to the `TP_STRUCT__entry()` part of
5204 ftrace's `TRACE_EVENT()` macro.
5206 See <<lttng-modules-tp-fields,Tracepoint fields macros>> for a
5207 complete description of the available `ctf_*()` macros.
5209 . Create the LTTng-modules probe's kernel module C source file,
5210 +probes/lttng-probe-__subsys__.c+, where +__subsys__+ is your
5215 .path:{probes/lttng-probe-my-subsys.c}
5217 #include <linux/module.h>
5218 #include "../lttng-tracer.h"
5221 * Build-time verification of mismatch between mainline
5222 * TRACE_EVENT() arguments and the LTTng-modules adaptation
5223 * layer LTTNG_TRACEPOINT_EVENT() arguments.
5225 #include <trace/events/my_subsys.h>
5227 /* Create LTTng tracepoint probes */
5228 #define LTTNG_PACKAGE_BUILD
5229 #define CREATE_TRACE_POINTS
5230 #define TRACE_INCLUDE_PATH ../instrumentation/events/lttng-module
5232 #include "../instrumentation/events/lttng-module/my_subsys.h"
5234 MODULE_LICENSE("GPL and additional rights");
5235 MODULE_AUTHOR("Your name <your-email>");
5236 MODULE_DESCRIPTION("LTTng my_subsys probes");
5237 MODULE_VERSION(__stringify(LTTNG_MODULES_MAJOR_VERSION) "."
5238 __stringify(LTTNG_MODULES_MINOR_VERSION) "."
5239 __stringify(LTTNG_MODULES_PATCHLEVEL_VERSION)
5240 LTTNG_MODULES_EXTRAVERSION);
5244 . Edit path:{probes/Makefile} and add your new kernel module object
5245 next to the existing ones:
5249 .path:{probes/Makefile}
5253 obj-m += lttng-probe-module.o
5254 obj-m += lttng-probe-power.o
5256 obj-m += lttng-probe-my-subsys.o
5262 . Build and install the LTTng kernel modules:
5267 make KERNELDIR=/path/to/linux
5268 sudo make modules_install
5272 Replace `/path/to/linux` with the path to the Linux source tree where
5273 you defined and used tracepoints with ftrace's `TRACE_EVENT()` macro.
5275 Note that you can also use the
5276 <<lttng-tracepoint-event-code,`LTTNG_TRACEPOINT_EVENT_CODE()` macro>>
5277 instead of `LTTNG_TRACEPOINT_EVENT()` to use custom local variables and
5278 C code that need to be executed before the event fields are recorded.
5280 The best way to learn how to use the previous LTTng-modules macros is to
5281 inspect the existing LTTng-modules tracepoint definitions in the
5282 dir:{instrumentation/events/lttng-module} header files. Compare them
5283 with the Linux kernel mainline versions in the
5284 dir:{include/trace/events} directory of the Linux source tree.
5288 [[lttng-tracepoint-event-code]]
5289 ===== Use custom C code to access the data for tracepoint fields
5291 Although we recommended to always use the
5292 <<lttng-adaptation-layer,`LTTNG_TRACEPOINT_EVENT()`>> macro to describe
5293 the arguments and fields of an LTTng-modules tracepoint when possible,
5294 sometimes you need a more complex process to access the data that the
5295 tracer records as event record fields. In other words, you need local
5296 variables and multiple C{nbsp}statements instead of simple
5297 argument-based expressions that you pass to the
5298 <<lttng-modules-tp-fields,`ctf_*()` macros of `TP_FIELDS()`>>.
5300 You can use the `LTTNG_TRACEPOINT_EVENT_CODE()` macro instead of
5301 `LTTNG_TRACEPOINT_EVENT()` to declare custom local variables and define
5302 a block of C{nbsp}code to be executed before LTTng records the fields.
5303 The structure of this macro is:
5306 .`LTTNG_TRACEPOINT_EVENT_CODE()` macro syntax.
5308 LTTNG_TRACEPOINT_EVENT_CODE(
5310 * Format identical to the LTTNG_TRACEPOINT_EVENT()
5311 * version for the following three macro parameters:
5314 TP_PROTO(int my_int, const char *my_string),
5315 TP_ARGS(my_int, my_string),
5317 /* Declarations of custom local variables */
5320 unsigned long b = 0;
5321 const char *name = "(undefined)";
5322 struct my_struct *my_struct;
5326 * Custom code which uses both tracepoint arguments
5327 * (in TP_ARGS()) and local variables (in TP_locvar()).
5329 * Local variables are actually members of a structure pointed
5330 * to by the special variable tp_locvar.
5334 tp_locvar->a = my_int + 17;
5335 tp_locvar->my_struct = get_my_struct_at(tp_locvar->a);
5336 tp_locvar->b = my_struct_compute_b(tp_locvar->my_struct);
5337 tp_locvar->name = my_struct_get_name(tp_locvar->my_struct);
5338 put_my_struct(tp_locvar->my_struct);
5347 * Format identical to the LTTNG_TRACEPOINT_EVENT()
5348 * version for this, except that tp_locvar members can be
5349 * used in the argument expression parameters of
5350 * the ctf_*() macros.
5353 ctf_integer(unsigned long, my_struct_b, tp_locvar->b)
5354 ctf_integer(int, my_struct_a, tp_locvar->a)
5355 ctf_string(my_string_field, my_string)
5356 ctf_string(my_struct_name, tp_locvar->name)
5361 IMPORTANT: The C code defined in `TP_code()` must not have any side
5362 effects when executed. In particular, the code must not allocate
5363 memory or get resources without deallocating this memory or putting
5364 those resources afterwards.
5367 [[instrumenting-linux-kernel-tracing]]
5368 ==== Load and unload a custom probe kernel module
5370 You must load a <<lttng-adaptation-layer,created LTTng-modules probe
5371 kernel module>> in the kernel before it can emit LTTng events.
5373 To load the default probe kernel modules and a custom probe kernel
5376 * Use the opt:lttng-sessiond(8):--extra-kmod-probes option to give extra
5377 probe modules to load when starting a root <<lttng-sessiond,session
5381 .Load the `my_subsys`, `usb`, and the default probe modules.
5385 sudo lttng-sessiond --extra-kmod-probes=my_subsys,usb
5390 You only need to pass the subsystem name, not the whole kernel module
5393 To load _only_ a given custom probe kernel module:
5395 * Use the opt:lttng-sessiond(8):--kmod-probes option to give the probe
5396 modules to load when starting a root session daemon:
5399 .Load only the `my_subsys` and `usb` probe modules.
5403 sudo lttng-sessiond --kmod-probes=my_subsys,usb
5408 To confirm that a probe module is loaded:
5415 lsmod | grep lttng_probe_usb
5419 To unload the loaded probe modules:
5421 * Kill the session daemon with `SIGTERM`:
5426 sudo pkill lttng-sessiond
5430 You can also use man:modprobe(8)'s `--remove` option if the session
5431 daemon terminates abnormally.
5434 [[controlling-tracing]]
5437 Once an application or a Linux kernel is
5438 <<instrumenting,instrumented>> for LTTng tracing,
5441 This section is divided in topics on how to use the various
5442 <<plumbing,components of LTTng>>, in particular the <<lttng-cli,cmd:lttng
5443 command-line tool>>, to _control_ the LTTng daemons and tracers.
5445 NOTE: In the following subsections, we refer to an man:lttng(1) command
5446 using its man page name. For example, instead of _Run the `create`
5447 command to..._, we use _Run the man:lttng-create(1) command to..._.
5451 === Start a session daemon
5453 In some situations, you need to run a <<lttng-sessiond,session daemon>>
5454 (man:lttng-sessiond(8)) _before_ you can use the man:lttng(1)
5457 You will see the following error when you run a command while no session
5461 Error: No session daemon is available
5464 The only command that automatically runs a session daemon is
5465 man:lttng-create(1), which you use to
5466 <<creating-destroying-tracing-sessions,create a tracing session>>. While
5467 this is most of the time the first operation that you do, sometimes it's
5468 not. Some examples are:
5470 * <<list-instrumentation-points,List the available instrumentation points>>.
5471 * <<saving-loading-tracing-session,Load a tracing session configuration>>.
5473 [[tracing-group]] Each Unix user must have its own running session
5474 daemon to trace user applications. The session daemon that the root user
5475 starts is the only one allowed to control the LTTng kernel tracer. Users
5476 that are part of the _tracing group_ can control the root session
5477 daemon. The default tracing group name is `tracing`; you can set it to
5478 something else with the opt:lttng-sessiond(8):--group option when you
5479 start the root session daemon.
5481 To start a user session daemon:
5483 * Run man:lttng-sessiond(8):
5488 lttng-sessiond --daemonize
5492 To start the root session daemon:
5494 * Run man:lttng-sessiond(8) as the root user:
5499 sudo lttng-sessiond --daemonize
5503 In both cases, remove the opt:lttng-sessiond(8):--daemonize option to
5504 start the session daemon in foreground.
5506 To stop a session daemon, use man:kill(1) on its process ID (standard
5509 Note that some Linux distributions could manage the LTTng session daemon
5510 as a service. In this case, you should use the service manager to
5511 start, restart, and stop session daemons.
5514 [[creating-destroying-tracing-sessions]]
5515 === Create and destroy a tracing session
5517 Almost all the LTTng control operations happen in the scope of
5518 a <<tracing-session,tracing session>>, which is the dialogue between the
5519 <<lttng-sessiond,session daemon>> and you.
5521 To create a tracing session with a generated name:
5523 * Use the man:lttng-create(1) command:
5532 The created tracing session's name is `auto` followed by the
5535 To create a tracing session with a specific name:
5537 * Use the optional argument of the man:lttng-create(1) command:
5542 lttng create my-session
5546 Replace `my-session` with the specific tracing session name.
5548 LTTng appends the creation date to the created tracing session's name.
5550 LTTng writes the traces of a tracing session in
5551 +$LTTNG_HOME/lttng-trace/__name__+ by default, where +__name__+ is the
5552 name of the tracing session. Note that the env:LTTNG_HOME environment
5553 variable defaults to `$HOME` if not set.
5555 To output LTTng traces to a non-default location:
5557 * Use the opt:lttng-create(1):--output option of the man:lttng-create(1) command:
5562 lttng create --output=/tmp/some-directory my-session
5566 You may create as many tracing sessions as you wish.
5568 To list all the existing tracing sessions for your Unix user:
5570 * Use the man:lttng-list(1) command:
5579 When you create a tracing session, it is set as the _current tracing
5580 session_. The following man:lttng(1) commands operate on the current
5581 tracing session when you don't specify one:
5583 [role="list-3-cols"]
5599 To change the current tracing session:
5601 * Use the man:lttng-set-session(1) command:
5606 lttng set-session new-session
5610 Replace `new-session` by the name of the new current tracing session.
5612 When you are done tracing in a given tracing session, you can destroy
5613 it. This operation frees the resources taken by the tracing session
5614 to destroy; it does not destroy the trace data that LTTng wrote for
5615 this tracing session.
5617 To destroy the current tracing session:
5619 * Use the man:lttng-destroy(1) command:
5629 [[list-instrumentation-points]]
5630 === List the available instrumentation points
5632 The <<lttng-sessiond,session daemon>> can query the running instrumented
5633 user applications and the Linux kernel to get a list of available
5634 instrumentation points. For the Linux kernel <<domain,tracing domain>>,
5635 they are tracepoints and system calls. For the user space tracing
5636 domain, they are tracepoints. For the other tracing domains, they are
5639 To list the available instrumentation points:
5641 * Use the man:lttng-list(1) command with the requested tracing domain's
5645 * opt:lttng-list(1):--kernel: Linux kernel tracepoints (your Unix user
5646 must be a root user, or it must be a member of the
5647 <<tracing-group,tracing group>>).
5648 * opt:lttng-list(1):--kernel with opt:lttng-list(1):--syscall: Linux
5649 kernel system calls (your Unix user must be a root user, or it must be
5650 a member of the tracing group).
5651 * opt:lttng-list(1):--userspace: user space tracepoints.
5652 * opt:lttng-list(1):--jul: `java.util.logging` loggers.
5653 * opt:lttng-list(1):--log4j: Apache log4j loggers.
5654 * opt:lttng-list(1):--python: Python loggers.
5657 .List the available user space tracepoints.
5661 lttng list --userspace
5665 .List the available Linux kernel system call tracepoints.
5669 lttng list --kernel --syscall
5674 [[enabling-disabling-events]]
5675 === Create and enable an event rule
5677 Once you <<creating-destroying-tracing-sessions,create a tracing
5678 session>>, you can create <<event,event rules>> with the
5679 man:lttng-enable-event(1) command.
5681 You specify each condition with a command-line option. The available
5682 condition options are shown in the following table.
5684 [role="growable",cols="asciidoc,asciidoc,default"]
5685 .Condition command-line options for the man:lttng-enable-event(1) command.
5687 |Option |Description |Applicable tracing domains
5693 . +--probe=__ADDR__+
5694 . +--function=__ADDR__+
5697 Instead of using the default _tracepoint_ instrumentation type, use:
5699 . A Linux system call.
5700 . A Linux https://lwn.net/Articles/132196/[KProbe] (symbol or address).
5701 . The entry and return points of a Linux function (symbol or address).
5705 |First positional argument.
5708 Tracepoint or system call name. In the case of a Linux KProbe or
5709 function, this is a custom name given to the event rule. With the
5710 JUL, log4j, and Python domains, this is a logger name.
5712 With a tracepoint, logger, or system call name, the last character
5713 can be `*` to match anything that remains.
5720 . +--loglevel=__LEVEL__+
5721 . +--loglevel-only=__LEVEL__+
5724 . Match only tracepoints or log statements with a logging level at
5725 least as severe as +__LEVEL__+.
5726 . Match only tracepoints or log statements with a logging level
5727 equal to +__LEVEL__+.
5729 See man:lttng-enable-event(1) for the list of available logging level
5732 |User space, JUL, log4j, and Python.
5734 |+--exclude=__EXCLUSIONS__+
5737 When you use a `*` character at the end of the tracepoint or logger
5738 name (first positional argument), exclude the specific names in the
5739 comma-delimited list +__EXCLUSIONS__+.
5742 User space, JUL, log4j, and Python.
5744 |+--filter=__EXPR__+
5747 Match only events which satisfy the expression +__EXPR__+.
5749 See man:lttng-enable-event(1) to learn more about the syntax of a
5756 You attach an event rule to a <<channel,channel>> on creation. If you do
5757 not specify the channel with the opt:lttng-enable-event(1):--channel
5758 option, and if the event rule to create is the first in its
5759 <<domain,tracing domain>> for a given tracing session, then LTTng
5760 creates a _default channel_ for you. This default channel is reused in
5761 subsequent invocations of the man:lttng-enable-event(1) command for the
5762 same tracing domain.
5764 An event rule is always enabled at creation time.
5766 The following examples show how you can combine the previous
5767 command-line options to create simple to more complex event rules.
5769 .Create an event rule targetting a Linux kernel tracepoint (default channel).
5773 lttng enable-event --kernel sched_switch
5777 .Create an event rule matching four Linux kernel system calls (default channel).
5781 lttng enable-event --kernel --syscall open,write,read,close
5785 .Create event rules matching tracepoints with filter expressions (default channel).
5789 lttng enable-event --kernel sched_switch --filter='prev_comm == "bash"'
5794 lttng enable-event --kernel --all \
5795 --filter='$ctx.tid == 1988 || $ctx.tid == 1534'
5800 lttng enable-event --jul my_logger \
5801 --filter='$app.retriever:cur_msg_id > 3'
5804 IMPORTANT: Make sure to always quote the filter string when you
5805 use man:lttng(1) from a shell.
5808 .Create an event rule matching any user space tracepoint of a given tracepoint provider with a log level range (default channel).
5812 lttng enable-event --userspace my_app:'*' --loglevel=TRACE_INFO
5815 IMPORTANT: Make sure to always quote the wildcard character when you
5816 use man:lttng(1) from a shell.
5819 .Create an event rule matching multiple Python loggers with a wildcard and with exclusions (default channel).
5823 lttng enable-event --python my-app.'*' \
5824 --exclude='my-app.module,my-app.hello'
5828 .Create an event rule matching any Apache log4j logger with a specific log level (default channel).
5832 lttng enable-event --log4j --all --loglevel-only=LOG4J_WARN
5836 .Create an event rule attached to a specific channel matching a specific user space tracepoint provider and tracepoint.
5840 lttng enable-event --userspace my_app:my_tracepoint --channel=my-channel
5844 The event rules of a given channel form a whitelist: as soon as an
5845 emitted event passes one of them, LTTng can record the event. For
5846 example, an event named `my_app:my_tracepoint` emitted from a user space
5847 tracepoint with a `TRACE_ERROR` log level passes both of the following
5852 lttng enable-event --userspace my_app:my_tracepoint
5853 lttng enable-event --userspace my_app:my_tracepoint \
5854 --loglevel=TRACE_INFO
5857 The second event rule is redundant: the first one includes
5861 [[disable-event-rule]]
5862 === Disable an event rule
5864 To disable an event rule that you <<enabling-disabling-events,created>>
5865 previously, use the man:lttng-disable-event(1) command. This command
5866 disables _all_ the event rules (of a given tracing domain and channel)
5867 which match an instrumentation point. The other conditions are not
5868 supported as of LTTng{nbsp}{revision}.
5870 The LTTng tracer does not record an emitted event which passes
5871 a _disabled_ event rule.
5873 .Disable an event rule matching a Python logger (default channel).
5877 lttng disable-event --python my-logger
5881 .Disable an event rule matching all `java.util.logging` loggers (default channel).
5885 lttng disable-event --jul '*'
5889 .Disable _all_ the event rules of the default channel.
5891 The opt:lttng-disable-event(1):--all-events option is not, like the
5892 opt:lttng-enable-event(1):--all option of man:lttng-enable-event(1), the
5893 equivalent of the event name `*` (wildcard): it disables _all_ the event
5894 rules of a given channel.
5898 lttng disable-event --jul --all-events
5902 NOTE: You cannot delete an event rule once you create it.
5906 === Get the status of a tracing session
5908 To get the status of the current tracing session, that is, its
5909 parameters, its channels, event rules, and their attributes:
5911 * Use the man:lttng-status(1) command:
5921 To get the status of any tracing session:
5923 * Use the man:lttng-list(1) command with the tracing session's name:
5928 lttng list my-session
5932 Replace `my-session` with the desired tracing session's name.
5935 [[basic-tracing-session-control]]
5936 === Start and stop a tracing session
5938 Once you <<creating-destroying-tracing-sessions,create a tracing
5940 <<enabling-disabling-events,create one or more event rules>>,
5941 you can start and stop the tracers for this tracing session.
5943 To start tracing in the current tracing session:
5945 * Use the man:lttng-start(1) command:
5954 LTTng is very flexible: you can launch user applications before
5955 or after the you start the tracers. The tracers only record the events
5956 if they pass enabled event rules and if they occur while the tracers are
5959 To stop tracing in the current tracing session:
5961 * Use the man:lttng-stop(1) command:
5970 If there were <<channel-overwrite-mode-vs-discard-mode,lost event
5971 records>> or lost sub-buffers since the last time you ran
5972 man:lttng-start(1), warnings are printed when you run the
5973 man:lttng-stop(1) command.
5976 [[enabling-disabling-channels]]
5977 === Create a channel
5979 Once you create a tracing session, you can create a <<channel,channel>>
5980 with the man:lttng-enable-channel(1) command.
5982 Note that LTTng automatically creates a default channel when, for a
5983 given <<domain,tracing domain>>, no channels exist and you
5984 <<enabling-disabling-events,create>> the first event rule. This default
5985 channel is named `channel0` and its attributes are set to reasonable
5986 values. Therefore, you only need to create a channel when you need
5987 non-default attributes.
5989 You specify each non-default channel attribute with a command-line
5990 option when you use the man:lttng-enable-channel(1) command. The
5991 available command-line options are:
5993 [role="growable",cols="asciidoc,asciidoc"]
5994 .Command-line options for the man:lttng-enable-channel(1) command.
5996 |Option |Description
6002 <<channel-overwrite-mode-vs-discard-mode,event loss mode>> instead of
6003 the default _discard_ mode.
6005 |`--buffers-pid` (user space tracing domain only)
6008 Use the per-process <<channel-buffering-schemes,buffering scheme>>
6009 instead of the default per-user buffering scheme.
6011 |+--subbuf-size=__SIZE__+
6014 Allocate sub-buffers of +__SIZE__+ bytes (power of two), for each CPU,
6015 either for each Unix user (default), or for each instrumented process.
6017 See <<channel-subbuf-size-vs-subbuf-count,Sub-buffer count and size>>.
6019 |+--num-subbuf=__COUNT__+
6022 Allocate +__COUNT__+ sub-buffers (power of two), for each CPU, either
6023 for each Unix user (default), or for each instrumented process.
6025 See <<channel-subbuf-size-vs-subbuf-count,Sub-buffer count and size>>.
6027 |+--tracefile-size=__SIZE__+
6030 Set the maximum size of each trace file that this channel writes within
6031 a stream to +__SIZE__+ bytes instead of no maximum.
6033 See <<tracefile-rotation,Trace file count and size>>.
6035 |+--tracefile-count=__COUNT__+
6038 Limit the number of trace files that this channel creates to
6039 +__COUNT__+ channels instead of no limit.
6041 See <<tracefile-rotation,Trace file count and size>>.
6043 |+--switch-timer=__PERIODUS__+
6046 Set the <<channel-switch-timer,switch timer period>>
6047 to +__PERIODUS__+{nbsp}µs.
6049 |+--read-timer=__PERIODUS__+
6052 Set the <<channel-read-timer,read timer period>>
6053 to +__PERIODUS__+{nbsp}µs.
6055 |+--output=__TYPE__+ (Linux kernel tracing domain only)
6058 Set the channel's output type to +__TYPE__+, either `mmap` or `splice`.
6062 You can only create a channel in the Linux kernel and user space
6063 <<domain,tracing domains>>: other tracing domains have their own channel
6064 created on the fly when <<enabling-disabling-events,creating event
6069 Because of a current LTTng limitation, you must create all channels
6070 _before_ you <<basic-tracing-session-control,start tracing>> in a given
6071 tracing session, that is, before the first time you run
6074 Since LTTng automatically creates a default channel when you use the
6075 man:lttng-enable-event(1) command with a specific tracing domain, you
6076 cannot, for example, create a Linux kernel event rule, start tracing,
6077 and then create a user space event rule, because no user space channel
6078 exists yet and it's too late to create one.
6080 For this reason, make sure to configure your channels properly
6081 before starting the tracers for the first time!
6084 The following examples show how you can combine the previous
6085 command-line options to create simple to more complex channels.
6087 .Create a Linux kernel channel with default attributes.
6091 lttng enable-channel --kernel my-channel
6095 .Create a user space channel with 4 sub-buffers or 1{nbsp}MiB each, per CPU, per instrumented process.
6099 lttng enable-channel --userspace --num-subbuf=4 --subbuf-size=1M \
6100 --buffers-pid my-channel
6104 .Create a Linux kernel channel which rotates 8 trace files of 4{nbsp}MiB each for each stream
6108 lttng enable-channel --kernel --tracefile-count=8 \
6109 --tracefile-size=4194304 my-channel
6113 .Create a user space channel in overwrite (or _flight recorder_) mode.
6117 lttng enable-channel --userspace --overwrite my-channel
6121 You can <<enabling-disabling-events,create>> the same event rule in
6122 two different channels:
6126 lttng enable-event --userspace --channel=my-channel app:tp
6127 lttng enable-event --userspace --channel=other-channel app:tp
6130 If both channels are enabled, when a tracepoint named `app:tp` is
6131 reached, LTTng records two events, one for each channel.
6135 === Disable a channel
6137 To disable a specific channel that you <<enabling-disabling-channels,created>>
6138 previously, use the man:lttng-disable-channel(1) command.
6140 .Disable a specific Linux kernel channel.
6144 lttng disable-channel --kernel my-channel
6148 The state of a channel precedes the individual states of event rules
6149 attached to it: event rules which belong to a disabled channel, even if
6150 they are enabled, are also considered disabled.
6154 === Add context fields to a channel
6156 Event record fields in trace files provide important information about
6157 events that occured previously, but sometimes some external context may
6158 help you solve a problem faster. Examples of context fields are:
6160 * The **process ID**, **thread ID**, **process name**, and
6161 **process priority** of the thread in which the event occurs.
6162 * The **hostname** of the system on which the event occurs.
6163 * The current values of many possible **performance counters** using
6165 ** CPU cycles, stalled cycles, idle cycles, and the other cycle types.
6167 ** Branch instructions, misses, and loads.
6169 * Any context defined at the application level (supported for the
6170 JUL and log4j <<domain,tracing domains>>).
6172 To get the full list of available context fields, see
6173 `lttng add-context --list`. Some context fields are reserved for a
6174 specific <<domain,tracing domain>> (Linux kernel or user space).
6176 You add context fields to <<channel,channels>>. All the events
6177 that a channel with added context fields records contain those fields.
6179 To add context fields to one or all the channels of a given tracing
6182 * Use the man:lttng-add-context(1) command.
6184 .Add context fields to all the channels of the current tracing session.
6186 The following command line adds the virtual process identifier and
6187 the per-thread CPU cycles count fields to all the user space channels
6188 of the current tracing session.
6192 lttng add-context --userspace --type=vpid --type=perf:thread:cpu-cycles
6196 .Add a context field to a specific channel.
6198 The following command line adds the thread identifier context field
6199 to the Linux kernel channel named `my-channel` in the current
6204 lttng add-context --kernel --channel=my-channel --type=tid
6208 .Add an application-specific context field to a specific channel.
6210 The following command line adds the `cur_msg_id` context field of the
6211 `retriever` context retriever for all the instrumented
6212 <<java-application,Java applications>> recording <<event,event records>>
6213 in the channel named `my-channel`:
6217 lttng add-context --kernel --channel=my-channel \
6218 --type='$app:retriever:cur_msg_id'
6221 IMPORTANT: Make sure to always quote the `$` character when you
6222 use man:lttng-add-context(1) from a shell.
6225 NOTE: You cannot remove context fields from a channel once you add it.
6230 === Track process IDs
6232 It's often useful to allow only specific process IDs (PIDs) to emit
6233 events. For example, you may wish to record all the system calls made by
6234 a given process (à la http://linux.die.net/man/1/strace[strace]).
6236 The man:lttng-track(1) and man:lttng-untrack(1) commands serve this
6237 purpose. Both commands operate on a whitelist of process IDs. You _add_
6238 entries to this whitelist with the man:lttng-track(1) command and remove
6239 entries with the man:lttng-untrack(1) command. Any process which has one
6240 of the PIDs in the whitelist is allowed to emit LTTng events which pass
6241 an enabled <<event,event rule>>.
6243 NOTE: The PID tracker tracks the _numeric process IDs_. Should a
6244 process with a given tracked ID exit and another process be given this
6245 ID, then the latter would also be allowed to emit events.
6247 .Track and untrack process IDs.
6249 For the sake of the following example, assume the target system has 16
6253 <<creating-destroying-tracing-sessions,create a tracing session>>,
6254 the whitelist contains all the possible PIDs:
6257 .All PIDs are tracked.
6258 image::track-all.png[]
6260 When the whitelist is full and you use the man:lttng-track(1) command to
6261 specify some PIDs to track, LTTng first clears the whitelist, then it
6262 tracks the specific PIDs. After:
6266 lttng track --pid=3,4,7,10,13
6272 .PIDs 3, 4, 7, 10, and 13 are tracked.
6273 image::track-3-4-7-10-13.png[]
6275 You can add more PIDs to the whitelist afterwards:
6279 lttng track --pid=1,15,16
6285 .PIDs 1, 15, and 16 are added to the whitelist.
6286 image::track-1-3-4-7-10-13-15-16.png[]
6288 The man:lttng-untrack(1) command removes entries from the PID tracker's
6289 whitelist. Given the previous example, the following command:
6293 lttng untrack --pid=3,7,10,13
6296 leads to this whitelist:
6299 .PIDs 3, 7, 10, and 13 are removed from the whitelist.
6300 image::track-1-4-15-16.png[]
6302 LTTng can track all possible PIDs again using the opt:track(1):--all
6307 lttng track --pid --all
6310 The result is, again:
6313 .All PIDs are tracked.
6314 image::track-all.png[]
6317 .Track only specific PIDs
6319 A very typical use case with PID tracking is to start with an empty
6320 whitelist, then <<basic-tracing-session-control,start the tracers>>, and
6321 then add PIDs manually while tracers are active. You can accomplish this
6322 by using the opt:lttng-untrack(1):--all option of the
6323 man:lttng-untrack(1) command to clear the whitelist after you
6324 <<creating-destroying-tracing-sessions,create a tracing session>>:
6328 lttng untrack --pid --all
6334 .No PIDs are tracked.
6335 image::untrack-all.png[]
6337 If you trace with this whitelist configuration, the tracer records no
6338 events for this <<domain,tracing domain>> because no processes are
6339 tracked. You can use the man:lttng-track(1) command as usual to track
6340 specific PIDs, for example:
6344 lttng track --pid=6,11
6350 .PIDs 6 and 11 are tracked.
6351 image::track-6-11.png[]
6356 [[saving-loading-tracing-session]]
6357 === Save and load tracing session configurations
6359 Configuring a <<tracing-session,tracing session>> can be long. Some of
6360 the tasks involved are:
6362 * <<enabling-disabling-channels,Create channels>> with
6363 specific attributes.
6364 * <<adding-context,Add context fields>> to specific channels.
6365 * <<enabling-disabling-events,Create event rules>> with specific log
6366 level and filter conditions.
6368 If you use LTTng to solve real world problems, chances are you have to
6369 record events using the same tracing session setup over and over,
6370 modifying a few variables each time in your instrumented program
6371 or environment. To avoid constant tracing session reconfiguration,
6372 the man:lttng(1) command-line tool can save and load tracing session
6373 configurations to/from XML files.
6375 To save a given tracing session configuration:
6377 * Use the man:lttng-save(1) command:
6382 lttng save my-session
6386 Replace `my-session` with the name of the tracing session to save.
6388 LTTng saves tracing session configurations to
6389 dir:{$LTTNG_HOME/.lttng/sessions} by default. Note that the
6390 env:LTTNG_HOME environment variable defaults to `$HOME` if not set. Use
6391 the opt:lttng-save(1):--output-path option to change this destination
6394 LTTng saves all configuration parameters, for example:
6396 * The tracing session name.
6397 * The trace data output path.
6398 * The channels with their state and all their attributes.
6399 * The context fields you added to channels.
6400 * The event rules with their state, log level and filter conditions.
6402 To load a tracing session:
6404 * Use the man:lttng-load(1) command:
6409 lttng load my-session
6413 Replace `my-session` with the name of the tracing session to load.
6415 When LTTng loads a configuration, it restores your saved tracing session
6416 as if you just configured it manually.
6418 See man:lttng(1) for the complete list of command-line options. You
6419 can also save and load all many sessions at a time, and decide in which
6420 directory to output the XML files.
6423 [[sending-trace-data-over-the-network]]
6424 === Send trace data over the network
6426 LTTng can send the recorded trace data to a remote system over the
6427 network instead of writing it to the local file system.
6429 To send the trace data over the network:
6431 . On the _remote_ system (which can also be the target system),
6432 start an LTTng <<lttng-relayd,relay daemon>> (man:lttng-relayd(8)):
6441 . On the _target_ system, create a tracing session configured to
6442 send trace data over the network:
6447 lttng create my-session --set-url=net://remote-system
6451 Replace `remote-system` by the host name or IP address of the
6452 remote system. See man:lttng-create(1) for the exact URL format.
6454 . On the target system, use the man:lttng(1) command-line tool as usual.
6455 When tracing is active, the target's consumer daemon sends sub-buffers
6456 to the relay daemon running on the remote system intead of flushing
6457 them to the local file system. The relay daemon writes the received
6458 packets to the local file system.
6460 The relay daemon writes trace files to
6461 +$LTTNG_HOME/lttng-traces/__hostname__/__session__+ by default, where
6462 +__hostname__+ is the host name of the target system and +__session__+
6463 is the tracing session name. Note that the env:LTTNG_HOME environment
6464 variable defaults to `$HOME` if not set. Use the
6465 opt:lttng-relayd(8):--output option of man:lttng-relayd(8) to write
6466 trace files to another base directory.
6471 === View events as LTTng emits them (noch:{LTTng} live)
6473 LTTng live is a network protocol implemented by the <<lttng-relayd,relay
6474 daemon>> (man:lttng-relayd(8)) to allow compatible trace viewers to
6475 display events as LTTng emits them on the target system while tracing is
6478 The relay daemon creates a _tee_: it forwards the trace data to both
6479 the local file system and to connected live viewers:
6482 .The relay daemon creates a _tee_, forwarding the trace data to both trace files and a connected live viewer.
6487 . On the _target system_, create a <<tracing-session,tracing session>>
6493 lttng create --live my-session
6497 This spawns a local relay daemon.
6499 . Start the live viewer and configure it to connect to the relay
6500 daemon. For example, with http://diamon.org/babeltrace[Babeltrace]:
6505 babeltrace --input-format=lttng-live net://localhost/host/hostname/my-session
6512 * `hostname` with the host name of the target system.
6513 * `my-session` with the name of the tracing session to view.
6516 . Configure the tracing session as usual with the man:lttng(1)
6517 command-line tool, and <<basic-tracing-session-control,start tracing>>.
6519 You can list the available live tracing sessions with Babeltrace:
6523 babeltrace --input-format=lttng-live net://localhost
6526 You can start the relay daemon on another system. In this case, you need
6527 to specify the relay daemon's URL when you create the tracing session
6528 with the opt:lttng-create(1):--set-url option. You also need to replace
6529 `localhost` in the procedure above with the host name of the system on
6530 which the relay daemon is running.
6532 See man:lttng-create(1) and man:lttng-relayd(8) for the complete list of
6533 command-line options.
6537 [[taking-a-snapshot]]
6538 === Take a snapshot of the current sub-buffers of a tracing session
6540 The normal behavior of LTTng is to append full sub-buffers to growing
6541 trace data files. This is ideal to keep a full history of the events
6542 that occurred on the target system, but it can
6543 represent too much data in some situations. For example, you may wish
6544 to trace your application continuously until some critical situation
6545 happens, in which case you only need the latest few recorded
6546 events to perform the desired analysis, not multi-gigabyte trace files.
6548 With the man:lttng-snapshot(1) command, you can take a snapshot of the
6549 current sub-buffers of a given <<tracing-session,tracing session>>.
6550 LTTng can write the snapshot to the local file system or send it over
6555 . Create a tracing session in _snapshot mode_:
6560 lttng create --snapshot my-session
6564 The <<channel-overwrite-mode-vs-discard-mode,event loss mode>> of
6565 <<channel,channels>> created in this mode is automatically set to
6566 _overwrite_ (flight recorder mode).
6568 . Configure the tracing session as usual with the man:lttng(1)
6569 command-line tool, and <<basic-tracing-session-control,start tracing>>.
6571 . **Optional**: When you need to take a snapshot,
6572 <<basic-tracing-session-control,stop tracing>>.
6574 You can take a snapshot when the tracers are active, but if you stop
6575 them first, you are sure that the data in the sub-buffers does not
6576 change before you actually take the snapshot.
6583 lttng snapshot record --name=my-first-snapshot
6587 LTTng writes the current sub-buffers of all the current tracing
6588 session's channels to trace files on the local file system. Those trace
6589 files have `my-first-snapshot` in their name.
6591 There is no difference between the format of a normal trace file and the
6592 format of a snapshot: viewers of LTTng traces also support LTTng
6595 By default, LTTng writes snapshot files to the path shown by
6596 `lttng snapshot list-output`. You can change this path or decide to send
6597 snapshots over the network using either:
6599 . An output path or URL that you specify when you create the
6601 . An snapshot output path or URL that you add using
6602 `lttng snapshot add-output`
6603 . An output path or URL that you provide directly to the
6604 `lttng snapshot record` command.
6606 Method 3 overrides method 2, which overrides method 1. When you
6607 specify a URL, a relay daemon must listen on a remote system (see
6608 <<sending-trace-data-over-the-network,Send trace data over the network>>).
6613 === Use the machine interface
6615 With any command of the man:lttng(1) command-line tool, you can set the
6616 opt:lttng(1):--mi option to `xml` (before the command name) to get an
6617 XML machine interface output, for example:
6621 lttng --mi=xml enable-event --kernel --syscall open
6624 A schema definition (XSD) is
6625 https://github.com/lttng/lttng-tools/blob/stable-2.8/src/common/mi-lttng-3.0.xsd[available]
6626 to ease the integration with external tools as much as possible.
6630 [[metadata-regenerate]]
6631 === Regenerate the metadata of an LTTng trace
6633 An LTTng trace, which is a http://diamon.org/ctf[CTF] trace, has both
6634 data stream files and a metadata file. This metadata file contains,
6635 amongst other things, information about the offset of the clock sources
6636 used to timestamp <<event,event records>> when tracing.
6638 If, once a <<tracing-session,tracing session>> is
6639 <<basic-tracing-session-control,started>>, a major
6640 https://en.wikipedia.org/wiki/Network_Time_Protocol[NTP] correction
6641 happens, the trace's clock offset also needs to be updated. You
6642 can use the man:lttng-metadata(1) command to do so.
6644 The main use case of this command is to allow a system to boot with
6645 an incorrect wall time and trace it with LTTng before its wall time
6646 is corrected. Once the system is known to be in a state where its
6647 wall time is correct, it can run `lttng metadata regenerate`.
6649 To regenerate the metadata of an LTTng trace:
6651 * Use the `regenerate` action of the man:lttng-metadata(1) command:
6656 lttng metadata regenerate
6662 `lttng metadata regenerate` has the following limitations:
6664 * Tracing session <<creating-destroying-tracing-sessions,created>>
6666 * User space <<channel,channels>>, if any, using
6667 <<channel-buffering-schemes,per-user buffering>>.
6672 [[persistent-memory-file-systems]]
6673 === Record trace data on persistent memory file systems
6675 https://en.wikipedia.org/wiki/Non-volatile_random-access_memory[Non-volatile random-access memory]
6676 (NVRAM) is random-access memory that retains its information when power
6677 is turned off (non-volatile). Systems with such memory can store data
6678 structures in RAM and retrieve them after a reboot, without flushing
6679 to typical _storage_.
6681 Linux supports NVRAM file systems thanks to either
6682 http://pramfs.sourceforge.net/[PRAMFS] or
6683 https://www.kernel.org/doc/Documentation/filesystems/dax.txt[DAX]{nbsp}+{nbsp}http://lkml.iu.edu/hypermail/linux/kernel/1504.1/03463.html[pmem]
6684 (requires Linux 4.1+).
6686 This section does not describe how to operate such file systems;
6687 we assume that you have a working persistent memory file system.
6689 When you create a <<tracing-session,tracing session>>, you can specify
6690 the path of the shared memory holding the sub-buffers. If you specify a
6691 location on an NVRAM file system, then you can retrieve the latest
6692 recorded trace data when the system reboots after a crash.
6694 To record trace data on a persistent memory file system and retrieve the
6695 trace data after a system crash:
6697 . Create a tracing session with a sub-buffer shared memory path located
6698 on an NVRAM file system:
6703 lttng create --shm-path=/path/to/shm
6707 . Configure the tracing session as usual with the man:lttng(1)
6708 command-line tool, and <<basic-tracing-session-control,start tracing>>.
6710 . After a system crash, use the man:lttng-crash(1) command-line tool to
6711 view the trace data recorded on the NVRAM file system:
6716 lttng-crash /path/to/shm
6720 The binary layout of the ring buffer files is not exactly the same as
6721 the trace files layout. This is why you need to use man:lttng-crash(1)
6722 instead of your preferred trace viewer directly.
6724 To convert the ring buffer files to LTTng trace files:
6726 * Use the opt:lttng-crash(1):--extract option of man:lttng-crash(1):
6731 lttng-crash --extract=/path/to/trace /path/to/shm
6739 [[lttng-modules-ref]]
6740 === noch:{LTTng-modules}
6743 [[lttng-modules-tp-fields]]
6744 ==== Tracepoint fields macros (for `TP_FIELDS()`)
6746 [[tp-fast-assign]][[tp-struct-entry]]The available macros to define
6747 tracepoint fields, which must be listed within `TP_FIELDS()` in
6748 `LTTNG_TRACEPOINT_EVENT()`, are:
6750 [role="func-desc growable",cols="asciidoc,asciidoc"]
6751 .Available macros to define LTTng-modules tracepoint fields
6753 |Macro |Description and parameters
6756 +ctf_integer(__t__, __n__, __e__)+
6758 +ctf_integer_nowrite(__t__, __n__, __e__)+
6760 +ctf_user_integer(__t__, __n__, __e__)+
6762 +ctf_user_integer_nowrite(__t__, __n__, __e__)+
6764 Standard integer, displayed in base 10.
6767 Integer C type (`int`, `long`, `size_t`, ...).
6773 Argument expression.
6776 +ctf_integer_hex(__t__, __n__, __e__)+
6778 +ctf_user_integer_hex(__t__, __n__, __e__)+
6780 Standard integer, displayed in base 16.
6789 Argument expression.
6791 |+ctf_integer_oct(__t__, __n__, __e__)+
6793 Standard integer, displayed in base 8.
6802 Argument expression.
6805 +ctf_integer_network(__t__, __n__, __e__)+
6807 +ctf_user_integer_network(__t__, __n__, __e__)+
6809 Integer in network byte order (big-endian), displayed in base 10.
6818 Argument expression.
6821 +ctf_integer_network_hex(__t__, __n__, __e__)+
6823 +ctf_user_integer_network_hex(__t__, __n__, __e__)+
6825 Integer in network byte order, displayed in base 16.
6834 Argument expression.
6837 +ctf_string(__n__, __e__)+
6839 +ctf_string_nowrite(__n__, __e__)+
6841 +ctf_user_string(__n__, __e__)+
6843 +ctf_user_string_nowrite(__n__, __e__)+
6845 Null-terminated string; undefined behavior if +__e__+ is `NULL`.
6851 Argument expression.
6854 +ctf_array(__t__, __n__, __e__, __s__)+
6856 +ctf_array_nowrite(__t__, __n__, __e__, __s__)+
6858 +ctf_user_array(__t__, __n__, __e__, __s__)+
6860 +ctf_user_array_nowrite(__t__, __n__, __e__, __s__)+
6862 Statically-sized array of integers.
6865 Array element C type.
6871 Argument expression.
6877 +ctf_array_bitfield(__t__, __n__, __e__, __s__)+
6879 +ctf_array_bitfield_nowrite(__t__, __n__, __e__, __s__)+
6881 +ctf_user_array_bitfield(__t__, __n__, __e__, __s__)+
6883 +ctf_user_array_bitfield_nowrite(__t__, __n__, __e__, __s__)+
6885 Statically-sized array of bits.
6887 The type of +__e__+ must be an integer type. +__s__+ is the number
6888 of elements of such type in +__e__+, not the number of bits.
6891 Array element C type.
6897 Argument expression.
6903 +ctf_array_text(__t__, __n__, __e__, __s__)+
6905 +ctf_array_text_nowrite(__t__, __n__, __e__, __s__)+
6907 +ctf_user_array_text(__t__, __n__, __e__, __s__)+
6909 +ctf_user_array_text_nowrite(__t__, __n__, __e__, __s__)+
6911 Statically-sized array, printed as text.
6913 The string does not need to be null-terminated.
6916 Array element C type (always `char`).
6922 Argument expression.
6928 +ctf_sequence(__t__, __n__, __e__, __T__, __E__)+
6930 +ctf_sequence_nowrite(__t__, __n__, __e__, __T__, __E__)+
6932 +ctf_user_sequence(__t__, __n__, __e__, __T__, __E__)+
6934 +ctf_user_sequence_nowrite(__t__, __n__, __e__, __T__, __E__)+
6936 Dynamically-sized array of integers.
6938 The type of +__E__+ must be unsigned.
6941 Array element C type.
6947 Argument expression.
6950 Length expression C type.
6956 +ctf_sequence_hex(__t__, __n__, __e__, __T__, __E__)+
6958 +ctf_user_sequence_hex(__t__, __n__, __e__, __T__, __E__)+
6960 Dynamically-sized array of integers, displayed in base 16.
6962 The type of +__E__+ must be unsigned.
6965 Array element C type.
6971 Argument expression.
6974 Length expression C type.
6979 |+ctf_sequence_network(__t__, __n__, __e__, __T__, __E__)+
6981 Dynamically-sized array of integers in network byte order (big-endian),
6982 displayed in base 10.
6984 The type of +__E__+ must be unsigned.
6987 Array element C type.
6993 Argument expression.
6996 Length expression C type.
7002 +ctf_sequence_bitfield(__t__, __n__, __e__, __T__, __E__)+
7004 +ctf_sequence_bitfield_nowrite(__t__, __n__, __e__, __T__, __E__)+
7006 +ctf_user_sequence_bitfield(__t__, __n__, __e__, __T__, __E__)+
7008 +ctf_user_sequence_bitfield_nowrite(__t__, __n__, __e__, __T__, __E__)+
7010 Dynamically-sized array of bits.
7012 The type of +__e__+ must be an integer type. +__s__+ is the number
7013 of elements of such type in +__e__+, not the number of bits.
7015 The type of +__E__+ must be unsigned.
7018 Array element C type.
7024 Argument expression.
7027 Length expression C type.
7033 +ctf_sequence_text(__t__, __n__, __e__, __T__, __E__)+
7035 +ctf_sequence_text_nowrite(__t__, __n__, __e__, __T__, __E__)+
7037 +ctf_user_sequence_text(__t__, __n__, __e__, __T__, __E__)+
7039 +ctf_user_sequence_text_nowrite(__t__, __n__, __e__, __T__, __E__)+
7041 Dynamically-sized array, displayed as text.
7043 The string does not need to be null-terminated.
7045 The type of +__E__+ must be unsigned.
7047 The behaviour is undefined if +__e__+ is `NULL`.
7050 Sequence element C type (always `char`).
7056 Argument expression.
7059 Length expression C type.
7065 Use the `_user` versions when the argument expression, `e`, is
7066 a user space address. In the cases of `ctf_user_integer*()` and
7067 `ctf_user_float*()`, `&e` must be a user space address, thus `e` must
7070 The `_nowrite` versions omit themselves from the session trace, but are
7071 otherwise identical. This means the `_nowrite` fields won't be written
7072 in the recorded trace. Their primary purpose is to make some
7073 of the event context available to the
7074 <<enabling-disabling-events,event filters>> without having to
7075 commit the data to sub-buffers.
7081 Terms related to LTTng and to tracing in general:
7084 The http://diamon.org/babeltrace[Babeltrace] project, which includes
7085 the cmd:babeltrace command, some libraries, and Python bindings.
7087 <<channel-buffering-schemes,buffering scheme>>::
7088 A layout of sub-buffers applied to a given channel.
7090 <<channel,channel>>::
7091 An entity which is responsible for a set of ring buffers.
7093 <<event,Event rules>> are always attached to a specific channel.
7096 A reference of time for a tracer.
7098 <<lttng-consumerd,consumer daemon>>::
7099 A process which is responsible for consuming the full sub-buffers
7100 and write them to a file system or send them over the network.
7102 <<channel-overwrite-mode-vs-discard-mode,discard mode>>:: The event loss
7103 mode in which the tracer _discards_ new event records when there's no
7104 sub-buffer space left to store them.
7107 The consequence of the execution of an instrumentation
7108 point, like a tracepoint that you manually place in some source code,
7109 or a Linux kernel KProbe.
7111 An event is said to _occur_ at a specific time. Different actions can
7112 be taken upon the occurance of an event, like record the event's payload
7115 <<channel-overwrite-mode-vs-discard-mode,event loss mode>>::
7116 The mechanism by which event records of a given channel are lost
7117 (not recorded) when there is no sub-buffer space left to store them.
7119 [[def-event-name]]event name::
7120 The name of an event, which is also the name of the event record.
7121 This is also called the _instrumentation point name_.
7124 A record, in a trace, of the payload of an event which occured.
7126 <<event,event rule>>::
7127 Set of conditions which must be satisfied for one or more occuring
7128 events to be recorded.
7130 `java.util.logging`::
7132 https://docs.oracle.com/javase/7/docs/api/java/util/logging/package-summary.html[core logging facilities].
7134 <<instrumenting,instrumentation>>::
7135 The use of LTTng probes to make a piece of software traceable.
7137 instrumentation point::
7138 A point in the execution path of a piece of software that, when
7139 reached by this execution, can emit an event.
7141 instrumentation point name::
7142 See _<<def-event-name,event name>>_.
7145 A http://logging.apache.org/log4j/1.2/[logging library] for Java
7146 developed by the Apache Software Foundation.
7149 Level of severity of a log statement or user space
7150 instrumentation point.
7153 The _Linux Trace Toolkit: next generation_ project.
7155 <<lttng-cli,cmd:lttng>>::
7156 A command-line tool provided by the LTTng-tools project which you
7157 can use to send and receive control messages to and from a
7161 The https://github.com/lttng/lttng-analyses[LTTng analyses] project,
7162 which is a set of analyzing programs that are used to obtain a
7163 higher level view of an LTTng trace.
7165 cmd:lttng-consumerd::
7166 The name of the consumer daemon program.
7169 A utility provided by the LTTng-tools project which can convert
7170 ring buffer files (usually
7171 <<persistent-memory-file-systems,saved on a persistent memory file system>>)
7174 LTTng Documentation::
7177 <<lttng-live,LTTng live>>::
7178 A communication protocol between the relay daemon and live viewers
7179 which makes it possible to see events "live", as they are received by
7182 <<lttng-modules,LTTng-modules>>::
7183 The https://github.com/lttng/lttng-modules[LTTng-modules] project,
7184 which contains the Linux kernel modules to make the Linux kernel
7185 instrumentation points available for LTTng tracing.
7188 The name of the relay daemon program.
7190 cmd:lttng-sessiond::
7191 The name of the session daemon program.
7194 The https://github.com/lttng/lttng-tools[LTTng-tools] project, which
7195 contains the various programs and libraries used to
7196 <<controlling-tracing,control tracing>>.
7198 <<lttng-ust,LTTng-UST>>::
7199 The https://github.com/lttng/lttng-ust[LTTng-UST] project, which
7200 contains libraries to instrument user applications.
7202 <<lttng-ust-agents,LTTng-UST Java agent>>::
7203 A Java package provided by the LTTng-UST project to allow the
7204 LTTng instrumentation of `java.util.logging` and Apache log4j 1.2
7207 <<lttng-ust-agents,LTTng-UST Python agent>>::
7208 A Python package provided by the LTTng-UST project to allow the
7209 LTTng instrumentation of Python logging statements.
7211 <<channel-overwrite-mode-vs-discard-mode,overwrite mode>>::
7212 The event loss mode in which new event records overwrite older
7213 event records when there's no sub-buffer space left to store them.
7215 <<channel-buffering-schemes,per-process buffering>>::
7216 A buffering scheme in which each instrumented process has its own
7217 sub-buffers for a given user space channel.
7219 <<channel-buffering-schemes,per-user buffering>>::
7220 A buffering scheme in which all the processes of a Unix user share the
7221 same sub-buffer for a given user space channel.
7223 <<lttng-relayd,relay daemon>>::
7224 A process which is responsible for receiving the trace data sent by
7225 a distant consumer daemon.
7228 A set of sub-buffers.
7230 <<lttng-sessiond,session daemon>>::
7231 A process which receives control commands from you and orchestrates
7232 the tracers and various LTTng daemons.
7234 <<taking-a-snapshot,snapshot>>::
7235 A copy of the current data of all the sub-buffers of a given tracing
7236 session, saved as trace files.
7239 One part of an LTTng ring buffer which contains event records.
7242 The time information attached to an event when it is emitted.
7245 A set of files which are the concatenations of one or more
7246 flushed sub-buffers.
7249 The action of recording the events emitted by an application
7250 or by a system, or to initiate such recording by controlling
7254 The http://tracecompass.org[Trace Compass] project and application.
7257 An instrumentation point using the tracepoint mechanism of the Linux
7258 kernel or of LTTng-UST.
7260 tracepoint definition::
7261 The definition of a single tracepoint.
7264 The name of a tracepoint.
7266 tracepoint provider::
7267 A set of functions providing tracepoints to an instrumented user
7270 Not to be confused with a _tracepoint provider package_: many tracepoint
7271 providers can exist within a tracepoint provider package.
7273 tracepoint provider package::
7274 One or more tracepoint providers compiled as an object file or as
7278 A software which records emitted events.
7280 <<domain,tracing domain>>::
7281 A namespace for event sources.
7284 The Unix group in which a Unix user can be to be allowed to trace the
7287 <<tracing-session,tracing session>>::
7288 A stateful dialogue between you and a <<lttng-sessiond,session
7292 An application running in user space, as opposed to a Linux kernel
7293 module, for example.