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 LTTng{nbsp}{revision} bears the name _KeKriek_. From
78 http://brasseriedunham.com/[Brasserie Dunham], the _**KeKriek**_ is a
79 sour mashed golden wheat ale fermented with local sour cherries from
80 Tougas orchards. Fresh sweet cherry notes with some tartness, lively
81 carbonation with a dry finish.
83 New features and changes in LTTng{nbsp}{revision}:
85 * **Tracing control**:
86 ** You can put more than one wildcard special character (`*`), and not
87 only at the end, when you <<enabling-disabling-events,create an event
88 rule>>, in both the instrumentation point name and the literal
90 link:/man/1/lttng-enable-event/v{revision}/#doc-filter-syntax[filter expressions]:
95 # lttng enable-event --kernel 'x86_*_local_timer_*' \
96 --filter='name == "*a*b*c*d*e" && count >= 23'
103 $ lttng enable-event --userspace '*_my_org:*msg*'
107 ** New trigger and notification API for
108 <<liblttng-ctl-lttng,`liblttng-ctl`>>. This new subsystem allows you
109 to register triggers which emit a notification when a given
110 condition is satisfied. As of LTTng{nbsp}{revision}, only
111 <<channel,channel>> buffer usage conditions are available.
112 Documentation is available in the
113 https://github.com/lttng/lttng-tools/tree/stable-{revision}/include/lttng[`liblttng-ctl`
116 ** You can now embed the whole textual LTTng-tools man pages into the
117 executables at build time with the `--enable-embedded-help`
118 configuration option. Thanks to this option, you don't need the
119 http://www.methods.co.nz/asciidoc/[AsciiDoc] and
120 https://directory.fsf.org/wiki/Xmlto[xmlto] tools at build time, and
121 a manual pager at run time, to get access to this documentation.
123 * **User space tracing**:
124 ** New blocking mode: an LTTng-UST tracepoint can now block until
125 <<channel,sub-buffer>> space is available instead of discarding event
126 records in <<channel-overwrite-mode-vs-discard-mode,discard mode>>.
127 With this feature, you can be sure that no event records are
128 discarded during your application's execution at the expense of
131 For example, the following command lines create a user space tracing
132 channel with an infinite blocking timeout and run an application
133 instrumented with LTTng-UST which is explicitly allowed to block:
139 $ lttng enable-channel --userspace --blocking-timeout=-1 blocking-channel
140 $ lttng enable-event --userspace --channel=blocking-channel --all
142 $ LTTNG_UST_ALLOW_BLOCKING=1 my-app
146 See the complete <<blocking-timeout-example,blocking timeout example>>.
148 * **Linux kernel tracing**:
149 ** Linux 4.10, 4.11, and 4.12 support.
150 ** The thread state dump events recorded by LTTng-modules now contain
151 the task's CPU identifier. This improves the precision of the
152 scheduler model for analyses.
153 ** Extended man:socketpair(2) system call tracing data.
159 What is LTTng? As its name suggests, the _Linux Trace Toolkit: next
160 generation_ is a modern toolkit for tracing Linux systems and
161 applications. So your first question might be:
168 As the history of software engineering progressed and led to what
169 we now take for granted--complex, numerous and
170 interdependent software applications running in parallel on
171 sophisticated operating systems like Linux--the authors of such
172 components, software developers, began feeling a natural
173 urge to have tools that would ensure the robustness and good performance
174 of their masterpieces.
176 One major achievement in this field is, inarguably, the
177 https://www.gnu.org/software/gdb/[GNU debugger (GDB)],
178 an essential tool for developers to find and fix bugs. But even the best
179 debugger won't help make your software run faster, and nowadays, faster
180 software means either more work done by the same hardware, or cheaper
181 hardware for the same work.
183 A _profiler_ is often the tool of choice to identify performance
184 bottlenecks. Profiling is suitable to identify _where_ performance is
185 lost in a given software. The profiler outputs a profile, a statistical
186 summary of observed events, which you may use to discover which
187 functions took the most time to execute. However, a profiler won't
188 report _why_ some identified functions are the bottleneck. Bottlenecks
189 might only occur when specific conditions are met, conditions that are
190 sometimes impossible to capture by a statistical profiler, or impossible
191 to reproduce with an application altered by the overhead of an
192 event-based profiler. For a thorough investigation of software
193 performance issues, a history of execution is essential, with the
194 recorded values of variables and context fields you choose, and
195 with as little influence as possible on the instrumented software. This
196 is where tracing comes in handy.
198 _Tracing_ is a technique used to understand what goes on in a running
199 software system. The software used for tracing is called a _tracer_,
200 which is conceptually similar to a tape recorder. When recording,
201 specific instrumentation points placed in the software source code
202 generate events that are saved on a giant tape: a _trace_ file. You
203 can trace user applications and the operating system at the same time,
204 opening the possibility of resolving a wide range of problems that would
205 otherwise be extremely challenging.
207 Tracing is often compared to _logging_. However, tracers and loggers are
208 two different tools, serving two different purposes. Tracers are
209 designed to record much lower-level events that occur much more
210 frequently than log messages, often in the range of thousands per
211 second, with very little execution overhead. Logging is more appropriate
212 for a very high-level analysis of less frequent events: user accesses,
213 exceptional conditions (errors and warnings, for example), database
214 transactions, instant messaging communications, and such. Simply put,
215 logging is one of the many use cases that can be satisfied with tracing.
217 The list of recorded events inside a trace file can be read manually
218 like a log file for the maximum level of detail, but it is generally
219 much more interesting to perform application-specific analyses to
220 produce reduced statistics and graphs that are useful to resolve a
221 given problem. Trace viewers and analyzers are specialized tools
224 In the end, this is what LTTng is: a powerful, open source set of
225 tools to trace the Linux kernel and user applications at the same time.
226 LTTng is composed of several components actively maintained and
227 developed by its link:/community/#where[community].
230 [[lttng-alternatives]]
231 === Alternatives to noch:{LTTng}
233 Excluding proprietary solutions, a few competing software tracers
236 * https://github.com/dtrace4linux/linux[dtrace4linux] is a port of
237 Sun Microsystems's DTrace to Linux. The cmd:dtrace tool interprets
238 user scripts and is responsible for loading code into the
239 Linux kernel for further execution and collecting the outputted data.
240 * https://en.wikipedia.org/wiki/Berkeley_Packet_Filter[eBPF] is a
241 subsystem in the Linux kernel in which a virtual machine can execute
242 programs passed from the user space to the kernel. You can attach
243 such programs to tracepoints and KProbes thanks to a system call, and
244 they can output data to the user space when executed thanks to
245 different mechanisms (pipe, VM register values, and eBPF maps, to name
247 * https://www.kernel.org/doc/Documentation/trace/ftrace.txt[ftrace]
248 is the de facto function tracer of the Linux kernel. Its user
249 interface is a set of special files in sysfs.
250 * https://perf.wiki.kernel.org/[perf] is
251 a performance analyzing tool for Linux which supports hardware
252 performance counters, tracepoints, as well as other counters and
253 types of probes. perf's controlling utility is the cmd:perf command
255 * http://linux.die.net/man/1/strace[strace]
256 is a command-line utility which records system calls made by a
257 user process, as well as signal deliveries and changes of process
258 state. strace makes use of https://en.wikipedia.org/wiki/Ptrace[ptrace]
259 to fulfill its function.
260 * http://www.sysdig.org/[sysdig], like SystemTap, uses scripts to
261 analyze Linux kernel events. You write scripts, or _chisels_ in
262 sysdig's jargon, in Lua and sysdig executes them while the system is
263 being traced or afterwards. sysdig's interface is the cmd:sysdig
264 command-line tool as well as the curses-based cmd:csysdig tool.
265 * https://sourceware.org/systemtap/[SystemTap] is a Linux kernel and
266 user space tracer which uses custom user scripts to produce plain text
267 traces. SystemTap converts the scripts to the C language, and then
268 compiles them as Linux kernel modules which are loaded to produce
269 trace data. SystemTap's primary user interface is the cmd:stap
272 The main distinctive features of LTTng is that it produces correlated
273 kernel and user space traces, as well as doing so with the lowest
274 overhead amongst other solutions. It produces trace files in the
275 http://diamon.org/ctf[CTF] format, a file format optimized
276 for the production and analyses of multi-gigabyte data.
278 LTTng is the result of more than 10 years of active open source
279 development by a community of passionate developers.
280 LTTng{nbsp}{revision} is currently available on major desktop and server
283 The main interface for tracing control is a single command-line tool
284 named cmd:lttng. The latter can create several tracing sessions, enable
285 and disable events on the fly, filter events efficiently with custom
286 user expressions, start and stop tracing, and much more. LTTng can
287 record the traces on the file system or send them over the network, and
288 keep them totally or partially. You can view the traces once tracing
289 becomes inactive or in real-time.
291 <<installing-lttng,Install LTTng now>> and
292 <<getting-started,start tracing>>!
298 **LTTng** is a set of software <<plumbing,components>> which interact to
299 <<instrumenting,instrument>> the Linux kernel and user applications, and
300 to <<controlling-tracing,control tracing>> (start and stop
301 tracing, enable and disable event rules, and the rest). Those
302 components are bundled into the following packages:
304 * **LTTng-tools**: Libraries and command-line interface to
306 * **LTTng-modules**: Linux kernel modules to instrument and
308 * **LTTng-UST**: Libraries and Java/Python packages to instrument and
309 trace user applications.
311 Most distributions mark the LTTng-modules and LTTng-UST packages as
312 optional when installing LTTng-tools (which is always required). In the
313 following sections, we always provide the steps to install all three,
316 * You only need to install LTTng-modules if you intend to trace the
318 * You only need to install LTTng-UST if you intend to trace user
322 .Availability of LTTng{nbsp}{revision} for major Linux distributions as of 25 July 2017.
324 |Distribution |Available in releases |Alternatives
326 |https://www.ubuntu.com/[Ubuntu]
327 |Ubuntu{nbsp}14.04 _Trusty Tahr_ and Ubuntu{nbsp}16.04 _Xenial Xerus_:
328 <<ubuntu-ppa,use the LTTng Stable{nbsp}{revision} PPA>>.
329 |link:/docs/v2.9#doc-ubuntu[LTTng{nbsp}2.9 for Ubuntu{nbsp}17.04 _Zesty Zapus_].
331 <<building-from-source,Build LTTng{nbsp}{revision} from source>> for
332 other Ubuntu releases.
334 |https://getfedora.org/[Fedora]
336 |link:/docs/v2.9#doc-fedora[LTTng{nbsp}2.9 for Fedora 26].
338 <<building-from-source,Build LTTng{nbsp}{revision} from source>>.
340 |https://www.debian.org/[Debian]
342 |link:/docs/v2.9#doc-debian[LTTng{nbsp}2.9 for Debian "stretch"
343 (stable), Debian "buster" (testing), and Debian "sid" (unstable)].
345 <<building-from-source,Build LTTng{nbsp}{revision} from source>>.
347 |https://www.archlinux.org/[Arch Linux]
349 |link:/docs/v2.9#doc-arch-linux[LTTng{nbsp}2.9 in the latest AUR packages].
351 |https://alpinelinux.org/[Alpine Linux]
353 |link:/docs/v2.9#doc-alpine-linux[LTTng{nbsp}2.9 for Alpine Linux "edge"].
355 <<building-from-source,Build LTTng{nbsp}{revision} from source>>.
357 |https://www.redhat.com/[RHEL] and https://www.suse.com/[SLES]
358 |See http://packages.efficios.com/[EfficiOS Enterprise Packages].
361 |https://buildroot.org/[Buildroot]
363 |link:/docs/v2.9#doc-buildroot[LTTng{nbsp}2.9 for Buildroot{nbsp}2017.02 and
364 Buildroot{nbsp}2017.05].
366 <<building-from-source,Build LTTng{nbsp}{revision} from source>>.
368 |http://www.openembedded.org/wiki/Main_Page[OpenEmbedded] and
369 https://www.yoctoproject.org/[Yocto]
371 |link:/docs/v2.9#doc-oe-yocto[LTTng{nbsp}2.9 for Yocto Project{nbsp}2.3 _Pyro_]
372 (`openembedded-core` layer).
374 <<building-from-source,Build LTTng{nbsp}{revision} from source>>.
379 === [[ubuntu-official-repositories]]Ubuntu
382 ==== noch:{LTTng} Stable {revision} PPA
384 The https://launchpad.net/~lttng/+archive/ubuntu/stable-{revision}[LTTng
385 Stable{nbsp}{revision} PPA] offers the latest stable
386 LTTng{nbsp}{revision} packages for:
388 * Ubuntu{nbsp}14.04 _Trusty Tahr_
389 * Ubuntu{nbsp}16.04 _Xenial Xerus_
391 To install LTTng{nbsp}{revision} from the LTTng Stable{nbsp}{revision} PPA:
393 . Add the LTTng Stable{nbsp}{revision} PPA repository and update the
399 # apt-add-repository ppa:lttng/stable-2.10
404 . Install the main LTTng{nbsp}{revision} packages:
409 # apt-get install lttng-tools
410 # apt-get install lttng-modules-dkms
411 # apt-get install liblttng-ust-dev
415 . **If you need to instrument and trace
416 <<java-application,Java applications>>**, install the LTTng-UST
422 # apt-get install liblttng-ust-agent-java
426 . **If you need to instrument and trace
427 <<python-application,Python{nbsp}3 applications>>**, install the
428 LTTng-UST Python agent:
433 # apt-get install python3-lttngust
438 [[enterprise-distributions]]
439 === RHEL, SUSE, and other enterprise distributions
441 To install LTTng on enterprise Linux distributions, such as Red Hat
442 Enterprise Linux (RHEL) and SUSE Linux Enterprise Server (SUSE), please
443 see http://packages.efficios.com/[EfficiOS Enterprise Packages].
446 [[building-from-source]]
447 === Build from source
449 To build and install LTTng{nbsp}{revision} from source:
451 . Using your distribution's package manager, or from source, install
452 the following dependencies of LTTng-tools and LTTng-UST:
455 * https://sourceforge.net/projects/libuuid/[libuuid]
456 * http://directory.fsf.org/wiki/Popt[popt]
457 * http://liburcu.org/[Userspace RCU]
458 * http://www.xmlsoft.org/[libxml2]
461 . Download, build, and install the latest LTTng-modules{nbsp}{revision}:
467 wget http://lttng.org/files/lttng-modules/lttng-modules-latest-2.10.tar.bz2 &&
468 tar -xf lttng-modules-latest-2.10.tar.bz2 &&
469 cd lttng-modules-2.10.* &&
471 sudo make modules_install &&
476 . Download, build, and install the latest LTTng-UST{nbsp}{revision}:
482 wget http://lttng.org/files/lttng-ust/lttng-ust-latest-2.10.tar.bz2 &&
483 tar -xf lttng-ust-latest-2.10.tar.bz2 &&
484 cd lttng-ust-2.10.* &&
494 .Java and Python application tracing
496 If you need to instrument and trace <<java-application,Java
497 applications>>, pass the `--enable-java-agent-jul`,
498 `--enable-java-agent-log4j`, or `--enable-java-agent-all` options to the
499 `configure` script, depending on which Java logging framework you use.
501 If you need to instrument and trace <<python-application,Python
502 applications>>, pass the `--enable-python-agent` option to the
503 `configure` script. You can set the `PYTHON` environment variable to the
504 path to the Python interpreter for which to install the LTTng-UST Python
512 By default, LTTng-UST libraries are installed to
513 dir:{/usr/local/lib}, which is the de facto directory in which to
514 keep self-compiled and third-party libraries.
516 When <<building-tracepoint-providers-and-user-application,linking an
517 instrumented user application with `liblttng-ust`>>:
519 * Append `/usr/local/lib` to the env:LD_LIBRARY_PATH environment
521 * Pass the `-L/usr/local/lib` and `-Wl,-rpath,/usr/local/lib` options to
522 man:gcc(1), man:g++(1), or man:clang(1).
526 . Download, build, and install the latest LTTng-tools{nbsp}{revision}:
532 wget http://lttng.org/files/lttng-tools/lttng-tools-latest-2.10.tar.bz2 &&
533 tar -xf lttng-tools-latest-2.10.tar.bz2 &&
534 cd lttng-tools-2.10.* &&
542 TIP: The https://github.com/eepp/vlttng[vlttng tool] can do all the
543 previous steps automatically for a given version of LTTng and confine
544 the installed files in a specific directory. This can be useful to test
545 LTTng without installing it on your system.
551 This is a short guide to get started quickly with LTTng kernel and user
554 Before you follow this guide, make sure to <<installing-lttng,install>>
557 This tutorial walks you through the steps to:
559 . <<tracing-the-linux-kernel,Trace the Linux kernel>>.
560 . <<tracing-your-own-user-application,Trace a user application>> written
562 . <<viewing-and-analyzing-your-traces,View and analyze the
566 [[tracing-the-linux-kernel]]
567 === Trace the Linux kernel
569 The following command lines start with the `#` prompt because you need
570 root privileges to trace the Linux kernel. You can also trace the kernel
571 as a regular user if your Unix user is a member of the
572 <<tracing-group,tracing group>>.
574 . Create a <<tracing-session,tracing session>> which writes its traces
575 to dir:{/tmp/my-kernel-trace}:
580 # lttng create my-kernel-session --output=/tmp/my-kernel-trace
584 . List the available kernel tracepoints and system calls:
589 # lttng list --kernel
590 # lttng list --kernel --syscall
594 . Create <<event,event rules>> which match the desired instrumentation
595 point names, for example the `sched_switch` and `sched_process_fork`
596 tracepoints, and the man:open(2) and man:close(2) system calls:
601 # lttng enable-event --kernel sched_switch,sched_process_fork
602 # lttng enable-event --kernel --syscall open,close
606 You can also create an event rule which matches _all_ the Linux kernel
607 tracepoints (this will generate a lot of data when tracing):
612 # lttng enable-event --kernel --all
616 . <<basic-tracing-session-control,Start tracing>>:
625 . Do some operation on your system for a few seconds. For example,
626 load a website, or list the files of a directory.
627 . <<basic-tracing-session-control,Stop tracing>> and destroy the
638 The man:lttng-destroy(1) command does not destroy the trace data; it
639 only destroys the state of the tracing session.
641 . For the sake of this example, make the recorded trace accessible to
647 # chown -R $(whoami) /tmp/my-kernel-trace
651 See <<viewing-and-analyzing-your-traces,View and analyze the
652 recorded events>> to view the recorded events.
655 [[tracing-your-own-user-application]]
656 === Trace a user application
658 This section steps you through a simple example to trace a
659 _Hello world_ program written in C.
661 To create the traceable user application:
663 . Create the tracepoint provider header file, which defines the
664 tracepoints and the events they can generate:
670 #undef TRACEPOINT_PROVIDER
671 #define TRACEPOINT_PROVIDER hello_world
673 #undef TRACEPOINT_INCLUDE
674 #define TRACEPOINT_INCLUDE "./hello-tp.h"
676 #if !defined(_HELLO_TP_H) || defined(TRACEPOINT_HEADER_MULTI_READ)
679 #include <lttng/tracepoint.h>
689 ctf_string(my_string_field, my_string_arg)
690 ctf_integer(int, my_integer_field, my_integer_arg)
694 #endif /* _HELLO_TP_H */
696 #include <lttng/tracepoint-event.h>
700 . Create the tracepoint provider package source file:
706 #define TRACEPOINT_CREATE_PROBES
707 #define TRACEPOINT_DEFINE
709 #include "hello-tp.h"
713 . Build the tracepoint provider package:
718 $ gcc -c -I. hello-tp.c
722 . Create the _Hello World_ application source file:
729 #include "hello-tp.h"
731 int main(int argc, char *argv[])
735 puts("Hello, World!\nPress Enter to continue...");
738 * The following getchar() call is only placed here for the purpose
739 * of this demonstration, to pause the application in order for
740 * you to have time to list its tracepoints. It is not
746 * A tracepoint() call.
748 * Arguments, as defined in hello-tp.h:
750 * 1. Tracepoint provider name (required)
751 * 2. Tracepoint name (required)
752 * 3. my_integer_arg (first user-defined argument)
753 * 4. my_string_arg (second user-defined argument)
755 * Notice the tracepoint provider and tracepoint names are
756 * NOT strings: they are in fact parts of variables that the
757 * macros in hello-tp.h create.
759 tracepoint(hello_world, my_first_tracepoint, 23, "hi there!");
761 for (x = 0; x < argc; ++x) {
762 tracepoint(hello_world, my_first_tracepoint, x, argv[x]);
765 puts("Quitting now!");
766 tracepoint(hello_world, my_first_tracepoint, x * x, "x^2");
773 . Build the application:
782 . Link the application with the tracepoint provider package,
783 `liblttng-ust`, and `libdl`:
788 $ gcc -o hello hello.o hello-tp.o -llttng-ust -ldl
792 Here's the whole build process:
795 .User space tracing tutorial's build steps.
796 image::ust-flow.png[]
798 To trace the user application:
800 . Run the application with a few arguments:
805 $ ./hello world and beyond
814 Press Enter to continue...
818 . Start an LTTng <<lttng-sessiond,session daemon>>:
823 $ lttng-sessiond --daemonize
827 Note that a session daemon might already be running, for example as
828 a service that the distribution's service manager started.
830 . List the available user space tracepoints:
835 $ lttng list --userspace
839 You see the `hello_world:my_first_tracepoint` tracepoint listed
840 under the `./hello` process.
842 . Create a <<tracing-session,tracing session>>:
847 $ lttng create my-user-space-session
851 . Create an <<event,event rule>> which matches the
852 `hello_world:my_first_tracepoint` event name:
857 $ lttng enable-event --userspace hello_world:my_first_tracepoint
861 . <<basic-tracing-session-control,Start tracing>>:
870 . Go back to the running `hello` application and press Enter. The
871 program executes all `tracepoint()` instrumentation points and exits.
872 . <<basic-tracing-session-control,Stop tracing>> and destroy the
883 The man:lttng-destroy(1) command does not destroy the trace data; it
884 only destroys the state of the tracing session.
886 By default, LTTng saves the traces in
887 +$LTTNG_HOME/lttng-traces/__name__-__date__-__time__+,
888 where +__name__+ is the tracing session name. The
889 env:LTTNG_HOME environment variable defaults to `$HOME` if not set.
891 See <<viewing-and-analyzing-your-traces,View and analyze the
892 recorded events>> to view the recorded events.
895 [[viewing-and-analyzing-your-traces]]
896 === View and analyze the recorded events
898 Once you have completed the <<tracing-the-linux-kernel,Trace the Linux
899 kernel>> and <<tracing-your-own-user-application,Trace a user
900 application>> tutorials, you can inspect the recorded events.
902 Many tools are available to read LTTng traces:
904 * **cmd:babeltrace** is a command-line utility which converts trace
905 formats; it supports the format that LTTng produces, CTF, as well as a
906 basic text output which can be ++grep++ed. The cmd:babeltrace command
907 is part of the http://diamon.org/babeltrace[Babeltrace] project.
908 * Babeltrace also includes
909 **https://www.python.org/[Python] bindings** so
910 that you can easily open and read an LTTng trace with your own script,
911 benefiting from the power of Python.
912 * http://tracecompass.org/[**Trace Compass**]
913 is a graphical user interface for viewing and analyzing any type of
914 logs or traces, including LTTng's.
915 * https://github.com/lttng/lttng-analyses[**LTTng analyses**] is a
916 project which includes many high-level analyses of LTTng kernel
917 traces, like scheduling statistics, interrupt frequency distribution,
918 top CPU usage, and more.
920 NOTE: This section assumes that the traces recorded during the previous
921 tutorials were saved to their default location, in the
922 dir:{$LTTNG_HOME/lttng-traces} directory. The env:LTTNG_HOME
923 environment variable defaults to `$HOME` if not set.
926 [[viewing-and-analyzing-your-traces-bt]]
927 ==== Use the cmd:babeltrace command-line tool
929 The simplest way to list all the recorded events of a trace is to pass
930 its path to cmd:babeltrace with no options:
934 $ babeltrace ~/lttng-traces/my-user-space-session*
937 cmd:babeltrace finds all traces recursively within the given path and
938 prints all their events, merging them in chronological order.
940 You can pipe the output of cmd:babeltrace into a tool like man:grep(1) for
945 $ babeltrace /tmp/my-kernel-trace | grep _switch
948 You can pipe the output of cmd:babeltrace into a tool like man:wc(1) to
949 count the recorded events:
953 $ babeltrace /tmp/my-kernel-trace | grep _open | wc --lines
957 [[viewing-and-analyzing-your-traces-bt-python]]
958 ==== Use the Babeltrace Python bindings
960 The <<viewing-and-analyzing-your-traces-bt,text output of cmd:babeltrace>>
961 is useful to isolate events by simple matching using man:grep(1) and
962 similar utilities. However, more elaborate filters, such as keeping only
963 event records with a field value falling within a specific range, are
964 not trivial to write using a shell. Moreover, reductions and even the
965 most basic computations involving multiple event records are virtually
966 impossible to implement.
968 Fortunately, Babeltrace ships with Python 3 bindings which makes it easy
969 to read the event records of an LTTng trace sequentially and compute the
972 The following script accepts an LTTng Linux kernel trace path as its
973 first argument and prints the short names of the top 5 running processes
974 on CPU 0 during the whole trace:
979 from collections import Counter
985 if len(sys.argv) != 2:
986 msg = 'Usage: python3 {} TRACEPATH'.format(sys.argv[0])
987 print(msg, file=sys.stderr)
990 # A trace collection contains one or more traces
991 col = babeltrace.TraceCollection()
993 # Add the trace provided by the user (LTTng traces always have
995 if col.add_trace(sys.argv[1], 'ctf') is None:
996 raise RuntimeError('Cannot add trace')
998 # This counter dict contains execution times:
1000 # task command name -> total execution time (ns)
1001 exec_times = Counter()
1003 # This contains the last `sched_switch` timestamp
1007 for event in col.events:
1008 # Keep only `sched_switch` events
1009 if event.name != 'sched_switch':
1012 # Keep only events which happened on CPU 0
1013 if event['cpu_id'] != 0:
1017 cur_ts = event.timestamp
1023 # Previous task command (short) name
1024 prev_comm = event['prev_comm']
1026 # Initialize entry in our dict if not yet done
1027 if prev_comm not in exec_times:
1028 exec_times[prev_comm] = 0
1030 # Compute previous command execution time
1031 diff = cur_ts - last_ts
1033 # Update execution time of this command
1034 exec_times[prev_comm] += diff
1036 # Update last timestamp
1040 for name, ns in exec_times.most_common(5):
1042 print('{:20}{} s'.format(name, s))
1047 if __name__ == '__main__':
1048 sys.exit(0 if top5proc() else 1)
1055 $ python3 top5proc.py /tmp/my-kernel-trace/kernel
1061 swapper/0 48.607245889 s
1062 chromium 7.192738188 s
1063 pavucontrol 0.709894415 s
1064 Compositor 0.660867933 s
1065 Xorg.bin 0.616753786 s
1068 Note that `swapper/0` is the "idle" process of CPU 0 on Linux; since we
1069 weren't using the CPU that much when tracing, its first position in the
1074 == [[understanding-lttng]]Core concepts
1076 From a user's perspective, the LTTng system is built on a few concepts,
1077 or objects, on which the <<lttng-cli,cmd:lttng command-line tool>>
1078 operates by sending commands to the <<lttng-sessiond,session daemon>>.
1079 Understanding how those objects relate to eachother is key in mastering
1082 The core concepts are:
1084 * <<tracing-session,Tracing session>>
1085 * <<domain,Tracing domain>>
1086 * <<channel,Channel and ring buffer>>
1087 * <<"event","Instrumentation point, event rule, event, and event record">>
1093 A _tracing session_ is a stateful dialogue between you and
1094 a <<lttng-sessiond,session daemon>>. You can
1095 <<creating-destroying-tracing-sessions,create a new tracing
1096 session>> with the `lttng create` command.
1098 Anything that you do when you control LTTng tracers happens within a
1099 tracing session. In particular, a tracing session:
1102 * Has its own set of trace files.
1103 * Has its own state of activity (started or stopped).
1104 * Has its own <<tracing-session-mode,mode>> (local, network streaming,
1106 * Has its own <<channel,channels>> which have their own
1107 <<event,event rules>>.
1110 .A _tracing session_ contains <<channel,channels>> that are members of <<domain,tracing domains>> and contain <<event,event rules>>.
1111 image::concepts.png[]
1113 Those attributes and objects are completely isolated between different
1116 A tracing session is analogous to a cash machine session:
1117 the operations you do on the banking system through the cash machine do
1118 not alter the data of other users of the same system. In the case of
1119 the cash machine, a session lasts as long as your bank card is inside.
1120 In the case of LTTng, a tracing session lasts from the `lttng create`
1121 command to the `lttng destroy` command.
1124 .Each Unix user has its own set of tracing sessions.
1125 image::many-sessions.png[]
1128 [[tracing-session-mode]]
1129 ==== Tracing session mode
1131 LTTng can send the generated trace data to different locations. The
1132 _tracing session mode_ dictates where to send it. The following modes
1133 are available in LTTng{nbsp}{revision}:
1136 LTTng writes the traces to the file system of the machine being traced
1139 Network streaming mode::
1140 LTTng sends the traces over the network to a
1141 <<lttng-relayd,relay daemon>> running on a remote system.
1144 LTTng does not write the traces by default. Instead, you can request
1145 LTTng to <<taking-a-snapshot,take a snapshot>>, that is, a copy of the
1146 current tracing buffers, and to write it to the target's file system
1147 or to send it over the network to a <<lttng-relayd,relay daemon>>
1148 running on a remote system.
1151 This mode is similar to the network streaming mode, but a live
1152 trace viewer can connect to the distant relay daemon to
1153 <<lttng-live,view event records as LTTng generates them>> by
1160 A _tracing domain_ is a namespace for event sources. A tracing domain
1161 has its own properties and features.
1163 There are currently five available tracing domains:
1167 * `java.util.logging` (JUL)
1171 You must specify a tracing domain when using some commands to avoid
1172 ambiguity. For example, since all the domains support named tracepoints
1173 as event sources (instrumentation points that you manually insert in the
1174 source code), you need to specify a tracing domain when
1175 <<enabling-disabling-events,creating an event rule>> because all the
1176 tracing domains could have tracepoints with the same names.
1178 Some features are reserved to specific tracing domains. Dynamic function
1179 entry and return instrumentation points, for example, are currently only
1180 supported in the Linux kernel tracing domain, but support for other
1181 tracing domains could be added in the future.
1183 You can create <<channel,channels>> in the Linux kernel and user space
1184 tracing domains. The other tracing domains have a single default
1189 === Channel and ring buffer
1191 A _channel_ is an object which is responsible for a set of ring buffers.
1192 Each ring buffer is divided into multiple sub-buffers. When an LTTng
1193 tracer emits an event, it can record it to one or more
1194 sub-buffers. The attributes of a channel determine what to do when
1195 there's no space left for a new event record because all sub-buffers
1196 are full, where to send a full sub-buffer, and other behaviours.
1198 A channel is always associated to a <<domain,tracing domain>>. The
1199 `java.util.logging` (JUL), log4j, and Python tracing domains each have
1200 a default channel which you cannot configure.
1202 A channel also owns <<event,event rules>>. When an LTTng tracer emits
1203 an event, it records it to the sub-buffers of all
1204 the enabled channels with a satisfied event rule, as long as those
1205 channels are part of active <<tracing-session,tracing sessions>>.
1208 [[channel-buffering-schemes]]
1209 ==== Per-user vs. per-process buffering schemes
1211 A channel has at least one ring buffer _per CPU_. LTTng always
1212 records an event to the ring buffer associated to the CPU on which it
1215 Two _buffering schemes_ are available when you
1216 <<enabling-disabling-channels,create a channel>> in the
1217 user space <<domain,tracing domain>>:
1219 Per-user buffering::
1220 Allocate one set of ring buffers--one per CPU--shared by all the
1221 instrumented processes of each Unix user.
1225 .Per-user buffering scheme.
1226 image::per-user-buffering.png[]
1229 Per-process buffering::
1230 Allocate one set of ring buffers--one per CPU--for each
1231 instrumented process.
1235 .Per-process buffering scheme.
1236 image::per-process-buffering.png[]
1239 The per-process buffering scheme tends to consume more memory than the
1240 per-user option because systems generally have more instrumented
1241 processes than Unix users running instrumented processes. However, the
1242 per-process buffering scheme ensures that one process having a high
1243 event throughput won't fill all the shared sub-buffers of the same
1246 The Linux kernel tracing domain has only one available buffering scheme
1247 which is to allocate a single set of ring buffers for the whole system.
1248 This scheme is similar to the per-user option, but with a single, global
1249 user "running" the kernel.
1252 [[channel-overwrite-mode-vs-discard-mode]]
1253 ==== Overwrite vs. discard event loss modes
1255 When an event occurs, LTTng records it to a specific sub-buffer (yellow
1256 arc in the following animation) of a specific channel's ring buffer.
1257 When there's no space left in a sub-buffer, the tracer marks it as
1258 consumable (red) and another, empty sub-buffer starts receiving the
1259 following event records. A <<lttng-consumerd,consumer daemon>>
1260 eventually consumes the marked sub-buffer (returns to white).
1263 [role="docsvg-channel-subbuf-anim"]
1268 In an ideal world, sub-buffers are consumed faster than they are filled,
1269 as is the case in the previous animation. In the real world,
1270 however, all sub-buffers can be full at some point, leaving no space to
1271 record the following events.
1273 By default, LTTng-modules and LTTng-UST are _non-blocking_ tracers: when
1274 no empty sub-buffer is available, it is acceptable to lose event records
1275 when the alternative would be to cause substantial delays in the
1276 instrumented application's execution. LTTng privileges performance over
1277 integrity; it aims at perturbing the traced system as little as possible
1278 in order to make tracing of subtle race conditions and rare interrupt
1281 Starting from LTTng{nbsp}2.10, the LTTng user space tracer, LTTng-UST,
1282 supports a _blocking mode_. See the <<blocking-timeout-example,blocking
1283 timeout example>> to learn how to use the blocking mode.
1285 When it comes to losing event records because no empty sub-buffer is
1286 available, or because the <<opt-blocking-timeout,blocking timeout>> is
1287 reached, the channel's _event loss mode_ determines what to do. The
1288 available event loss modes are:
1291 Drop the newest event records until a the tracer
1292 releases a sub-buffer.
1295 Clear the sub-buffer containing the oldest event records and start
1296 writing the newest event records there.
1298 This mode is sometimes called _flight recorder mode_ because it's
1300 https://en.wikipedia.org/wiki/Flight_recorder[flight recorder]:
1301 always keep a fixed amount of the latest data.
1303 Which mechanism you should choose depends on your context: prioritize
1304 the newest or the oldest event records in the ring buffer?
1306 Beware that, in overwrite mode, the tracer abandons a whole sub-buffer
1307 as soon as a there's no space left for a new event record, whereas in
1308 discard mode, the tracer only discards the event record that doesn't
1311 In discard mode, LTTng increments a count of lost event records when
1312 an event record is lost and saves this count to the trace. In
1313 overwrite mode, LTTng keeps no information when it overwrites a
1314 sub-buffer before consuming it.
1316 There are a few ways to decrease your probability of losing event
1318 <<channel-subbuf-size-vs-subbuf-count,Sub-buffer count and size>> shows
1319 how you can fine-une the sub-buffer count and size of a channel to
1320 virtually stop losing event records, though at the cost of greater
1324 [[channel-subbuf-size-vs-subbuf-count]]
1325 ==== Sub-buffer count and size
1327 When you <<enabling-disabling-channels,create a channel>>, you can
1328 set its number of sub-buffers and their size.
1330 Note that there is noticeable CPU overhead introduced when
1331 switching sub-buffers (marking a full one as consumable and switching
1332 to an empty one for the following events to be recorded). Knowing this,
1333 the following list presents a few practical situations along with how
1334 to configure the sub-buffer count and size for them:
1336 * **High event throughput**: In general, prefer bigger sub-buffers to
1337 lower the risk of losing event records.
1339 Having bigger sub-buffers also ensures a lower
1340 <<channel-switch-timer,sub-buffer switching frequency>>.
1342 The number of sub-buffers is only meaningful if you create the channel
1343 in overwrite mode: in this case, if a sub-buffer overwrite happens, the
1344 other sub-buffers are left unaltered.
1346 * **Low event throughput**: In general, prefer smaller sub-buffers
1347 since the risk of losing event records is low.
1349 Because events occur less frequently, the sub-buffer switching frequency
1350 should remain low and thus the tracer's overhead should not be a
1353 * **Low memory system**: If your target system has a low memory
1354 limit, prefer fewer first, then smaller sub-buffers.
1356 Even if the system is limited in memory, you want to keep the
1357 sub-buffers as big as possible to avoid a high sub-buffer switching
1360 Note that LTTng uses http://diamon.org/ctf/[CTF] as its trace format,
1361 which means event data is very compact. For example, the average
1362 LTTng kernel event record weights about 32{nbsp}bytes. Thus, a
1363 sub-buffer size of 1{nbsp}MiB is considered big.
1365 The previous situations highlight the major trade-off between a few big
1366 sub-buffers and more, smaller sub-buffers: sub-buffer switching
1367 frequency vs. how much data is lost in overwrite mode. Assuming a
1368 constant event throughput and using the overwrite mode, the two
1369 following configurations have the same ring buffer total size:
1372 [role="docsvg-channel-subbuf-size-vs-count-anim"]
1377 * **2 sub-buffers of 4{nbsp}MiB each**: Expect a very low sub-buffer
1378 switching frequency, but if a sub-buffer overwrite happens, half of
1379 the event records so far (4{nbsp}MiB) are definitely lost.
1380 * **8 sub-buffers of 1{nbsp}MiB each**: Expect 4{nbsp}times the tracer's
1381 overhead as the previous configuration, but if a sub-buffer
1382 overwrite happens, only the eighth of event records so far are
1385 In discard mode, the sub-buffers count parameter is pointless: use two
1386 sub-buffers and set their size according to the requirements of your
1390 [[channel-switch-timer]]
1391 ==== Switch timer period
1393 The _switch timer period_ is an important configurable attribute of
1394 a channel to ensure periodic sub-buffer flushing.
1396 When the _switch timer_ expires, a sub-buffer switch happens. You can
1397 set the switch timer period attribute when you
1398 <<enabling-disabling-channels,create a channel>> to ensure that event
1399 data is consumed and committed to trace files or to a distant relay
1400 daemon periodically in case of a low event throughput.
1403 [role="docsvg-channel-switch-timer"]
1408 This attribute is also convenient when you use big sub-buffers to cope
1409 with a sporadic high event throughput, even if the throughput is
1413 [[channel-read-timer]]
1414 ==== Read timer period
1416 By default, the LTTng tracers use a notification mechanism to signal a
1417 full sub-buffer so that a consumer daemon can consume it. When such
1418 notifications must be avoided, for example in real-time applications,
1419 you can use the channel's _read timer_ instead. When the read timer
1420 fires, the <<lttng-consumerd,consumer daemon>> checks for full,
1421 consumable sub-buffers.
1424 [[tracefile-rotation]]
1425 ==== Trace file count and size
1427 By default, trace files can grow as large as needed. You can set the
1428 maximum size of each trace file that a channel writes when you
1429 <<enabling-disabling-channels,create a channel>>. When the size of
1430 a trace file reaches the channel's fixed maximum size, LTTng creates
1431 another file to contain the next event records. LTTng appends a file
1432 count to each trace file name in this case.
1434 If you set the trace file size attribute when you create a channel, the
1435 maximum number of trace files that LTTng creates is _unlimited_ by
1436 default. To limit them, you can also set a maximum number of trace
1437 files. When the number of trace files reaches the channel's fixed
1438 maximum count, the oldest trace file is overwritten. This mechanism is
1439 called _trace file rotation_.
1443 === Instrumentation point, event rule, event, and event record
1445 An _event rule_ is a set of conditions which must be **all** satisfied
1446 for LTTng to record an occuring event.
1448 You set the conditions when you <<enabling-disabling-events,create
1451 You always attach an event rule to <<channel,channel>> when you create
1454 When an event passes the conditions of an event rule, LTTng records it
1455 in one of the attached channel's sub-buffers.
1457 The available conditions, as of LTTng{nbsp}{revision}, are:
1459 * The event rule _is enabled_.
1460 * The instrumentation point's type _is{nbsp}T_.
1461 * The instrumentation point's name (sometimes called _event name_)
1462 _matches{nbsp}N_, but _is not{nbsp}E_.
1463 * The instrumentation point's log level _is as severe as{nbsp}L_, or
1464 _is exactly{nbsp}L_.
1465 * The fields of the event's payload _satisfy_ a filter
1466 expression{nbsp}__F__.
1468 As you can see, all the conditions but the dynamic filter are related to
1469 the event rule's status or to the instrumentation point, not to the
1470 occurring events. This is why, without a filter, checking if an event
1471 passes an event rule is not a dynamic task: when you create or modify an
1472 event rule, all the tracers of its tracing domain enable or disable the
1473 instrumentation points themselves once. This is possible because the
1474 attributes of an instrumentation point (type, name, and log level) are
1475 defined statically. In other words, without a dynamic filter, the tracer
1476 _does not evaluate_ the arguments of an instrumentation point unless it
1477 matches an enabled event rule.
1479 Note that, for LTTng to record an event, the <<channel,channel>> to
1480 which a matching event rule is attached must also be enabled, and the
1481 tracing session owning this channel must be active.
1484 .Logical path from an instrumentation point to an event record.
1485 image::event-rule.png[]
1487 .Event, event record, or event rule?
1489 With so many similar terms, it's easy to get confused.
1491 An **event** is the consequence of the execution of an _instrumentation
1492 point_, like a tracepoint that you manually place in some source code,
1493 or a Linux kernel KProbe. An event is said to _occur_ at a specific
1494 time. Different actions can be taken upon the occurrence of an event,
1495 like record the event's payload to a buffer.
1497 An **event record** is the representation of an event in a sub-buffer. A
1498 tracer is responsible for capturing the payload of an event, current
1499 context variables, the event's ID, and the event's timestamp. LTTng
1500 can append this sub-buffer to a trace file.
1502 An **event rule** is a set of conditions which must all be satisfied for
1503 LTTng to record an occuring event. Events still occur without
1504 satisfying event rules, but LTTng does not record them.
1509 == Components of noch:{LTTng}
1511 The second _T_ in _LTTng_ stands for _toolkit_: it would be wrong
1512 to call LTTng a simple _tool_ since it is composed of multiple
1513 interacting components. This section describes those components,
1514 explains their respective roles, and shows how they connect together to
1515 form the LTTng ecosystem.
1517 The following diagram shows how the most important components of LTTng
1518 interact with user applications, the Linux kernel, and you:
1521 .Control and trace data paths between LTTng components.
1522 image::plumbing.png[]
1524 The LTTng project incorporates:
1526 * **LTTng-tools**: Libraries and command-line interface to
1527 control tracing sessions.
1528 ** <<lttng-sessiond,Session daemon>> (man:lttng-sessiond(8)).
1529 ** <<lttng-consumerd,Consumer daemon>> (man:lttng-consumerd(8)).
1530 ** <<lttng-relayd,Relay daemon>> (man:lttng-relayd(8)).
1531 ** <<liblttng-ctl-lttng,Tracing control library>> (`liblttng-ctl`).
1532 ** <<lttng-cli,Tracing control command-line tool>> (man:lttng(1)).
1533 * **LTTng-UST**: Libraries and Java/Python packages to trace user
1535 ** <<lttng-ust,User space tracing library>> (`liblttng-ust`) and its
1536 headers to instrument and trace any native user application.
1537 ** <<prebuilt-ust-helpers,Preloadable user space tracing helpers>>:
1538 *** `liblttng-ust-libc-wrapper`
1539 *** `liblttng-ust-pthread-wrapper`
1540 *** `liblttng-ust-cyg-profile`
1541 *** `liblttng-ust-cyg-profile-fast`
1542 *** `liblttng-ust-dl`
1543 ** User space tracepoint provider source files generator command-line
1544 tool (man:lttng-gen-tp(1)).
1545 ** <<lttng-ust-agents,LTTng-UST Java agent>> to instrument and trace
1546 Java applications using `java.util.logging` or
1547 Apache log4j 1.2 logging.
1548 ** <<lttng-ust-agents,LTTng-UST Python agent>> to instrument
1549 Python applications using the standard `logging` package.
1550 * **LTTng-modules**: <<lttng-modules,Linux kernel modules>> to trace
1552 ** LTTng kernel tracer module.
1553 ** Tracing ring buffer kernel modules.
1554 ** Probe kernel modules.
1555 ** LTTng logger kernel module.
1559 === Tracing control command-line interface
1562 .The tracing control command-line interface.
1563 image::plumbing-lttng-cli.png[]
1565 The _man:lttng(1) command-line tool_ is the standard user interface to
1566 control LTTng <<tracing-session,tracing sessions>>. The cmd:lttng tool
1567 is part of LTTng-tools.
1569 The cmd:lttng tool is linked with
1570 <<liblttng-ctl-lttng,`liblttng-ctl`>> to communicate with
1571 one or more <<lttng-sessiond,session daemons>> behind the scenes.
1573 The cmd:lttng tool has a Git-like interface:
1577 $ lttng <GENERAL OPTIONS> <COMMAND> <COMMAND OPTIONS>
1580 The <<controlling-tracing,Tracing control>> section explores the
1581 available features of LTTng using the cmd:lttng tool.
1584 [[liblttng-ctl-lttng]]
1585 === Tracing control library
1588 .The tracing control library.
1589 image::plumbing-liblttng-ctl.png[]
1591 The _LTTng control library_, `liblttng-ctl`, is used to communicate
1592 with a <<lttng-sessiond,session daemon>> using a C API that hides the
1593 underlying protocol's details. `liblttng-ctl` is part of LTTng-tools.
1595 The <<lttng-cli,cmd:lttng command-line tool>>
1596 is linked with `liblttng-ctl`.
1598 You can use `liblttng-ctl` in C or $$C++$$ source code by including its
1603 #include <lttng/lttng.h>
1606 Some objects are referenced by name (C string), such as tracing
1607 sessions, but most of them require to create a handle first using
1608 `lttng_create_handle()`.
1610 The best available developer documentation for `liblttng-ctl` is, as of
1611 LTTng{nbsp}{revision}, its installed header files. Every function and
1612 structure is thoroughly documented.
1616 === User space tracing library
1619 .The user space tracing library.
1620 image::plumbing-liblttng-ust.png[]
1622 The _user space tracing library_, `liblttng-ust` (see man:lttng-ust(3)),
1623 is the LTTng user space tracer. It receives commands from a
1624 <<lttng-sessiond,session daemon>>, for example to
1625 enable and disable specific instrumentation points, and writes event
1626 records to ring buffers shared with a
1627 <<lttng-consumerd,consumer daemon>>.
1628 `liblttng-ust` is part of LTTng-UST.
1630 Public C header files are installed beside `liblttng-ust` to
1631 instrument any <<c-application,C or $$C++$$ application>>.
1633 <<lttng-ust-agents,LTTng-UST agents>>, which are regular Java and Python
1634 packages, use their own library providing tracepoints which is
1635 linked with `liblttng-ust`.
1637 An application or library does not have to initialize `liblttng-ust`
1638 manually: its constructor does the necessary tasks to properly register
1639 to a session daemon. The initialization phase also enables the
1640 instrumentation points matching the <<event,event rules>> that you
1644 [[lttng-ust-agents]]
1645 === User space tracing agents
1648 .The user space tracing agents.
1649 image::plumbing-lttng-ust-agents.png[]
1651 The _LTTng-UST Java and Python agents_ are regular Java and Python
1652 packages which add LTTng tracing capabilities to the
1653 native logging frameworks. The LTTng-UST agents are part of LTTng-UST.
1655 In the case of Java, the
1656 https://docs.oracle.com/javase/7/docs/api/java/util/logging/package-summary.html[`java.util.logging`
1657 core logging facilities] and
1658 https://logging.apache.org/log4j/1.2/[Apache log4j 1.2] are supported.
1659 Note that Apache Log4{nbsp}2 is not supported.
1661 In the case of Python, the standard
1662 https://docs.python.org/3/library/logging.html[`logging`] package
1663 is supported. Both Python 2 and Python 3 modules can import the
1664 LTTng-UST Python agent package.
1666 The applications using the LTTng-UST agents are in the
1667 `java.util.logging` (JUL),
1668 log4j, and Python <<domain,tracing domains>>.
1670 Both agents use the same mechanism to trace the log statements. When an
1671 agent is initialized, it creates a log handler that attaches to the root
1672 logger. The agent also registers to a <<lttng-sessiond,session daemon>>.
1673 When the application executes a log statement, it is passed to the
1674 agent's log handler by the root logger. The agent's log handler calls a
1675 native function in a tracepoint provider package shared library linked
1676 with <<lttng-ust,`liblttng-ust`>>, passing the formatted log message and
1677 other fields, like its logger name and its log level. This native
1678 function contains a user space instrumentation point, hence tracing the
1681 The log level condition of an
1682 <<event,event rule>> is considered when tracing
1683 a Java or a Python application, and it's compatible with the standard
1684 JUL, log4j, and Python log levels.
1688 === LTTng kernel modules
1691 .The LTTng kernel modules.
1692 image::plumbing-lttng-modules.png[]
1694 The _LTTng kernel modules_ are a set of Linux kernel modules
1695 which implement the kernel tracer of the LTTng project. The LTTng
1696 kernel modules are part of LTTng-modules.
1698 The LTTng kernel modules include:
1700 * A set of _probe_ modules.
1702 Each module attaches to a specific subsystem
1703 of the Linux kernel using its tracepoint instrument points. There are
1704 also modules to attach to the entry and return points of the Linux
1705 system call functions.
1707 * _Ring buffer_ modules.
1709 A ring buffer implementation is provided as kernel modules. The LTTng
1710 kernel tracer writes to the ring buffer; a
1711 <<lttng-consumerd,consumer daemon>> reads from the ring buffer.
1713 * The _LTTng kernel tracer_ module.
1714 * The _LTTng logger_ module.
1716 The LTTng logger module implements the special path:{/proc/lttng-logger}
1717 file so that any executable can generate LTTng events by opening and
1718 writing to this file.
1720 See <<proc-lttng-logger-abi,LTTng logger>>.
1722 Generally, you do not have to load the LTTng kernel modules manually
1723 (using man:modprobe(8), for example): a root <<lttng-sessiond,session
1724 daemon>> loads the necessary modules when starting. If you have extra
1725 probe modules, you can specify to load them to the session daemon on
1728 The LTTng kernel modules are installed in
1729 +/usr/lib/modules/__release__/extra+ by default, where +__release__+ is
1730 the kernel release (see `uname --kernel-release`).
1737 .The session daemon.
1738 image::plumbing-sessiond.png[]
1740 The _session daemon_, man:lttng-sessiond(8), is a daemon responsible for
1741 managing tracing sessions and for controlling the various components of
1742 LTTng. The session daemon is part of LTTng-tools.
1744 The session daemon sends control requests to and receives control
1747 * The <<lttng-ust,user space tracing library>>.
1749 Any instance of the user space tracing library first registers to
1750 a session daemon. Then, the session daemon can send requests to
1751 this instance, such as:
1754 ** Get the list of tracepoints.
1755 ** Share an <<event,event rule>> so that the user space tracing library
1756 can enable or disable tracepoints. Amongst the possible conditions
1757 of an event rule is a filter expression which `liblttng-ust` evalutes
1758 when an event occurs.
1759 ** Share <<channel,channel>> attributes and ring buffer locations.
1762 The session daemon and the user space tracing library use a Unix
1763 domain socket for their communication.
1765 * The <<lttng-ust-agents,user space tracing agents>>.
1767 Any instance of a user space tracing agent first registers to
1768 a session daemon. Then, the session daemon can send requests to
1769 this instance, such as:
1772 ** Get the list of loggers.
1773 ** Enable or disable a specific logger.
1776 The session daemon and the user space tracing agent use a TCP connection
1777 for their communication.
1779 * The <<lttng-modules,LTTng kernel tracer>>.
1780 * The <<lttng-consumerd,consumer daemon>>.
1782 The session daemon sends requests to the consumer daemon to instruct
1783 it where to send the trace data streams, amongst other information.
1785 * The <<lttng-relayd,relay daemon>>.
1787 The session daemon receives commands from the
1788 <<liblttng-ctl-lttng,tracing control library>>.
1790 The root session daemon loads the appropriate
1791 <<lttng-modules,LTTng kernel modules>> on startup. It also spawns
1792 a <<lttng-consumerd,consumer daemon>> as soon as you create
1793 an <<event,event rule>>.
1795 The session daemon does not send and receive trace data: this is the
1796 role of the <<lttng-consumerd,consumer daemon>> and
1797 <<lttng-relayd,relay daemon>>. It does, however, generate the
1798 http://diamon.org/ctf/[CTF] metadata stream.
1800 Each Unix user can have its own session daemon instance. The
1801 tracing sessions managed by different session daemons are completely
1804 The root user's session daemon is the only one which is
1805 allowed to control the LTTng kernel tracer, and its spawned consumer
1806 daemon is the only one which is allowed to consume trace data from the
1807 LTTng kernel tracer. Note, however, that any Unix user which is a member
1808 of the <<tracing-group,tracing group>> is allowed
1809 to create <<channel,channels>> in the
1810 Linux kernel <<domain,tracing domain>>, and thus to trace the Linux
1813 The <<lttng-cli,cmd:lttng command-line tool>> automatically starts a
1814 session daemon when using its `create` command if none is currently
1815 running. You can also start the session daemon manually.
1822 .The consumer daemon.
1823 image::plumbing-consumerd.png[]
1825 The _consumer daemon_, man:lttng-consumerd(8), is a daemon which shares
1826 ring buffers with user applications or with the LTTng kernel modules to
1827 collect trace data and send it to some location (on disk or to a
1828 <<lttng-relayd,relay daemon>> over the network). The consumer daemon
1829 is part of LTTng-tools.
1831 You do not start a consumer daemon manually: a consumer daemon is always
1832 spawned by a <<lttng-sessiond,session daemon>> as soon as you create an
1833 <<event,event rule>>, that is, before you start tracing. When you kill
1834 its owner session daemon, the consumer daemon also exits because it is
1835 the session daemon's child process. Command-line options of
1836 man:lttng-sessiond(8) target the consumer daemon process.
1838 There are up to two running consumer daemons per Unix user, whereas only
1839 one session daemon can run per user. This is because each process can be
1840 either 32-bit or 64-bit: if the target system runs a mixture of 32-bit
1841 and 64-bit processes, it is more efficient to have separate
1842 corresponding 32-bit and 64-bit consumer daemons. The root user is an
1843 exception: it can have up to _three_ running consumer daemons: 32-bit
1844 and 64-bit instances for its user applications, and one more
1845 reserved for collecting kernel trace data.
1853 image::plumbing-relayd.png[]
1855 The _relay daemon_, man:lttng-relayd(8), is a daemon acting as a bridge
1856 between remote session and consumer daemons, local trace files, and a
1857 remote live trace viewer. The relay daemon is part of LTTng-tools.
1859 The main purpose of the relay daemon is to implement a receiver of
1860 <<sending-trace-data-over-the-network,trace data over the network>>.
1861 This is useful when the target system does not have much file system
1862 space to record trace files locally.
1864 The relay daemon is also a server to which a
1865 <<lttng-live,live trace viewer>> can
1866 connect. The live trace viewer sends requests to the relay daemon to
1867 receive trace data as the target system emits events. The
1868 communication protocol is named _LTTng live_; it is used over TCP
1871 Note that you can start the relay daemon on the target system directly.
1872 This is the setup of choice when the use case is to view events as
1873 the target system emits them without the need of a remote system.
1877 == [[using-lttng]]Instrumentation
1879 There are many examples of tracing and monitoring in our everyday life:
1881 * You have access to real-time and historical weather reports and
1882 forecasts thanks to weather stations installed around the country.
1883 * You know your heart is safe thanks to an electrocardiogram.
1884 * You make sure not to drive your car too fast and to have enough fuel
1885 to reach your destination thanks to gauges visible on your dashboard.
1887 All the previous examples have something in common: they rely on
1888 **instruments**. Without the electrodes attached to the surface of your
1889 body's skin, cardiac monitoring is futile.
1891 LTTng, as a tracer, is no different from those real life examples. If
1892 you're about to trace a software system or, in other words, record its
1893 history of execution, you better have **instrumentation points** in the
1894 subject you're tracing, that is, the actual software.
1896 Various ways were developed to instrument a piece of software for LTTng
1897 tracing. The most straightforward one is to manually place
1898 instrumentation points, called _tracepoints_, in the software's source
1899 code. It is also possible to add instrumentation points dynamically in
1900 the Linux kernel <<domain,tracing domain>>.
1902 If you're only interested in tracing the Linux kernel, your
1903 instrumentation needs are probably already covered by LTTng's built-in
1904 <<lttng-modules,Linux kernel tracepoints>>. You may also wish to trace a
1905 user application which is already instrumented for LTTng tracing.
1906 In such cases, you can skip this whole section and read the topics of
1907 the <<controlling-tracing,Tracing control>> section.
1909 Many methods are available to instrument a piece of software for LTTng
1912 * <<c-application,User space instrumentation for C and $$C++$$
1914 * <<prebuilt-ust-helpers,Prebuilt user space tracing helpers>>.
1915 * <<java-application,User space Java agent>>.
1916 * <<python-application,User space Python agent>>.
1917 * <<proc-lttng-logger-abi,LTTng logger>>.
1918 * <<instrumenting-linux-kernel,LTTng kernel tracepoints>>.
1922 === [[cxx-application]]User space instrumentation for C and $$C++$$ applications
1924 The procedure to instrument a C or $$C++$$ user application with
1925 the <<lttng-ust,LTTng user space tracing library>>, `liblttng-ust`, is:
1927 . <<tracepoint-provider,Create the source files of a tracepoint provider
1929 . <<probing-the-application-source-code,Add tracepoints to
1930 the application's source code>>.
1931 . <<building-tracepoint-providers-and-user-application,Build and link
1932 a tracepoint provider package and the user application>>.
1934 If you need quick, man:printf(3)-like instrumentation, you can skip
1935 those steps and use <<tracef,`tracef()`>> or <<tracelog,`tracelog()`>>
1938 IMPORTANT: You need to <<installing-lttng,install>> LTTng-UST to
1939 instrument a user application with `liblttng-ust`.
1942 [[tracepoint-provider]]
1943 ==== Create the source files of a tracepoint provider package
1945 A _tracepoint provider_ is a set of compiled functions which provide
1946 **tracepoints** to an application, the type of instrumentation point
1947 supported by LTTng-UST. Those functions can emit events with
1948 user-defined fields and serialize those events as event records to one
1949 or more LTTng-UST <<channel,channel>> sub-buffers. The `tracepoint()`
1950 macro, which you <<probing-the-application-source-code,insert in a user
1951 application's source code>>, calls those functions.
1953 A _tracepoint provider package_ is an object file (`.o`) or a shared
1954 library (`.so`) which contains one or more tracepoint providers.
1955 Its source files are:
1957 * One or more <<tpp-header,tracepoint provider header>> (`.h`).
1958 * A <<tpp-source,tracepoint provider package source>> (`.c`).
1960 A tracepoint provider package is dynamically linked with `liblttng-ust`,
1961 the LTTng user space tracer, at run time.
1964 .User application linked with `liblttng-ust` and containing a tracepoint provider.
1965 image::ust-app.png[]
1967 NOTE: If you need quick, man:printf(3)-like instrumentation, you can
1968 skip creating and using a tracepoint provider and use
1969 <<tracef,`tracef()`>> or <<tracelog,`tracelog()`>> instead.
1973 ===== Create a tracepoint provider header file template
1975 A _tracepoint provider header file_ contains the tracepoint
1976 definitions of a tracepoint provider.
1978 To create a tracepoint provider header file:
1980 . Start from this template:
1984 .Tracepoint provider header file template (`.h` file extension).
1986 #undef TRACEPOINT_PROVIDER
1987 #define TRACEPOINT_PROVIDER provider_name
1989 #undef TRACEPOINT_INCLUDE
1990 #define TRACEPOINT_INCLUDE "./tp.h"
1992 #if !defined(_TP_H) || defined(TRACEPOINT_HEADER_MULTI_READ)
1995 #include <lttng/tracepoint.h>
1998 * Use TRACEPOINT_EVENT(), TRACEPOINT_EVENT_CLASS(),
1999 * TRACEPOINT_EVENT_INSTANCE(), and TRACEPOINT_LOGLEVEL() here.
2004 #include <lttng/tracepoint-event.h>
2010 * `provider_name` with the name of your tracepoint provider.
2011 * `"tp.h"` with the name of your tracepoint provider header file.
2013 . Below the `#include <lttng/tracepoint.h>` line, put your
2014 <<defining-tracepoints,tracepoint definitions>>.
2016 Your tracepoint provider name must be unique amongst all the possible
2017 tracepoint provider names used on the same target system. We
2018 suggest to include the name of your project or company in the name,
2019 for example, `org_lttng_my_project_tpp`.
2021 TIP: [[lttng-gen-tp]]You can use the man:lttng-gen-tp(1) tool to create
2022 this boilerplate for you. When using cmd:lttng-gen-tp, all you need to
2023 write are the <<defining-tracepoints,tracepoint definitions>>.
2026 [[defining-tracepoints]]
2027 ===== Create a tracepoint definition
2029 A _tracepoint definition_ defines, for a given tracepoint:
2031 * Its **input arguments**. They are the macro parameters that the
2032 `tracepoint()` macro accepts for this particular tracepoint
2033 in the user application's source code.
2034 * Its **output event fields**. They are the sources of event fields
2035 that form the payload of any event that the execution of the
2036 `tracepoint()` macro emits for this particular tracepoint.
2038 You can create a tracepoint definition by using the
2039 `TRACEPOINT_EVENT()` macro below the `#include <lttng/tracepoint.h>`
2041 <<tpp-header,tracepoint provider header file template>>.
2043 The syntax of the `TRACEPOINT_EVENT()` macro is:
2046 .`TRACEPOINT_EVENT()` macro syntax.
2049 /* Tracepoint provider name */
2052 /* Tracepoint name */
2055 /* Input arguments */
2060 /* Output event fields */
2069 * `provider_name` with your tracepoint provider name.
2070 * `tracepoint_name` with your tracepoint name.
2071 * `arguments` with the <<tpp-def-input-args,input arguments>>.
2072 * `fields` with the <<tpp-def-output-fields,output event field>>
2075 This tracepoint emits events named `provider_name:tracepoint_name`.
2078 .Event name's length limitation
2080 The concatenation of the tracepoint provider name and the
2081 tracepoint name must not exceed **254 characters**. If it does, the
2082 instrumented application compiles and runs, but LTTng throws multiple
2083 warnings and you could experience serious issues.
2086 [[tpp-def-input-args]]The syntax of the `TP_ARGS()` macro is:
2089 .`TP_ARGS()` macro syntax.
2098 * `type` with the C type of the argument.
2099 * `arg_name` with the argument name.
2101 You can repeat `type` and `arg_name` up to 10 times to have
2102 more than one argument.
2104 .`TP_ARGS()` usage with three arguments.
2116 The `TP_ARGS()` and `TP_ARGS(void)` forms are valid to create a
2117 tracepoint definition with no input arguments.
2119 [[tpp-def-output-fields]]The `TP_FIELDS()` macro contains a list of
2120 `ctf_*()` macros. Each `ctf_*()` macro defines one event field. See
2121 man:lttng-ust(3) for a complete description of the available `ctf_*()`
2122 macros. A `ctf_*()` macro specifies the type, size, and byte order of
2125 Each `ctf_*()` macro takes an _argument expression_ parameter. This is a
2126 C expression that the tracer evalutes at the `tracepoint()` macro site
2127 in the application's source code. This expression provides a field's
2128 source of data. The argument expression can include input argument names
2129 listed in the `TP_ARGS()` macro.
2131 Each `ctf_*()` macro also takes a _field name_ parameter. Field names
2132 must be unique within a given tracepoint definition.
2134 Here's a complete tracepoint definition example:
2136 .Tracepoint definition.
2138 The following tracepoint definition defines a tracepoint which takes
2139 three input arguments and has four output event fields.
2143 #include "my-custom-structure.h"
2149 const struct my_custom_structure*, my_custom_structure,
2154 ctf_string(query_field, query)
2155 ctf_float(double, ratio_field, ratio)
2156 ctf_integer(int, recv_size, my_custom_structure->recv_size)
2157 ctf_integer(int, send_size, my_custom_structure->send_size)
2162 You can refer to this tracepoint definition with the `tracepoint()`
2163 macro in your application's source code like this:
2167 tracepoint(my_provider, my_tracepoint,
2168 my_structure, some_ratio, the_query);
2172 NOTE: The LTTng tracer only evaluates tracepoint arguments at run time
2173 if they satisfy an enabled <<event,event rule>>.
2176 [[using-tracepoint-classes]]
2177 ===== Use a tracepoint class
2179 A _tracepoint class_ is a class of tracepoints which share the same
2180 output event field definitions. A _tracepoint instance_ is one
2181 instance of such a defined tracepoint class, with its own tracepoint
2184 The <<defining-tracepoints,`TRACEPOINT_EVENT()` macro>> is actually a
2185 shorthand which defines both a tracepoint class and a tracepoint
2186 instance at the same time.
2188 When you build a tracepoint provider package, the C or $$C++$$ compiler
2189 creates one serialization function for each **tracepoint class**. A
2190 serialization function is responsible for serializing the event fields
2191 of a tracepoint to a sub-buffer when tracing.
2193 For various performance reasons, when your situation requires multiple
2194 tracepoint definitions with different names, but with the same event
2195 fields, we recommend that you manually create a tracepoint class
2196 and instantiate as many tracepoint instances as needed. One positive
2197 effect of such a design, amongst other advantages, is that all
2198 tracepoint instances of the same tracepoint class reuse the same
2199 serialization function, thus reducing
2200 https://en.wikipedia.org/wiki/Cache_pollution[cache pollution].
2202 .Use a tracepoint class and tracepoint instances.
2204 Consider the following three tracepoint definitions:
2216 ctf_integer(int, userid, userid)
2217 ctf_integer(size_t, len, len)
2229 ctf_integer(int, userid, userid)
2230 ctf_integer(size_t, len, len)
2242 ctf_integer(int, userid, userid)
2243 ctf_integer(size_t, len, len)
2248 In this case, we create three tracepoint classes, with one implicit
2249 tracepoint instance for each of them: `get_account`, `get_settings`, and
2250 `get_transaction`. However, they all share the same event field names
2251 and types. Hence three identical, yet independent serialization
2252 functions are created when you build the tracepoint provider package.
2254 A better design choice is to define a single tracepoint class and three
2255 tracepoint instances:
2259 /* The tracepoint class */
2260 TRACEPOINT_EVENT_CLASS(
2261 /* Tracepoint provider name */
2264 /* Tracepoint class name */
2267 /* Input arguments */
2273 /* Output event fields */
2275 ctf_integer(int, userid, userid)
2276 ctf_integer(size_t, len, len)
2280 /* The tracepoint instances */
2281 TRACEPOINT_EVENT_INSTANCE(
2282 /* Tracepoint provider name */
2285 /* Tracepoint class name */
2288 /* Tracepoint name */
2291 /* Input arguments */
2297 TRACEPOINT_EVENT_INSTANCE(
2306 TRACEPOINT_EVENT_INSTANCE(
2319 [[assigning-log-levels]]
2320 ===== Assign a log level to a tracepoint definition
2322 You can assign an optional _log level_ to a
2323 <<defining-tracepoints,tracepoint definition>>.
2325 Assigning different levels of severity to tracepoint definitions can
2326 be useful: when you <<enabling-disabling-events,create an event rule>>,
2327 you can target tracepoints having a log level as severe as a specific
2330 The concept of LTTng-UST log levels is similar to the levels found
2331 in typical logging frameworks:
2333 * In a logging framework, the log level is given by the function
2334 or method name you use at the log statement site: `debug()`,
2335 `info()`, `warn()`, `error()`, and so on.
2336 * In LTTng-UST, you statically assign the log level to a tracepoint
2337 definition; any `tracepoint()` macro invocation which refers to
2338 this definition has this log level.
2340 You can assign a log level to a tracepoint definition with the
2341 `TRACEPOINT_LOGLEVEL()` macro. You must use this macro _after_ the
2342 <<defining-tracepoints,`TRACEPOINT_EVENT()`>> or
2343 <<using-tracepoint-classes,`TRACEPOINT_INSTANCE()`>> macro for a given
2346 The syntax of the `TRACEPOINT_LOGLEVEL()` macro is:
2349 .`TRACEPOINT_LOGLEVEL()` macro syntax.
2351 TRACEPOINT_LOGLEVEL(provider_name, tracepoint_name, log_level)
2356 * `provider_name` with the tracepoint provider name.
2357 * `tracepoint_name` with the tracepoint name.
2358 * `log_level` with the log level to assign to the tracepoint
2359 definition named `tracepoint_name` in the `provider_name`
2360 tracepoint provider.
2362 See man:lttng-ust(3) for a list of available log level names.
2364 .Assign the `TRACE_DEBUG_UNIT` log level to a tracepoint definition.
2368 /* Tracepoint definition */
2377 ctf_integer(int, userid, userid)
2378 ctf_integer(size_t, len, len)
2382 /* Log level assignment */
2383 TRACEPOINT_LOGLEVEL(my_app, get_transaction, TRACE_DEBUG_UNIT)
2389 ===== Create a tracepoint provider package source file
2391 A _tracepoint provider package source file_ is a C source file which
2392 includes a <<tpp-header,tracepoint provider header file>> to expand its
2393 macros into event serialization and other functions.
2395 You can always use the following tracepoint provider package source
2399 .Tracepoint provider package source file template.
2401 #define TRACEPOINT_CREATE_PROBES
2406 Replace `tp.h` with the name of your <<tpp-header,tracepoint provider
2407 header file>> name. You may also include more than one tracepoint
2408 provider header file here to create a tracepoint provider package
2409 holding more than one tracepoint providers.
2412 [[probing-the-application-source-code]]
2413 ==== Add tracepoints to an application's source code
2415 Once you <<tpp-header,create a tracepoint provider header file>>, you
2416 can use the `tracepoint()` macro in your application's
2417 source code to insert the tracepoints that this header
2418 <<defining-tracepoints,defines>>.
2420 The `tracepoint()` macro takes at least two parameters: the tracepoint
2421 provider name and the tracepoint name. The corresponding tracepoint
2422 definition defines the other parameters.
2424 .`tracepoint()` usage.
2426 The following <<defining-tracepoints,tracepoint definition>> defines a
2427 tracepoint which takes two input arguments and has two output event
2431 .Tracepoint provider header file.
2433 #include "my-custom-structure.h"
2440 const char*, cmd_name
2443 ctf_string(cmd_name, cmd_name)
2444 ctf_integer(int, number_of_args, argc)
2449 You can refer to this tracepoint definition with the `tracepoint()`
2450 macro in your application's source code like this:
2453 .Application's source file.
2457 int main(int argc, char* argv[])
2459 tracepoint(my_provider, my_tracepoint, argc, argv[0]);
2465 Note how the application's source code includes
2466 the tracepoint provider header file containing the tracepoint
2467 definitions to use, path:{tp.h}.
2470 .`tracepoint()` usage with a complex tracepoint definition.
2472 Consider this complex tracepoint definition, where multiple event
2473 fields refer to the same input arguments in their argument expression
2477 .Tracepoint provider header file.
2479 /* For `struct stat` */
2480 #include <sys/types.h>
2481 #include <sys/stat.h>
2493 ctf_integer(int, my_constant_field, 23 + 17)
2494 ctf_integer(int, my_int_arg_field, my_int_arg)
2495 ctf_integer(int, my_int_arg_field2, my_int_arg * my_int_arg)
2496 ctf_integer(int, sum4_field, my_str_arg[0] + my_str_arg[1] +
2497 my_str_arg[2] + my_str_arg[3])
2498 ctf_string(my_str_arg_field, my_str_arg)
2499 ctf_integer_hex(off_t, size_field, st->st_size)
2500 ctf_float(double, size_dbl_field, (double) st->st_size)
2501 ctf_sequence_text(char, half_my_str_arg_field, my_str_arg,
2502 size_t, strlen(my_str_arg) / 2)
2507 You can refer to this tracepoint definition with the `tracepoint()`
2508 macro in your application's source code like this:
2511 .Application's source file.
2513 #define TRACEPOINT_DEFINE
2520 stat("/etc/fstab", &s);
2521 tracepoint(my_provider, my_tracepoint, 23, "Hello, World!", &s);
2527 If you look at the event record that LTTng writes when tracing this
2528 program, assuming the file size of path:{/etc/fstab} is 301{nbsp}bytes,
2529 it should look like this:
2531 .Event record fields
2533 |Field's name |Field's value
2534 |`my_constant_field` |40
2535 |`my_int_arg_field` |23
2536 |`my_int_arg_field2` |529
2538 |`my_str_arg_field` |`Hello, World!`
2539 |`size_field` |0x12d
2540 |`size_dbl_field` |301.0
2541 |`half_my_str_arg_field` |`Hello,`
2545 Sometimes, the arguments you pass to `tracepoint()` are expensive to
2546 compute--they use the call stack, for example. To avoid this
2547 computation when the tracepoint is disabled, you can use the
2548 `tracepoint_enabled()` and `do_tracepoint()` macros.
2550 The syntax of the `tracepoint_enabled()` and `do_tracepoint()` macros
2554 .`tracepoint_enabled()` and `do_tracepoint()` macros syntax.
2556 tracepoint_enabled(provider_name, tracepoint_name)
2557 do_tracepoint(provider_name, tracepoint_name, ...)
2562 * `provider_name` with the tracepoint provider name.
2563 * `tracepoint_name` with the tracepoint name.
2565 `tracepoint_enabled()` returns a non-zero value if the tracepoint named
2566 `tracepoint_name` from the provider named `provider_name` is enabled
2569 `do_tracepoint()` is like `tracepoint()`, except that it doesn't check
2570 if the tracepoint is enabled. Using `tracepoint()` with
2571 `tracepoint_enabled()` is dangerous since `tracepoint()` also contains
2572 the `tracepoint_enabled()` check, thus a race condition is
2573 possible in this situation:
2576 .Possible race condition when using `tracepoint_enabled()` with `tracepoint()`.
2578 if (tracepoint_enabled(my_provider, my_tracepoint)) {
2579 stuff = prepare_stuff();
2582 tracepoint(my_provider, my_tracepoint, stuff);
2585 If the tracepoint is enabled after the condition, then `stuff` is not
2586 prepared: the emitted event will either contain wrong data, or the whole
2587 application could crash (segmentation fault, for example).
2589 NOTE: Neither `tracepoint_enabled()` nor `do_tracepoint()` have an
2590 `STAP_PROBEV()` call. If you need it, you must emit
2594 [[building-tracepoint-providers-and-user-application]]
2595 ==== Build and link a tracepoint provider package and an application
2597 Once you have one or more <<tpp-header,tracepoint provider header
2598 files>> and a <<tpp-source,tracepoint provider package source file>>,
2599 you can create the tracepoint provider package by compiling its source
2600 file. From here, multiple build and run scenarios are possible. The
2601 following table shows common application and library configurations
2602 along with the required command lines to achieve them.
2604 In the following diagrams, we use the following file names:
2607 Executable application.
2610 Application's object file.
2613 Tracepoint provider package object file.
2616 Tracepoint provider package archive file.
2619 Tracepoint provider package shared object file.
2622 User library object file.
2625 User library shared object file.
2627 We use the following symbols in the diagrams of table below:
2630 .Symbols used in the build scenario diagrams.
2631 image::ust-sit-symbols.png[]
2633 We assume that path:{.} is part of the env:LD_LIBRARY_PATH environment
2634 variable in the following instructions.
2636 [role="growable ust-scenarios",cols="asciidoc,asciidoc"]
2637 .Common tracepoint provider package scenarios.
2639 |Scenario |Instructions
2642 The instrumented application is statically linked with
2643 the tracepoint provider package object.
2645 image::ust-sit+app-linked-with-tp-o+app-instrumented.png[]
2648 include::../common/ust-sit-step-tp-o.txt[]
2650 To build the instrumented application:
2652 . In path:{app.c}, before including path:{tpp.h}, add the following line:
2657 #define TRACEPOINT_DEFINE
2661 . Compile the application source file:
2670 . Build the application:
2675 $ gcc -o app app.o tpp.o -llttng-ust -ldl
2679 To run the instrumented application:
2681 * Start the application:
2691 The instrumented application is statically linked with the
2692 tracepoint provider package archive file.
2694 image::ust-sit+app-linked-with-tp-a+app-instrumented.png[]
2697 To create the tracepoint provider package archive file:
2699 . Compile the <<tpp-source,tracepoint provider package source file>>:
2708 . Create the tracepoint provider package archive file:
2713 $ ar rcs tpp.a tpp.o
2717 To build the instrumented application:
2719 . In path:{app.c}, before including path:{tpp.h}, add the following line:
2724 #define TRACEPOINT_DEFINE
2728 . Compile the application source file:
2737 . Build the application:
2742 $ gcc -o app app.o tpp.a -llttng-ust -ldl
2746 To run the instrumented application:
2748 * Start the application:
2758 The instrumented application is linked with the tracepoint provider
2759 package shared object.
2761 image::ust-sit+app-linked-with-tp-so+app-instrumented.png[]
2764 include::../common/ust-sit-step-tp-so.txt[]
2766 To build the instrumented application:
2768 . In path:{app.c}, before including path:{tpp.h}, add the following line:
2773 #define TRACEPOINT_DEFINE
2777 . Compile the application source file:
2786 . Build the application:
2791 $ gcc -o app app.o -ldl -L. -ltpp
2795 To run the instrumented application:
2797 * Start the application:
2807 The tracepoint provider package shared object is preloaded before the
2808 instrumented application starts.
2810 image::ust-sit+tp-so-preloaded+app-instrumented.png[]
2813 include::../common/ust-sit-step-tp-so.txt[]
2815 To build the instrumented application:
2817 . In path:{app.c}, before including path:{tpp.h}, add the
2823 #define TRACEPOINT_DEFINE
2824 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
2828 . Compile the application source file:
2837 . Build the application:
2842 $ gcc -o app app.o -ldl
2846 To run the instrumented application with tracing support:
2848 * Preload the tracepoint provider package shared object and
2849 start the application:
2854 $ LD_PRELOAD=./libtpp.so ./app
2858 To run the instrumented application without tracing support:
2860 * Start the application:
2870 The instrumented application dynamically loads the tracepoint provider
2871 package shared object.
2873 See the <<dlclose-warning,warning about `dlclose()`>>.
2875 image::ust-sit+app-dlopens-tp-so+app-instrumented.png[]
2878 include::../common/ust-sit-step-tp-so.txt[]
2880 To build the instrumented application:
2882 . In path:{app.c}, before including path:{tpp.h}, add the
2888 #define TRACEPOINT_DEFINE
2889 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
2893 . Compile the application source file:
2902 . Build the application:
2907 $ gcc -o app app.o -ldl
2911 To run the instrumented application:
2913 * Start the application:
2923 The application is linked with the instrumented user library.
2925 The instrumented user library is statically linked with the tracepoint
2926 provider package object file.
2928 image::ust-sit+app-linked-with-lib+lib-linked-with-tp-o+lib-instrumented.png[]
2931 include::../common/ust-sit-step-tp-o-fpic.txt[]
2933 To build the instrumented user library:
2935 . In path:{emon.c}, before including path:{tpp.h}, add the
2941 #define TRACEPOINT_DEFINE
2945 . Compile the user library source file:
2950 $ gcc -I. -fpic -c emon.c
2954 . Build the user library shared object:
2959 $ gcc -shared -o libemon.so emon.o tpp.o -llttng-ust -ldl
2963 To build the application:
2965 . Compile the application source file:
2974 . Build the application:
2979 $ gcc -o app app.o -L. -lemon
2983 To run the application:
2985 * Start the application:
2995 The application is linked with the instrumented user library.
2997 The instrumented user library is linked with the tracepoint provider
2998 package shared object.
3000 image::ust-sit+app-linked-with-lib+lib-linked-with-tp-so+lib-instrumented.png[]
3003 include::../common/ust-sit-step-tp-so.txt[]
3005 To build the instrumented user library:
3007 . In path:{emon.c}, before including path:{tpp.h}, add the
3013 #define TRACEPOINT_DEFINE
3017 . Compile the user library source file:
3022 $ gcc -I. -fpic -c emon.c
3026 . Build the user library shared object:
3031 $ gcc -shared -o libemon.so emon.o -ldl -L. -ltpp
3035 To build the application:
3037 . Compile the application source file:
3046 . Build the application:
3051 $ gcc -o app app.o -L. -lemon
3055 To run the application:
3057 * Start the application:
3067 The tracepoint provider package shared object is preloaded before the
3070 The application is linked with the instrumented user library.
3072 image::ust-sit+tp-so-preloaded+app-linked-with-lib+lib-instrumented.png[]
3075 include::../common/ust-sit-step-tp-so.txt[]
3077 To build the instrumented user library:
3079 . In path:{emon.c}, before including path:{tpp.h}, add the
3085 #define TRACEPOINT_DEFINE
3086 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3090 . Compile the user library source file:
3095 $ gcc -I. -fpic -c emon.c
3099 . Build the user library shared object:
3104 $ gcc -shared -o libemon.so emon.o -ldl
3108 To build the application:
3110 . Compile the application source file:
3119 . Build the application:
3124 $ gcc -o app app.o -L. -lemon
3128 To run the application with tracing support:
3130 * Preload the tracepoint provider package shared object and
3131 start the application:
3136 $ LD_PRELOAD=./libtpp.so ./app
3140 To run the application without tracing support:
3142 * Start the application:
3152 The application is linked with the instrumented user library.
3154 The instrumented user library dynamically loads the tracepoint provider
3155 package shared object.
3157 See the <<dlclose-warning,warning about `dlclose()`>>.
3159 image::ust-sit+app-linked-with-lib+lib-dlopens-tp-so+lib-instrumented.png[]
3162 include::../common/ust-sit-step-tp-so.txt[]
3164 To build the instrumented user library:
3166 . In path:{emon.c}, before including path:{tpp.h}, add the
3172 #define TRACEPOINT_DEFINE
3173 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3177 . Compile the user library source file:
3182 $ gcc -I. -fpic -c emon.c
3186 . Build the user library shared object:
3191 $ gcc -shared -o libemon.so emon.o -ldl
3195 To build the application:
3197 . Compile the application source file:
3206 . Build the application:
3211 $ gcc -o app app.o -L. -lemon
3215 To run the application:
3217 * Start the application:
3227 The application dynamically loads the instrumented user library.
3229 The instrumented user library is linked with the tracepoint provider
3230 package shared object.
3232 See the <<dlclose-warning,warning about `dlclose()`>>.
3234 image::ust-sit+app-dlopens-lib+lib-linked-with-tp-so+lib-instrumented.png[]
3237 include::../common/ust-sit-step-tp-so.txt[]
3239 To build the instrumented user library:
3241 . In path:{emon.c}, before including path:{tpp.h}, add the
3247 #define TRACEPOINT_DEFINE
3251 . Compile the user library source file:
3256 $ gcc -I. -fpic -c emon.c
3260 . Build the user library shared object:
3265 $ gcc -shared -o libemon.so emon.o -ldl -L. -ltpp
3269 To build the application:
3271 . Compile the application source file:
3280 . Build the application:
3285 $ gcc -o app app.o -ldl -L. -lemon
3289 To run the application:
3291 * Start the application:
3301 The application dynamically loads the instrumented user library.
3303 The instrumented user library dynamically loads the tracepoint provider
3304 package shared object.
3306 See the <<dlclose-warning,warning about `dlclose()`>>.
3308 image::ust-sit+app-dlopens-lib+lib-dlopens-tp-so+lib-instrumented.png[]
3311 include::../common/ust-sit-step-tp-so.txt[]
3313 To build the instrumented user library:
3315 . In path:{emon.c}, before including path:{tpp.h}, add the
3321 #define TRACEPOINT_DEFINE
3322 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3326 . Compile the user library source file:
3331 $ gcc -I. -fpic -c emon.c
3335 . Build the user library shared object:
3340 $ gcc -shared -o libemon.so emon.o -ldl
3344 To build the application:
3346 . Compile the application source file:
3355 . Build the application:
3360 $ gcc -o app app.o -ldl -L. -lemon
3364 To run the application:
3366 * Start the application:
3376 The tracepoint provider package shared object is preloaded before the
3379 The application dynamically loads the instrumented user library.
3381 image::ust-sit+tp-so-preloaded+app-dlopens-lib+lib-instrumented.png[]
3384 include::../common/ust-sit-step-tp-so.txt[]
3386 To build the instrumented user library:
3388 . In path:{emon.c}, before including path:{tpp.h}, add the
3394 #define TRACEPOINT_DEFINE
3395 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3399 . Compile the user library source file:
3404 $ gcc -I. -fpic -c emon.c
3408 . Build the user library shared object:
3413 $ gcc -shared -o libemon.so emon.o -ldl
3417 To build the application:
3419 . Compile the application source file:
3428 . Build the application:
3433 $ gcc -o app app.o -L. -lemon
3437 To run the application with tracing support:
3439 * Preload the tracepoint provider package shared object and
3440 start the application:
3445 $ LD_PRELOAD=./libtpp.so ./app
3449 To run the application without tracing support:
3451 * Start the application:
3461 The application is statically linked with the tracepoint provider
3462 package object file.
3464 The application is linked with the instrumented user library.
3466 image::ust-sit+app-linked-with-tp-o+app-linked-with-lib+lib-instrumented.png[]
3469 include::../common/ust-sit-step-tp-o.txt[]
3471 To build the instrumented user library:
3473 . In path:{emon.c}, before including path:{tpp.h}, add the
3479 #define TRACEPOINT_DEFINE
3483 . Compile the user library source file:
3488 $ gcc -I. -fpic -c emon.c
3492 . Build the user library shared object:
3497 $ gcc -shared -o libemon.so emon.o
3501 To build the application:
3503 . Compile the application source file:
3512 . Build the application:
3517 $ gcc -o app app.o tpp.o -llttng-ust -ldl -L. -lemon
3521 To run the instrumented application:
3523 * Start the application:
3533 The application is statically linked with the tracepoint provider
3534 package object file.
3536 The application dynamically loads the instrumented user library.
3538 image::ust-sit+app-linked-with-tp-o+app-dlopens-lib+lib-instrumented.png[]
3541 include::../common/ust-sit-step-tp-o.txt[]
3543 To build the application:
3545 . In path:{app.c}, before including path:{tpp.h}, add the following line:
3550 #define TRACEPOINT_DEFINE
3554 . Compile the application source file:
3563 . Build the application:
3568 $ gcc -Wl,--export-dynamic -o app app.o tpp.o \
3573 The `--export-dynamic` option passed to the linker is necessary for the
3574 dynamically loaded library to ``see'' the tracepoint symbols defined in
3577 To build the instrumented user library:
3579 . Compile the user library source file:
3584 $ gcc -I. -fpic -c emon.c
3588 . Build the user library shared object:
3593 $ gcc -shared -o libemon.so emon.o
3597 To run the application:
3599 * Start the application:
3611 .Do not use man:dlclose(3) on a tracepoint provider package
3613 Never use man:dlclose(3) on any shared object which:
3615 * Is linked with, statically or dynamically, a tracepoint provider
3617 * Calls man:dlopen(3) itself to dynamically open a tracepoint provider
3618 package shared object.
3620 This is currently considered **unsafe** due to a lack of reference
3621 counting from LTTng-UST to the shared object.
3623 A known workaround (available since glibc 2.2) is to use the
3624 `RTLD_NODELETE` flag when calling man:dlopen(3) initially. This has the
3625 effect of not unloading the loaded shared object, even if man:dlclose(3)
3628 You can also preload the tracepoint provider package shared object with
3629 the env:LD_PRELOAD environment variable to overcome this limitation.
3633 [[using-lttng-ust-with-daemons]]
3634 ===== Use noch:{LTTng-UST} with daemons
3636 If your instrumented application calls man:fork(2), man:clone(2),
3637 or BSD's man:rfork(2), without a following man:exec(3)-family
3638 system call, you must preload the path:{liblttng-ust-fork.so} shared
3639 object when you start the application.
3643 $ LD_PRELOAD=liblttng-ust-fork.so ./my-app
3646 If your tracepoint provider package is
3647 a shared library which you also preload, you must put both
3648 shared objects in env:LD_PRELOAD:
3652 $ LD_PRELOAD=liblttng-ust-fork.so:/path/to/tp.so ./my-app
3658 ===== Use noch:{LTTng-UST} with applications which close file descriptors that don't belong to them
3660 If your instrumented application closes one or more file descriptors
3661 which it did not open itself, you must preload the
3662 path:{liblttng-ust-fd.so} shared object when you start the application:
3666 $ LD_PRELOAD=liblttng-ust-fd.so ./my-app
3669 Typical use cases include closing all the file descriptors after
3670 man:fork(2) or man:rfork(2) and buggy applications doing
3674 [[lttng-ust-pkg-config]]
3675 ===== Use noch:{pkg-config}
3677 On some distributions, LTTng-UST ships with a
3678 https://www.freedesktop.org/wiki/Software/pkg-config/[pkg-config]
3679 metadata file. If this is your case, then you can use cmd:pkg-config to
3680 build an application on the command line:
3684 $ gcc -o my-app my-app.o tp.o $(pkg-config --cflags --libs lttng-ust)
3688 [[instrumenting-32-bit-app-on-64-bit-system]]
3689 ===== [[advanced-instrumenting-techniques]]Build a 32-bit instrumented application for a 64-bit target system
3691 In order to trace a 32-bit application running on a 64-bit system,
3692 LTTng must use a dedicated 32-bit
3693 <<lttng-consumerd,consumer daemon>>.
3695 The following steps show how to build and install a 32-bit consumer
3696 daemon, which is _not_ part of the default 64-bit LTTng build, how to
3697 build and install the 32-bit LTTng-UST libraries, and how to build and
3698 link an instrumented 32-bit application in that context.
3700 To build a 32-bit instrumented application for a 64-bit target system,
3701 assuming you have a fresh target system with no installed Userspace RCU
3704 . Download, build, and install a 32-bit version of Userspace RCU:
3709 $ cd $(mktemp -d) &&
3710 wget http://lttng.org/files/urcu/userspace-rcu-latest-0.9.tar.bz2 &&
3711 tar -xf userspace-rcu-latest-0.9.tar.bz2 &&
3712 cd userspace-rcu-0.9.* &&
3713 ./configure --libdir=/usr/local/lib32 CFLAGS=-m32 &&
3715 sudo make install &&
3720 . Using your distribution's package manager, or from source, install
3721 the following 32-bit versions of the following dependencies of
3722 LTTng-tools and LTTng-UST:
3725 * https://sourceforge.net/projects/libuuid/[libuuid]
3726 * http://directory.fsf.org/wiki/Popt[popt]
3727 * http://www.xmlsoft.org/[libxml2]
3730 . Download, build, and install a 32-bit version of the latest
3731 LTTng-UST{nbsp}{revision}:
3736 $ cd $(mktemp -d) &&
3737 wget http://lttng.org/files/lttng-ust/lttng-ust-latest-2.10.tar.bz2 &&
3738 tar -xf lttng-ust-latest-2.10.tar.bz2 &&
3739 cd lttng-ust-2.10.* &&
3740 ./configure --libdir=/usr/local/lib32 \
3741 CFLAGS=-m32 CXXFLAGS=-m32 \
3742 LDFLAGS='-L/usr/local/lib32 -L/usr/lib32' &&
3744 sudo make install &&
3751 Depending on your distribution,
3752 32-bit libraries could be installed at a different location than
3753 `/usr/lib32`. For example, Debian is known to install
3754 some 32-bit libraries in `/usr/lib/i386-linux-gnu`.
3756 In this case, make sure to set `LDFLAGS` to all the
3757 relevant 32-bit library paths, for example:
3761 $ LDFLAGS='-L/usr/lib/i386-linux-gnu -L/usr/lib32'
3765 . Download the latest LTTng-tools{nbsp}{revision}, build, and install
3766 the 32-bit consumer daemon:
3771 $ cd $(mktemp -d) &&
3772 wget http://lttng.org/files/lttng-tools/lttng-tools-latest-2.10.tar.bz2 &&
3773 tar -xf lttng-tools-latest-2.10.tar.bz2 &&
3774 cd lttng-tools-2.10.* &&
3775 ./configure --libdir=/usr/local/lib32 CFLAGS=-m32 CXXFLAGS=-m32 \
3776 LDFLAGS='-L/usr/local/lib32 -L/usr/lib32' \
3777 --disable-bin-lttng --disable-bin-lttng-crash \
3778 --disable-bin-lttng-relayd --disable-bin-lttng-sessiond &&
3780 cd src/bin/lttng-consumerd &&
3781 sudo make install &&
3786 . From your distribution or from source,
3787 <<installing-lttng,install>> the 64-bit versions of
3788 LTTng-UST and Userspace RCU.
3789 . Download, build, and install the 64-bit version of the
3790 latest LTTng-tools{nbsp}{revision}:
3795 $ cd $(mktemp -d) &&
3796 wget http://lttng.org/files/lttng-tools/lttng-tools-latest-2.10.tar.bz2 &&
3797 tar -xf lttng-tools-latest-2.10.tar.bz2 &&
3798 cd lttng-tools-2.10.* &&
3799 ./configure --with-consumerd32-libdir=/usr/local/lib32 \
3800 --with-consumerd32-bin=/usr/local/lib32/lttng/libexec/lttng-consumerd &&
3802 sudo make install &&
3807 . Pass the following options to man:gcc(1), man:g++(1), or man:clang(1)
3808 when linking your 32-bit application:
3811 -m32 -L/usr/lib32 -L/usr/local/lib32 \
3812 -Wl,-rpath,/usr/lib32,-rpath,/usr/local/lib32
3815 For example, let's rebuild the quick start example in
3816 <<tracing-your-own-user-application,Trace a user application>> as an
3817 instrumented 32-bit application:
3822 $ gcc -m32 -c -I. hello-tp.c
3823 $ gcc -m32 -c hello.c
3824 $ gcc -m32 -o hello hello.o hello-tp.o \
3825 -L/usr/lib32 -L/usr/local/lib32 \
3826 -Wl,-rpath,/usr/lib32,-rpath,/usr/local/lib32 \
3831 No special action is required to execute the 32-bit application and
3832 to trace it: use the command-line man:lttng(1) tool as usual.
3839 man:tracef(3) is a small LTTng-UST API designed for quick,
3840 man:printf(3)-like instrumentation without the burden of
3841 <<tracepoint-provider,creating>> and
3842 <<building-tracepoint-providers-and-user-application,building>>
3843 a tracepoint provider package.
3845 To use `tracef()` in your application:
3847 . In the C or C++ source files where you need to use `tracef()`,
3848 include `<lttng/tracef.h>`:
3853 #include <lttng/tracef.h>
3857 . In the application's source code, use `tracef()` like you would use
3865 tracef("my message: %d (%s)", my_integer, my_string);
3871 . Link your application with `liblttng-ust`:
3876 $ gcc -o app app.c -llttng-ust
3880 To trace the events that `tracef()` calls emit:
3882 * <<enabling-disabling-events,Create an event rule>> which matches the
3883 `lttng_ust_tracef:*` event name:
3888 $ lttng enable-event --userspace 'lttng_ust_tracef:*'
3893 .Limitations of `tracef()`
3895 The `tracef()` utility function was developed to make user space tracing
3896 super simple, albeit with notable disadvantages compared to
3897 <<defining-tracepoints,user-defined tracepoints>>:
3899 * All the emitted events have the same tracepoint provider and
3900 tracepoint names, respectively `lttng_ust_tracef` and `event`.
3901 * There is no static type checking.
3902 * The only event record field you actually get, named `msg`, is a string
3903 potentially containing the values you passed to `tracef()`
3904 using your own format string. This also means that you cannot filter
3905 events with a custom expression at run time because there are no
3907 * Since `tracef()` uses the C standard library's man:vasprintf(3)
3908 function behind the scenes to format the strings at run time, its
3909 expected performance is lower than with user-defined tracepoints,
3910 which do not require a conversion to a string.
3912 Taking this into consideration, `tracef()` is useful for some quick
3913 prototyping and debugging, but you should not consider it for any
3914 permanent and serious applicative instrumentation.
3920 ==== Use `tracelog()`
3922 The man:tracelog(3) API is very similar to <<tracef,`tracef()`>>, with
3923 the difference that it accepts an additional log level parameter.
3925 The goal of `tracelog()` is to ease the migration from logging to
3928 To use `tracelog()` in your application:
3930 . In the C or C++ source files where you need to use `tracelog()`,
3931 include `<lttng/tracelog.h>`:
3936 #include <lttng/tracelog.h>
3940 . In the application's source code, use `tracelog()` like you would use
3941 man:printf(3), except for the first parameter which is the log
3949 tracelog(TRACE_WARNING, "my message: %d (%s)",
3950 my_integer, my_string);
3956 See man:lttng-ust(3) for a list of available log level names.
3958 . Link your application with `liblttng-ust`:
3963 $ gcc -o app app.c -llttng-ust
3967 To trace the events that `tracelog()` calls emit with a log level
3968 _as severe as_ a specific log level:
3970 * <<enabling-disabling-events,Create an event rule>> which matches the
3971 `lttng_ust_tracelog:*` event name and a minimum level
3977 $ lttng enable-event --userspace 'lttng_ust_tracelog:*'
3978 --loglevel=TRACE_WARNING
3982 To trace the events that `tracelog()` calls emit with a
3983 _specific log level_:
3985 * Create an event rule which matches the `lttng_ust_tracelog:*`
3986 event name and a specific log level:
3991 $ lttng enable-event --userspace 'lttng_ust_tracelog:*'
3992 --loglevel-only=TRACE_INFO
3997 [[prebuilt-ust-helpers]]
3998 === Prebuilt user space tracing helpers
4000 The LTTng-UST package provides a few helpers in the form or preloadable
4001 shared objects which automatically instrument system functions and
4004 The helper shared objects are normally found in dir:{/usr/lib}. If you
4005 built LTTng-UST <<building-from-source,from source>>, they are probably
4006 located in dir:{/usr/local/lib}.
4008 The installed user space tracing helpers in LTTng-UST{nbsp}{revision}
4011 path:{liblttng-ust-libc-wrapper.so}::
4012 path:{liblttng-ust-pthread-wrapper.so}::
4013 <<liblttng-ust-libc-pthread-wrapper,C{nbsp}standard library
4014 memory and POSIX threads function tracing>>.
4016 path:{liblttng-ust-cyg-profile.so}::
4017 path:{liblttng-ust-cyg-profile-fast.so}::
4018 <<liblttng-ust-cyg-profile,Function entry and exit tracing>>.
4020 path:{liblttng-ust-dl.so}::
4021 <<liblttng-ust-dl,Dynamic linker tracing>>.
4023 To use a user space tracing helper with any user application:
4025 * Preload the helper shared object when you start the application:
4030 $ LD_PRELOAD=liblttng-ust-libc-wrapper.so my-app
4034 You can preload more than one helper:
4039 $ LD_PRELOAD=liblttng-ust-libc-wrapper.so:liblttng-ust-dl.so my-app
4045 [[liblttng-ust-libc-pthread-wrapper]]
4046 ==== Instrument C standard library memory and POSIX threads functions
4048 The path:{liblttng-ust-libc-wrapper.so} and
4049 path:{liblttng-ust-pthread-wrapper.so} helpers
4050 add instrumentation to some C standard library and POSIX
4054 .Functions instrumented by preloading path:{liblttng-ust-libc-wrapper.so}.
4056 |TP provider name |TP name |Instrumented function
4058 .6+|`lttng_ust_libc` |`malloc` |man:malloc(3)
4059 |`calloc` |man:calloc(3)
4060 |`realloc` |man:realloc(3)
4061 |`free` |man:free(3)
4062 |`memalign` |man:memalign(3)
4063 |`posix_memalign` |man:posix_memalign(3)
4067 .Functions instrumented by preloading path:{liblttng-ust-pthread-wrapper.so}.
4069 |TP provider name |TP name |Instrumented function
4071 .4+|`lttng_ust_pthread` |`pthread_mutex_lock_req` |man:pthread_mutex_lock(3p) (request time)
4072 |`pthread_mutex_lock_acq` |man:pthread_mutex_lock(3p) (acquire time)
4073 |`pthread_mutex_trylock` |man:pthread_mutex_trylock(3p)
4074 |`pthread_mutex_unlock` |man:pthread_mutex_unlock(3p)
4077 When you preload the shared object, it replaces the functions listed
4078 in the previous tables by wrappers which contain tracepoints and call
4079 the replaced functions.
4082 [[liblttng-ust-cyg-profile]]
4083 ==== Instrument function entry and exit
4085 The path:{liblttng-ust-cyg-profile*.so} helpers can add instrumentation
4086 to the entry and exit points of functions.
4088 man:gcc(1) and man:clang(1) have an option named
4089 https://gcc.gnu.org/onlinedocs/gcc/Instrumentation-Options.html[`-finstrument-functions`]
4090 which generates instrumentation calls for entry and exit to functions.
4091 The LTTng-UST function tracing helpers,
4092 path:{liblttng-ust-cyg-profile.so} and
4093 path:{liblttng-ust-cyg-profile-fast.so}, take advantage of this feature
4094 to add tracepoints to the two generated functions (which contain
4095 `cyg_profile` in their names, hence the helper's name).
4097 To use the LTTng-UST function tracing helper, the source files to
4098 instrument must be built using the `-finstrument-functions` compiler
4101 There are two versions of the LTTng-UST function tracing helper:
4103 * **path:{liblttng-ust-cyg-profile-fast.so}** is a lightweight variant
4104 that you should only use when it can be _guaranteed_ that the
4105 complete event stream is recorded without any lost event record.
4106 Any kind of duplicate information is left out.
4108 Assuming no event record is lost, having only the function addresses on
4109 entry is enough to create a call graph, since an event record always
4110 contains the ID of the CPU that generated it.
4112 You can use a tool like man:addr2line(1) to convert function addresses
4113 back to source file names and line numbers.
4115 * **path:{liblttng-ust-cyg-profile.so}** is a more robust variant
4116 which also works in use cases where event records might get discarded or
4117 not recorded from application startup.
4118 In these cases, the trace analyzer needs more information to be
4119 able to reconstruct the program flow.
4121 See man:lttng-ust-cyg-profile(3) to learn more about the instrumentation
4122 points of this helper.
4124 All the tracepoints that this helper provides have the
4125 log level `TRACE_DEBUG_FUNCTION` (see man:lttng-ust(3)).
4127 TIP: It's sometimes a good idea to limit the number of source files that
4128 you compile with the `-finstrument-functions` option to prevent LTTng
4129 from writing an excessive amount of trace data at run time. When using
4130 man:gcc(1), you can use the
4131 `-finstrument-functions-exclude-function-list` option to avoid
4132 instrument entries and exits of specific function names.
4137 ==== Instrument the dynamic linker
4139 The path:{liblttng-ust-dl.so} helper adds instrumentation to the
4140 man:dlopen(3) and man:dlclose(3) function calls.
4142 See man:lttng-ust-dl(3) to learn more about the instrumentation points
4147 [[java-application]]
4148 === User space Java agent
4150 You can instrument any Java application which uses one of the following
4153 * The https://docs.oracle.com/javase/7/docs/api/java/util/logging/package-summary.html[**`java.util.logging`**]
4154 (JUL) core logging facilities.
4155 * http://logging.apache.org/log4j/1.2/[**Apache log4j 1.2**], since
4156 LTTng 2.6. Note that Apache Log4j{nbsp}2 is not supported.
4159 .LTTng-UST Java agent imported by a Java application.
4160 image::java-app.png[]
4162 Note that the methods described below are new in LTTng{nbsp}{revision}.
4163 Previous LTTng versions use another technique.
4165 NOTE: We use http://openjdk.java.net/[OpenJDK]{nbsp}8 for development
4166 and https://ci.lttng.org/[continuous integration], thus this version is
4167 directly supported. However, the LTTng-UST Java agent is also tested
4168 with OpenJDK{nbsp}7.
4173 ==== Use the LTTng-UST Java agent for `java.util.logging`
4175 To use the LTTng-UST Java agent in a Java application which uses
4176 `java.util.logging` (JUL):
4178 . In the Java application's source code, import the LTTng-UST
4179 log handler package for `java.util.logging`:
4184 import org.lttng.ust.agent.jul.LttngLogHandler;
4188 . Create an LTTng-UST JUL log handler:
4193 Handler lttngUstLogHandler = new LttngLogHandler();
4197 . Add this handler to the JUL loggers which should emit LTTng events:
4202 Logger myLogger = Logger.getLogger("some-logger");
4204 myLogger.addHandler(lttngUstLogHandler);
4208 . Use `java.util.logging` log statements and configuration as usual.
4209 The loggers with an attached LTTng-UST log handler can emit
4212 . Before exiting the application, remove the LTTng-UST log handler from
4213 the loggers attached to it and call its `close()` method:
4218 myLogger.removeHandler(lttngUstLogHandler);
4219 lttngUstLogHandler.close();
4223 This is not strictly necessary, but it is recommended for a clean
4224 disposal of the handler's resources.
4226 . Include the LTTng-UST Java agent's common and JUL-specific JAR files,
4227 path:{lttng-ust-agent-common.jar} and path:{lttng-ust-agent-jul.jar},
4229 https://docs.oracle.com/javase/tutorial/essential/environment/paths.html[class
4230 path] when you build the Java application.
4232 The JAR files are typically located in dir:{/usr/share/java}.
4234 IMPORTANT: The LTTng-UST Java agent must be
4235 <<installing-lttng,installed>> for the logging framework your
4238 .Use the LTTng-UST Java agent for `java.util.logging`.
4243 import java.io.IOException;
4244 import java.util.logging.Handler;
4245 import java.util.logging.Logger;
4246 import org.lttng.ust.agent.jul.LttngLogHandler;
4250 private static final int answer = 42;
4252 public static void main(String[] argv) throws Exception
4255 Logger logger = Logger.getLogger("jello");
4257 // Create an LTTng-UST log handler
4258 Handler lttngUstLogHandler = new LttngLogHandler();
4260 // Add the LTTng-UST log handler to our logger
4261 logger.addHandler(lttngUstLogHandler);
4264 logger.info("some info");
4265 logger.warning("some warning");
4267 logger.finer("finer information; the answer is " + answer);
4269 logger.severe("error!");
4271 // Not mandatory, but cleaner
4272 logger.removeHandler(lttngUstLogHandler);
4273 lttngUstLogHandler.close();
4282 $ javac -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-jul.jar Test.java
4285 <<creating-destroying-tracing-sessions,Create a tracing session>>,
4286 <<enabling-disabling-events,create an event rule>> matching the
4287 `jello` JUL logger, and <<basic-tracing-session-control,start tracing>>:
4292 $ lttng enable-event --jul jello
4296 Run the compiled class:
4300 $ java -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-jul.jar:. Test
4303 <<basic-tracing-session-control,Stop tracing>> and inspect the
4313 In the resulting trace, an <<event,event record>> generated by a Java
4314 application using `java.util.logging` is named `lttng_jul:event` and
4315 has the following fields:
4318 Log record's message.
4324 Name of the class in which the log statement was executed.
4327 Name of the method in which the log statement was executed.
4330 Logging time (timestamp in milliseconds).
4333 Log level integer value.
4336 ID of the thread in which the log statement was executed.
4338 You can use the opt:lttng-enable-event(1):--loglevel or
4339 opt:lttng-enable-event(1):--loglevel-only option of the
4340 man:lttng-enable-event(1) command to target a range of JUL log levels
4341 or a specific JUL log level.
4346 ==== Use the LTTng-UST Java agent for Apache log4j
4348 To use the LTTng-UST Java agent in a Java application which uses
4351 . In the Java application's source code, import the LTTng-UST
4352 log appender package for Apache log4j:
4357 import org.lttng.ust.agent.log4j.LttngLogAppender;
4361 . Create an LTTng-UST log4j log appender:
4366 Appender lttngUstLogAppender = new LttngLogAppender();
4370 . Add this appender to the log4j loggers which should emit LTTng events:
4375 Logger myLogger = Logger.getLogger("some-logger");
4377 myLogger.addAppender(lttngUstLogAppender);
4381 . Use Apache log4j log statements and configuration as usual. The
4382 loggers with an attached LTTng-UST log appender can emit LTTng events.
4384 . Before exiting the application, remove the LTTng-UST log appender from
4385 the loggers attached to it and call its `close()` method:
4390 myLogger.removeAppender(lttngUstLogAppender);
4391 lttngUstLogAppender.close();
4395 This is not strictly necessary, but it is recommended for a clean
4396 disposal of the appender's resources.
4398 . Include the LTTng-UST Java agent's common and log4j-specific JAR
4399 files, path:{lttng-ust-agent-common.jar} and
4400 path:{lttng-ust-agent-log4j.jar}, in the
4401 https://docs.oracle.com/javase/tutorial/essential/environment/paths.html[class
4402 path] when you build the Java application.
4404 The JAR files are typically located in dir:{/usr/share/java}.
4406 IMPORTANT: The LTTng-UST Java agent must be
4407 <<installing-lttng,installed>> for the logging framework your
4410 .Use the LTTng-UST Java agent for Apache log4j.
4415 import org.apache.log4j.Appender;
4416 import org.apache.log4j.Logger;
4417 import org.lttng.ust.agent.log4j.LttngLogAppender;
4421 private static final int answer = 42;
4423 public static void main(String[] argv) throws Exception
4426 Logger logger = Logger.getLogger("jello");
4428 // Create an LTTng-UST log appender
4429 Appender lttngUstLogAppender = new LttngLogAppender();
4431 // Add the LTTng-UST log appender to our logger
4432 logger.addAppender(lttngUstLogAppender);
4435 logger.info("some info");
4436 logger.warn("some warning");
4438 logger.debug("debug information; the answer is " + answer);
4440 logger.fatal("error!");
4442 // Not mandatory, but cleaner
4443 logger.removeAppender(lttngUstLogAppender);
4444 lttngUstLogAppender.close();
4450 Build this example (`$LOG4JPATH` is the path to the Apache log4j JAR
4455 $ javac -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-log4j.jar:$LOG4JPATH Test.java
4458 <<creating-destroying-tracing-sessions,Create a tracing session>>,
4459 <<enabling-disabling-events,create an event rule>> matching the
4460 `jello` log4j logger, and <<basic-tracing-session-control,start tracing>>:
4465 $ lttng enable-event --log4j jello
4469 Run the compiled class:
4473 $ java -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-log4j.jar:$LOG4JPATH:. Test
4476 <<basic-tracing-session-control,Stop tracing>> and inspect the
4486 In the resulting trace, an <<event,event record>> generated by a Java
4487 application using log4j is named `lttng_log4j:event` and
4488 has the following fields:
4491 Log record's message.
4497 Name of the class in which the log statement was executed.
4500 Name of the method in which the log statement was executed.
4503 Name of the file in which the executed log statement is located.
4506 Line number at which the log statement was executed.
4512 Log level integer value.
4515 Name of the Java thread in which the log statement was executed.
4517 You can use the opt:lttng-enable-event(1):--loglevel or
4518 opt:lttng-enable-event(1):--loglevel-only option of the
4519 man:lttng-enable-event(1) command to target a range of Apache log4j log levels
4520 or a specific log4j log level.
4524 [[java-application-context]]
4525 ==== Provide application-specific context fields in a Java application
4527 A Java application-specific context field is a piece of state provided
4528 by the application which <<adding-context,you can add>>, using the
4529 man:lttng-add-context(1) command, to each <<event,event record>>
4530 produced by the log statements of this application.
4532 For example, a given object might have a current request ID variable.
4533 You can create a context information retriever for this object and
4534 assign a name to this current request ID. You can then, using the
4535 man:lttng-add-context(1) command, add this context field by name to
4536 the JUL or log4j <<channel,channel>>.
4538 To provide application-specific context fields in a Java application:
4540 . In the Java application's source code, import the LTTng-UST
4541 Java agent context classes and interfaces:
4546 import org.lttng.ust.agent.context.ContextInfoManager;
4547 import org.lttng.ust.agent.context.IContextInfoRetriever;
4551 . Create a context information retriever class, that is, a class which
4552 implements the `IContextInfoRetriever` interface:
4557 class MyContextInfoRetriever implements IContextInfoRetriever
4560 public Object retrieveContextInfo(String key)
4562 if (key.equals("intCtx")) {
4564 } else if (key.equals("strContext")) {
4565 return "context value!";
4574 This `retrieveContextInfo()` method is the only member of the
4575 `IContextInfoRetriever` interface. Its role is to return the current
4576 value of a state by name to create a context field. The names of the
4577 context fields and which state variables they return depends on your
4580 All primitive types and objects are supported as context fields.
4581 When `retrieveContextInfo()` returns an object, the context field
4582 serializer calls its `toString()` method to add a string field to
4583 event records. The method can also return `null`, which means that
4584 no context field is available for the required name.
4586 . Register an instance of your context information retriever class to
4587 the context information manager singleton:
4592 IContextInfoRetriever cir = new MyContextInfoRetriever();
4593 ContextInfoManager cim = ContextInfoManager.getInstance();
4594 cim.registerContextInfoRetriever("retrieverName", cir);
4598 . Before exiting the application, remove your context information
4599 retriever from the context information manager singleton:
4604 ContextInfoManager cim = ContextInfoManager.getInstance();
4605 cim.unregisterContextInfoRetriever("retrieverName");
4609 This is not strictly necessary, but it is recommended for a clean
4610 disposal of some manager's resources.
4612 . Build your Java application with LTTng-UST Java agent support as
4613 usual, following the procedure for either the <<jul,JUL>> or
4614 <<log4j,Apache log4j>> framework.
4617 .Provide application-specific context fields in a Java application.
4622 import java.util.logging.Handler;
4623 import java.util.logging.Logger;
4624 import org.lttng.ust.agent.jul.LttngLogHandler;
4625 import org.lttng.ust.agent.context.ContextInfoManager;
4626 import org.lttng.ust.agent.context.IContextInfoRetriever;
4630 // Our context information retriever class
4631 private static class MyContextInfoRetriever
4632 implements IContextInfoRetriever
4635 public Object retrieveContextInfo(String key) {
4636 if (key.equals("intCtx")) {
4638 } else if (key.equals("strContext")) {
4639 return "context value!";
4646 private static final int answer = 42;
4648 public static void main(String args[]) throws Exception
4650 // Get the context information manager instance
4651 ContextInfoManager cim = ContextInfoManager.getInstance();
4653 // Create and register our context information retriever
4654 IContextInfoRetriever cir = new MyContextInfoRetriever();
4655 cim.registerContextInfoRetriever("myRetriever", cir);
4658 Logger logger = Logger.getLogger("jello");
4660 // Create an LTTng-UST log handler
4661 Handler lttngUstLogHandler = new LttngLogHandler();
4663 // Add the LTTng-UST log handler to our logger
4664 logger.addHandler(lttngUstLogHandler);
4667 logger.info("some info");
4668 logger.warning("some warning");
4670 logger.finer("finer information; the answer is " + answer);
4672 logger.severe("error!");
4674 // Not mandatory, but cleaner
4675 logger.removeHandler(lttngUstLogHandler);
4676 lttngUstLogHandler.close();
4677 cim.unregisterContextInfoRetriever("myRetriever");
4686 $ javac -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-jul.jar Test.java
4689 <<creating-destroying-tracing-sessions,Create a tracing session>>
4690 and <<enabling-disabling-events,create an event rule>> matching the
4696 $ lttng enable-event --jul jello
4699 <<adding-context,Add the application-specific context fields>> to the
4704 $ lttng add-context --jul --type='$app.myRetriever:intCtx'
4705 $ lttng add-context --jul --type='$app.myRetriever:strContext'
4708 <<basic-tracing-session-control,Start tracing>>:
4715 Run the compiled class:
4719 $ java -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-jul.jar:. Test
4722 <<basic-tracing-session-control,Stop tracing>> and inspect the
4734 [[python-application]]
4735 === User space Python agent
4737 You can instrument a Python 2 or Python 3 application which uses the
4738 standard https://docs.python.org/3/library/logging.html[`logging`]
4741 Each log statement emits an LTTng event once the
4742 application module imports the
4743 <<lttng-ust-agents,LTTng-UST Python agent>> package.
4746 .A Python application importing the LTTng-UST Python agent.
4747 image::python-app.png[]
4749 To use the LTTng-UST Python agent:
4751 . In the Python application's source code, import the LTTng-UST Python
4761 The LTTng-UST Python agent automatically adds its logging handler to the
4762 root logger at import time.
4764 Any log statement that the application executes before this import does
4765 not emit an LTTng event.
4767 IMPORTANT: The LTTng-UST Python agent must be
4768 <<installing-lttng,installed>>.
4770 . Use log statements and logging configuration as usual.
4771 Since the LTTng-UST Python agent adds a handler to the _root_
4772 logger, you can trace any log statement from any logger.
4774 .Use the LTTng-UST Python agent.
4785 logging.basicConfig()
4786 logger = logging.getLogger('my-logger')
4789 logger.debug('debug message')
4790 logger.info('info message')
4791 logger.warn('warn message')
4792 logger.error('error message')
4793 logger.critical('critical message')
4797 if __name__ == '__main__':
4801 NOTE: `logging.basicConfig()`, which adds to the root logger a basic
4802 logging handler which prints to the standard error stream, is not
4803 strictly required for LTTng-UST tracing to work, but in versions of
4804 Python preceding 3.2, you could see a warning message which indicates
4805 that no handler exists for the logger `my-logger`.
4807 <<creating-destroying-tracing-sessions,Create a tracing session>>,
4808 <<enabling-disabling-events,create an event rule>> matching the
4809 `my-logger` Python logger, and <<basic-tracing-session-control,start
4815 $ lttng enable-event --python my-logger
4819 Run the Python script:
4826 <<basic-tracing-session-control,Stop tracing>> and inspect the recorded
4836 In the resulting trace, an <<event,event record>> generated by a Python
4837 application is named `lttng_python:event` and has the following fields:
4840 Logging time (string).
4843 Log record's message.
4849 Name of the function in which the log statement was executed.
4852 Line number at which the log statement was executed.
4855 Log level integer value.
4858 ID of the Python thread in which the log statement was executed.
4861 Name of the Python thread in which the log statement was executed.
4863 You can use the opt:lttng-enable-event(1):--loglevel or
4864 opt:lttng-enable-event(1):--loglevel-only option of the
4865 man:lttng-enable-event(1) command to target a range of Python log levels
4866 or a specific Python log level.
4868 When an application imports the LTTng-UST Python agent, the agent tries
4869 to register to a <<lttng-sessiond,session daemon>>. Note that you must
4870 <<start-sessiond,start the session daemon>> _before_ you run the Python
4871 application. If a session daemon is found, the agent tries to register
4872 to it during 5{nbsp}seconds, after which the application continues
4873 without LTTng tracing support. You can override this timeout value with
4874 the env:LTTNG_UST_PYTHON_REGISTER_TIMEOUT environment variable
4877 If the session daemon stops while a Python application with an imported
4878 LTTng-UST Python agent runs, the agent retries to connect and to
4879 register to a session daemon every 3{nbsp}seconds. You can override this
4880 delay with the env:LTTNG_UST_PYTHON_REGISTER_RETRY_DELAY environment
4885 [[proc-lttng-logger-abi]]
4888 The `lttng-tracer` Linux kernel module, part of
4889 <<lttng-modules,LTTng-modules>>, creates the special LTTng logger file
4890 path:{/proc/lttng-logger} when it's loaded. Any application can write
4891 text data to this file to emit an LTTng event.
4894 .An application writes to the LTTng logger file to emit an LTTng event.
4895 image::lttng-logger.png[]
4897 The LTTng logger is the quickest method--not the most efficient,
4898 however--to add instrumentation to an application. It is designed
4899 mostly to instrument shell scripts:
4903 $ echo "Some message, some $variable" > /proc/lttng-logger
4906 Any event that the LTTng logger emits is named `lttng_logger` and
4907 belongs to the Linux kernel <<domain,tracing domain>>. However, unlike
4908 other instrumentation points in the kernel tracing domain, **any Unix
4909 user** can <<enabling-disabling-events,create an event rule>> which
4910 matches its event name, not only the root user or users in the
4911 <<tracing-group,tracing group>>.
4913 To use the LTTng logger:
4915 * From any application, write text data to the path:{/proc/lttng-logger}
4918 The `msg` field of `lttng_logger` event records contains the
4921 NOTE: The maximum message length of an LTTng logger event is
4922 1024{nbsp}bytes. Writing more than this makes the LTTng logger emit more
4923 than one event to contain the remaining data.
4925 You should not use the LTTng logger to trace a user application which
4926 can be instrumented in a more efficient way, namely:
4928 * <<c-application,C and $$C++$$ applications>>.
4929 * <<java-application,Java applications>>.
4930 * <<python-application,Python applications>>.
4932 .Use the LTTng logger.
4937 echo 'Hello, World!' > /proc/lttng-logger
4939 df --human-readable --print-type / > /proc/lttng-logger
4942 <<creating-destroying-tracing-sessions,Create a tracing session>>,
4943 <<enabling-disabling-events,create an event rule>> matching the
4944 `lttng_logger` Linux kernel tracepoint, and
4945 <<basic-tracing-session-control,start tracing>>:
4950 $ lttng enable-event --kernel lttng_logger
4954 Run the Bash script:
4961 <<basic-tracing-session-control,Stop tracing>> and inspect the recorded
4972 [[instrumenting-linux-kernel]]
4973 === LTTng kernel tracepoints
4975 NOTE: This section shows how to _add_ instrumentation points to the
4976 Linux kernel. The kernel's subsystems are already thoroughly
4977 instrumented at strategic places for LTTng when you
4978 <<installing-lttng,install>> the <<lttng-modules,LTTng-modules>>
4982 There are two methods to instrument the Linux kernel:
4984 . <<linux-add-lttng-layer,Add an LTTng layer>> over an existing ftrace
4985 tracepoint which uses the `TRACE_EVENT()` API.
4987 Choose this if you want to instrumentation a Linux kernel tree with an
4988 instrumentation point compatible with ftrace, perf, and SystemTap.
4990 . Use an <<linux-lttng-tracepoint-event,LTTng-only approach>> to
4991 instrument an out-of-tree kernel module.
4993 Choose this if you don't need ftrace, perf, or SystemTap support.
4997 [[linux-add-lttng-layer]]
4998 ==== [[instrumenting-linux-kernel-itself]][[mainline-trace-event]][[lttng-adaptation-layer]]Add an LTTng layer to an existing ftrace tracepoint
5000 This section shows how to add an LTTng layer to existing ftrace
5001 instrumentation using the `TRACE_EVENT()` API.
5003 This section does not document the `TRACE_EVENT()` macro. You can
5004 read the following articles to learn more about this API:
5006 * http://lwn.net/Articles/379903/[Using the TRACE_EVENT() macro (Part 1)]
5007 * http://lwn.net/Articles/381064/[Using the TRACE_EVENT() macro (Part 2)]
5008 * http://lwn.net/Articles/383362/[Using the TRACE_EVENT() macro (Part 3)]
5010 The following procedure assumes that your ftrace tracepoints are
5011 correctly defined in their own header and that they are created in
5012 one source file using the `CREATE_TRACE_POINTS` definition.
5014 To add an LTTng layer over an existing ftrace tracepoint:
5016 . Make sure the following kernel configuration options are
5022 * `CONFIG_HIGH_RES_TIMERS`
5023 * `CONFIG_TRACEPOINTS`
5026 . Build the Linux source tree with your custom ftrace tracepoints.
5027 . Boot the resulting Linux image on your target system.
5029 Confirm that the tracepoints exist by looking for their names in the
5030 dir:{/sys/kernel/debug/tracing/events/subsys} directory, where `subsys`
5031 is your subsystem's name.
5033 . Get a copy of the latest LTTng-modules{nbsp}{revision}:
5038 $ cd $(mktemp -d) &&
5039 wget http://lttng.org/files/lttng-modules/lttng-modules-latest-2.10.tar.bz2 &&
5040 tar -xf lttng-modules-latest-2.10.tar.bz2 &&
5041 cd lttng-modules-2.10.*
5045 . In dir:{instrumentation/events/lttng-module}, relative to the root
5046 of the LTTng-modules source tree, create a header file named
5047 +__subsys__.h+ for your custom subsystem +__subsys__+ and write your
5048 LTTng-modules tracepoint definitions using the LTTng-modules
5051 Start with this template:
5055 .path:{instrumentation/events/lttng-module/my_subsys.h}
5058 #define TRACE_SYSTEM my_subsys
5060 #if !defined(_LTTNG_MY_SUBSYS_H) || defined(TRACE_HEADER_MULTI_READ)
5061 #define _LTTNG_MY_SUBSYS_H
5063 #include "../../../probes/lttng-tracepoint-event.h"
5064 #include <linux/tracepoint.h>
5066 LTTNG_TRACEPOINT_EVENT(
5068 * Format is identical to TRACE_EVENT()'s version for the three
5069 * following macro parameters:
5072 TP_PROTO(int my_int, const char *my_string),
5073 TP_ARGS(my_int, my_string),
5075 /* LTTng-modules specific macros */
5077 ctf_integer(int, my_int_field, my_int)
5078 ctf_string(my_bar_field, my_bar)
5082 #endif /* !defined(_LTTNG_MY_SUBSYS_H) || defined(TRACE_HEADER_MULTI_READ) */
5084 #include "../../../probes/define_trace.h"
5088 The entries in the `TP_FIELDS()` section are the list of fields for the
5089 LTTng tracepoint. This is similar to the `TP_STRUCT__entry()` part of
5090 ftrace's `TRACE_EVENT()` macro.
5092 See <<lttng-modules-tp-fields,Tracepoint fields macros>> for a
5093 complete description of the available `ctf_*()` macros.
5095 . Create the LTTng-modules probe's kernel module C source file,
5096 +probes/lttng-probe-__subsys__.c+, where +__subsys__+ is your
5101 .path:{probes/lttng-probe-my-subsys.c}
5103 #include <linux/module.h>
5104 #include "../lttng-tracer.h"
5107 * Build-time verification of mismatch between mainline
5108 * TRACE_EVENT() arguments and the LTTng-modules adaptation
5109 * layer LTTNG_TRACEPOINT_EVENT() arguments.
5111 #include <trace/events/my_subsys.h>
5113 /* Create LTTng tracepoint probes */
5114 #define LTTNG_PACKAGE_BUILD
5115 #define CREATE_TRACE_POINTS
5116 #define TRACE_INCLUDE_PATH ../instrumentation/events/lttng-module
5118 #include "../instrumentation/events/lttng-module/my_subsys.h"
5120 MODULE_LICENSE("GPL and additional rights");
5121 MODULE_AUTHOR("Your name <your-email>");
5122 MODULE_DESCRIPTION("LTTng my_subsys probes");
5123 MODULE_VERSION(__stringify(LTTNG_MODULES_MAJOR_VERSION) "."
5124 __stringify(LTTNG_MODULES_MINOR_VERSION) "."
5125 __stringify(LTTNG_MODULES_PATCHLEVEL_VERSION)
5126 LTTNG_MODULES_EXTRAVERSION);
5130 . Edit path:{probes/KBuild} and add your new kernel module object
5131 next to the existing ones:
5135 .path:{probes/KBuild}
5139 obj-m += lttng-probe-module.o
5140 obj-m += lttng-probe-power.o
5142 obj-m += lttng-probe-my-subsys.o
5148 . Build and install the LTTng kernel modules:
5153 $ make KERNELDIR=/path/to/linux
5154 # make modules_install && depmod -a
5158 Replace `/path/to/linux` with the path to the Linux source tree where
5159 you defined and used tracepoints with ftrace's `TRACE_EVENT()` macro.
5161 Note that you can also use the
5162 <<lttng-tracepoint-event-code,`LTTNG_TRACEPOINT_EVENT_CODE()` macro>>
5163 instead of `LTTNG_TRACEPOINT_EVENT()` to use custom local variables and
5164 C code that need to be executed before the event fields are recorded.
5166 The best way to learn how to use the previous LTTng-modules macros is to
5167 inspect the existing LTTng-modules tracepoint definitions in the
5168 dir:{instrumentation/events/lttng-module} header files. Compare them
5169 with the Linux kernel mainline versions in the
5170 dir:{include/trace/events} directory of the Linux source tree.
5174 [[lttng-tracepoint-event-code]]
5175 ===== Use custom C code to access the data for tracepoint fields
5177 Although we recommended to always use the
5178 <<lttng-adaptation-layer,`LTTNG_TRACEPOINT_EVENT()`>> macro to describe
5179 the arguments and fields of an LTTng-modules tracepoint when possible,
5180 sometimes you need a more complex process to access the data that the
5181 tracer records as event record fields. In other words, you need local
5182 variables and multiple C{nbsp}statements instead of simple
5183 argument-based expressions that you pass to the
5184 <<lttng-modules-tp-fields,`ctf_*()` macros of `TP_FIELDS()`>>.
5186 You can use the `LTTNG_TRACEPOINT_EVENT_CODE()` macro instead of
5187 `LTTNG_TRACEPOINT_EVENT()` to declare custom local variables and define
5188 a block of C{nbsp}code to be executed before LTTng records the fields.
5189 The structure of this macro is:
5192 .`LTTNG_TRACEPOINT_EVENT_CODE()` macro syntax.
5194 LTTNG_TRACEPOINT_EVENT_CODE(
5196 * Format identical to the LTTNG_TRACEPOINT_EVENT()
5197 * version for the following three macro parameters:
5200 TP_PROTO(int my_int, const char *my_string),
5201 TP_ARGS(my_int, my_string),
5203 /* Declarations of custom local variables */
5206 unsigned long b = 0;
5207 const char *name = "(undefined)";
5208 struct my_struct *my_struct;
5212 * Custom code which uses both tracepoint arguments
5213 * (in TP_ARGS()) and local variables (in TP_locvar()).
5215 * Local variables are actually members of a structure pointed
5216 * to by the special variable tp_locvar.
5220 tp_locvar->a = my_int + 17;
5221 tp_locvar->my_struct = get_my_struct_at(tp_locvar->a);
5222 tp_locvar->b = my_struct_compute_b(tp_locvar->my_struct);
5223 tp_locvar->name = my_struct_get_name(tp_locvar->my_struct);
5224 put_my_struct(tp_locvar->my_struct);
5233 * Format identical to the LTTNG_TRACEPOINT_EVENT()
5234 * version for this, except that tp_locvar members can be
5235 * used in the argument expression parameters of
5236 * the ctf_*() macros.
5239 ctf_integer(unsigned long, my_struct_b, tp_locvar->b)
5240 ctf_integer(int, my_struct_a, tp_locvar->a)
5241 ctf_string(my_string_field, my_string)
5242 ctf_string(my_struct_name, tp_locvar->name)
5247 IMPORTANT: The C code defined in `TP_code()` must not have any side
5248 effects when executed. In particular, the code must not allocate
5249 memory or get resources without deallocating this memory or putting
5250 those resources afterwards.
5253 [[instrumenting-linux-kernel-tracing]]
5254 ==== Load and unload a custom probe kernel module
5256 You must load a <<lttng-adaptation-layer,created LTTng-modules probe
5257 kernel module>> in the kernel before it can emit LTTng events.
5259 To load the default probe kernel modules and a custom probe kernel
5262 * Use the opt:lttng-sessiond(8):--extra-kmod-probes option to give extra
5263 probe modules to load when starting a root <<lttng-sessiond,session
5267 .Load the `my_subsys`, `usb`, and the default probe modules.
5271 # lttng-sessiond --extra-kmod-probes=my_subsys,usb
5276 You only need to pass the subsystem name, not the whole kernel module
5279 To load _only_ a given custom probe kernel module:
5281 * Use the opt:lttng-sessiond(8):--kmod-probes option to give the probe
5282 modules to load when starting a root session daemon:
5285 .Load only the `my_subsys` and `usb` probe modules.
5289 # lttng-sessiond --kmod-probes=my_subsys,usb
5294 To confirm that a probe module is loaded:
5301 $ lsmod | grep lttng_probe_usb
5305 To unload the loaded probe modules:
5307 * Kill the session daemon with `SIGTERM`:
5312 # pkill lttng-sessiond
5316 You can also use man:modprobe(8)'s `--remove` option if the session
5317 daemon terminates abnormally.
5320 [[controlling-tracing]]
5323 Once an application or a Linux kernel is
5324 <<instrumenting,instrumented>> for LTTng tracing,
5327 This section is divided in topics on how to use the various
5328 <<plumbing,components of LTTng>>, in particular the <<lttng-cli,cmd:lttng
5329 command-line tool>>, to _control_ the LTTng daemons and tracers.
5331 NOTE: In the following subsections, we refer to an man:lttng(1) command
5332 using its man page name. For example, instead of _Run the `create`
5333 command to..._, we use _Run the man:lttng-create(1) command to..._.
5337 === Start a session daemon
5339 In some situations, you need to run a <<lttng-sessiond,session daemon>>
5340 (man:lttng-sessiond(8)) _before_ you can use the man:lttng(1)
5343 You will see the following error when you run a command while no session
5347 Error: No session daemon is available
5350 The only command that automatically runs a session daemon is
5351 man:lttng-create(1), which you use to
5352 <<creating-destroying-tracing-sessions,create a tracing session>>. While
5353 this is most of the time the first operation that you do, sometimes it's
5354 not. Some examples are:
5356 * <<list-instrumentation-points,List the available instrumentation points>>.
5357 * <<saving-loading-tracing-session,Load a tracing session configuration>>.
5359 [[tracing-group]] Each Unix user must have its own running session
5360 daemon to trace user applications. The session daemon that the root user
5361 starts is the only one allowed to control the LTTng kernel tracer. Users
5362 that are part of the _tracing group_ can control the root session
5363 daemon. The default tracing group name is `tracing`; you can set it to
5364 something else with the opt:lttng-sessiond(8):--group option when you
5365 start the root session daemon.
5367 To start a user session daemon:
5369 * Run man:lttng-sessiond(8):
5374 $ lttng-sessiond --daemonize
5378 To start the root session daemon:
5380 * Run man:lttng-sessiond(8) as the root user:
5385 # lttng-sessiond --daemonize
5389 In both cases, remove the opt:lttng-sessiond(8):--daemonize option to
5390 start the session daemon in foreground.
5392 To stop a session daemon, use man:kill(1) on its process ID (standard
5395 Note that some Linux distributions could manage the LTTng session daemon
5396 as a service. In this case, you should use the service manager to
5397 start, restart, and stop session daemons.
5400 [[creating-destroying-tracing-sessions]]
5401 === Create and destroy a tracing session
5403 Almost all the LTTng control operations happen in the scope of
5404 a <<tracing-session,tracing session>>, which is the dialogue between the
5405 <<lttng-sessiond,session daemon>> and you.
5407 To create a tracing session with a generated name:
5409 * Use the man:lttng-create(1) command:
5418 The created tracing session's name is `auto` followed by the
5421 To create a tracing session with a specific name:
5423 * Use the optional argument of the man:lttng-create(1) command:
5428 $ lttng create my-session
5432 Replace `my-session` with the specific tracing session name.
5434 LTTng appends the creation date to the created tracing session's name.
5436 LTTng writes the traces of a tracing session in
5437 +$LTTNG_HOME/lttng-trace/__name__+ by default, where +__name__+ is the
5438 name of the tracing session. Note that the env:LTTNG_HOME environment
5439 variable defaults to `$HOME` if not set.
5441 To output LTTng traces to a non-default location:
5443 * Use the opt:lttng-create(1):--output option of the man:lttng-create(1) command:
5448 $ lttng create my-session --output=/tmp/some-directory
5452 You may create as many tracing sessions as you wish.
5454 To list all the existing tracing sessions for your Unix user:
5456 * Use the man:lttng-list(1) command:
5465 When you create a tracing session, it is set as the _current tracing
5466 session_. The following man:lttng(1) commands operate on the current
5467 tracing session when you don't specify one:
5469 [role="list-3-cols"]
5486 To change the current tracing session:
5488 * Use the man:lttng-set-session(1) command:
5493 $ lttng set-session new-session
5497 Replace `new-session` by the name of the new current tracing session.
5499 When you are done tracing in a given tracing session, you can destroy
5500 it. This operation frees the resources taken by the tracing session
5501 to destroy; it does not destroy the trace data that LTTng wrote for
5502 this tracing session.
5504 To destroy the current tracing session:
5506 * Use the man:lttng-destroy(1) command:
5516 [[list-instrumentation-points]]
5517 === List the available instrumentation points
5519 The <<lttng-sessiond,session daemon>> can query the running instrumented
5520 user applications and the Linux kernel to get a list of available
5521 instrumentation points. For the Linux kernel <<domain,tracing domain>>,
5522 they are tracepoints and system calls. For the user space tracing
5523 domain, they are tracepoints. For the other tracing domains, they are
5526 To list the available instrumentation points:
5528 * Use the man:lttng-list(1) command with the requested tracing domain's
5532 * opt:lttng-list(1):--kernel: Linux kernel tracepoints (your Unix user
5533 must be a root user, or it must be a member of the
5534 <<tracing-group,tracing group>>).
5535 * opt:lttng-list(1):--kernel with opt:lttng-list(1):--syscall: Linux
5536 kernel system calls (your Unix user must be a root user, or it must be
5537 a member of the tracing group).
5538 * opt:lttng-list(1):--userspace: user space tracepoints.
5539 * opt:lttng-list(1):--jul: `java.util.logging` loggers.
5540 * opt:lttng-list(1):--log4j: Apache log4j loggers.
5541 * opt:lttng-list(1):--python: Python loggers.
5544 .List the available user space tracepoints.
5548 $ lttng list --userspace
5552 .List the available Linux kernel system call tracepoints.
5556 $ lttng list --kernel --syscall
5561 [[enabling-disabling-events]]
5562 === Create and enable an event rule
5564 Once you <<creating-destroying-tracing-sessions,create a tracing
5565 session>>, you can create <<event,event rules>> with the
5566 man:lttng-enable-event(1) command.
5568 You specify each condition with a command-line option. The available
5569 condition options are shown in the following table.
5571 [role="growable",cols="asciidoc,asciidoc,default"]
5572 .Condition command-line options for the man:lttng-enable-event(1) command.
5574 |Option |Description |Applicable tracing domains
5580 . +--probe=__ADDR__+
5581 . +--function=__ADDR__+
5584 Instead of using the default _tracepoint_ instrumentation type, use:
5586 . A Linux system call.
5587 . A Linux https://lwn.net/Articles/132196/[KProbe] (symbol or address).
5588 . The entry and return points of a Linux function (symbol or address).
5592 |First positional argument.
5595 Tracepoint or system call name. In the case of a Linux KProbe or
5596 function, this is a custom name given to the event rule. With the
5597 JUL, log4j, and Python domains, this is a logger name.
5599 With a tracepoint, logger, or system call name, the last character
5600 can be `*` to match anything that remains.
5607 . +--loglevel=__LEVEL__+
5608 . +--loglevel-only=__LEVEL__+
5611 . Match only tracepoints or log statements with a logging level at
5612 least as severe as +__LEVEL__+.
5613 . Match only tracepoints or log statements with a logging level
5614 equal to +__LEVEL__+.
5616 See man:lttng-enable-event(1) for the list of available logging level
5619 |User space, JUL, log4j, and Python.
5621 |+--exclude=__EXCLUSIONS__+
5624 When you use a `*` character at the end of the tracepoint or logger
5625 name (first positional argument), exclude the specific names in the
5626 comma-delimited list +__EXCLUSIONS__+.
5629 User space, JUL, log4j, and Python.
5631 |+--filter=__EXPR__+
5634 Match only events which satisfy the expression +__EXPR__+.
5636 See man:lttng-enable-event(1) to learn more about the syntax of a
5643 You attach an event rule to a <<channel,channel>> on creation. If you do
5644 not specify the channel with the opt:lttng-enable-event(1):--channel
5645 option, and if the event rule to create is the first in its
5646 <<domain,tracing domain>> for a given tracing session, then LTTng
5647 creates a _default channel_ for you. This default channel is reused in
5648 subsequent invocations of the man:lttng-enable-event(1) command for the
5649 same tracing domain.
5651 An event rule is always enabled at creation time.
5653 The following examples show how you can combine the previous
5654 command-line options to create simple to more complex event rules.
5656 .Create an event rule targetting a Linux kernel tracepoint (default channel).
5660 $ lttng enable-event --kernel sched_switch
5664 .Create an event rule matching four Linux kernel system calls (default channel).
5668 $ lttng enable-event --kernel --syscall open,write,read,close
5672 .Create event rules matching tracepoints with filter expressions (default channel).
5676 $ lttng enable-event --kernel sched_switch --filter='prev_comm == "bash"'
5681 $ lttng enable-event --kernel --all \
5682 --filter='$ctx.tid == 1988 || $ctx.tid == 1534'
5687 $ lttng enable-event --jul my_logger \
5688 --filter='$app.retriever:cur_msg_id > 3'
5691 IMPORTANT: Make sure to always quote the filter string when you
5692 use man:lttng(1) from a shell.
5695 .Create an event rule matching any user space tracepoint of a given tracepoint provider with a log level range (default channel).
5699 $ lttng enable-event --userspace my_app:'*' --loglevel=TRACE_INFO
5702 IMPORTANT: Make sure to always quote the wildcard character when you
5703 use man:lttng(1) from a shell.
5706 .Create an event rule matching multiple Python loggers with a wildcard and with exclusions (default channel).
5710 $ lttng enable-event --python my-app.'*' \
5711 --exclude='my-app.module,my-app.hello'
5715 .Create an event rule matching any Apache log4j logger with a specific log level (default channel).
5719 $ lttng enable-event --log4j --all --loglevel-only=LOG4J_WARN
5723 .Create an event rule attached to a specific channel matching a specific user space tracepoint provider and tracepoint.
5727 $ lttng enable-event --userspace my_app:my_tracepoint --channel=my-channel
5731 The event rules of a given channel form a whitelist: as soon as an
5732 emitted event passes one of them, LTTng can record the event. For
5733 example, an event named `my_app:my_tracepoint` emitted from a user space
5734 tracepoint with a `TRACE_ERROR` log level passes both of the following
5739 $ lttng enable-event --userspace my_app:my_tracepoint
5740 $ lttng enable-event --userspace my_app:my_tracepoint \
5741 --loglevel=TRACE_INFO
5744 The second event rule is redundant: the first one includes
5748 [[disable-event-rule]]
5749 === Disable an event rule
5751 To disable an event rule that you <<enabling-disabling-events,created>>
5752 previously, use the man:lttng-disable-event(1) command. This command
5753 disables _all_ the event rules (of a given tracing domain and channel)
5754 which match an instrumentation point. The other conditions are not
5755 supported as of LTTng{nbsp}{revision}.
5757 The LTTng tracer does not record an emitted event which passes
5758 a _disabled_ event rule.
5760 .Disable an event rule matching a Python logger (default channel).
5764 $ lttng disable-event --python my-logger
5768 .Disable an event rule matching all `java.util.logging` loggers (default channel).
5772 $ lttng disable-event --jul '*'
5776 .Disable _all_ the event rules of the default channel.
5778 The opt:lttng-disable-event(1):--all-events option is not, like the
5779 opt:lttng-enable-event(1):--all option of man:lttng-enable-event(1), the
5780 equivalent of the event name `*` (wildcard): it disables _all_ the event
5781 rules of a given channel.
5785 $ lttng disable-event --jul --all-events
5789 NOTE: You cannot delete an event rule once you create it.
5793 === Get the status of a tracing session
5795 To get the status of the current tracing session, that is, its
5796 parameters, its channels, event rules, and their attributes:
5798 * Use the man:lttng-status(1) command:
5808 To get the status of any tracing session:
5810 * Use the man:lttng-list(1) command with the tracing session's name:
5815 $ lttng list my-session
5819 Replace `my-session` with the desired tracing session's name.
5822 [[basic-tracing-session-control]]
5823 === Start and stop a tracing session
5825 Once you <<creating-destroying-tracing-sessions,create a tracing
5827 <<enabling-disabling-events,create one or more event rules>>,
5828 you can start and stop the tracers for this tracing session.
5830 To start tracing in the current tracing session:
5832 * Use the man:lttng-start(1) command:
5841 LTTng is very flexible: you can launch user applications before
5842 or after the you start the tracers. The tracers only record the events
5843 if they pass enabled event rules and if they occur while the tracers are
5846 To stop tracing in the current tracing session:
5848 * Use the man:lttng-stop(1) command:
5857 If there were <<channel-overwrite-mode-vs-discard-mode,lost event
5858 records>> or lost sub-buffers since the last time you ran
5859 man:lttng-start(1), warnings are printed when you run the
5860 man:lttng-stop(1) command.
5863 [[enabling-disabling-channels]]
5864 === Create a channel
5866 Once you create a tracing session, you can create a <<channel,channel>>
5867 with the man:lttng-enable-channel(1) command.
5869 Note that LTTng automatically creates a default channel when, for a
5870 given <<domain,tracing domain>>, no channels exist and you
5871 <<enabling-disabling-events,create>> the first event rule. This default
5872 channel is named `channel0` and its attributes are set to reasonable
5873 values. Therefore, you only need to create a channel when you need
5874 non-default attributes.
5876 You specify each non-default channel attribute with a command-line
5877 option when you use the man:lttng-enable-channel(1) command. The
5878 available command-line options are:
5880 [role="growable",cols="asciidoc,asciidoc"]
5881 .Command-line options for the man:lttng-enable-channel(1) command.
5883 |Option |Description
5889 <<channel-overwrite-mode-vs-discard-mode,event loss mode>> instead of
5890 the default _discard_ mode.
5892 |`--buffers-pid` (user space tracing domain only)
5895 Use the per-process <<channel-buffering-schemes,buffering scheme>>
5896 instead of the default per-user buffering scheme.
5898 |+--subbuf-size=__SIZE__+
5901 Allocate sub-buffers of +__SIZE__+ bytes (power of two), for each CPU,
5902 either for each Unix user (default), or for each instrumented process.
5904 See <<channel-subbuf-size-vs-subbuf-count,Sub-buffer count and size>>.
5906 |+--num-subbuf=__COUNT__+
5909 Allocate +__COUNT__+ sub-buffers (power of two), for each CPU, either
5910 for each Unix user (default), or for each instrumented process.
5912 See <<channel-subbuf-size-vs-subbuf-count,Sub-buffer count and size>>.
5914 |+--tracefile-size=__SIZE__+
5917 Set the maximum size of each trace file that this channel writes within
5918 a stream to +__SIZE__+ bytes instead of no maximum.
5920 See <<tracefile-rotation,Trace file count and size>>.
5922 |+--tracefile-count=__COUNT__+
5925 Limit the number of trace files that this channel creates to
5926 +__COUNT__+ channels instead of no limit.
5928 See <<tracefile-rotation,Trace file count and size>>.
5930 |+--switch-timer=__PERIODUS__+
5933 Set the <<channel-switch-timer,switch timer period>>
5934 to +__PERIODUS__+{nbsp}µs.
5936 |+--read-timer=__PERIODUS__+
5939 Set the <<channel-read-timer,read timer period>>
5940 to +__PERIODUS__+{nbsp}µs.
5942 |[[opt-blocking-timeout]]+--blocking-timeout=__TIMEOUTUS__+
5945 Set the timeout of user space applications which load LTTng-UST
5946 in blocking mode to +__TIMEOUTUS__+:
5949 Never block (non-blocking mode).
5952 Block forever until space is available in a sub-buffer to record
5955 __n__, a positive value::
5956 Wait for at most __n__ µs when trying to write into a sub-buffer.
5958 Note that, for this option to have any effect on an instrumented
5959 user space application, you need to run the application with a set
5960 env:LTTNG_UST_ALLOW_BLOCKING environment variable.
5962 |+--output=__TYPE__+ (Linux kernel tracing domain only)
5965 Set the channel's output type to +__TYPE__+, either `mmap` or `splice`.
5969 You can only create a channel in the Linux kernel and user space
5970 <<domain,tracing domains>>: other tracing domains have their own channel
5971 created on the fly when <<enabling-disabling-events,creating event
5976 Because of a current LTTng limitation, you must create all channels
5977 _before_ you <<basic-tracing-session-control,start tracing>> in a given
5978 tracing session, that is, before the first time you run
5981 Since LTTng automatically creates a default channel when you use the
5982 man:lttng-enable-event(1) command with a specific tracing domain, you
5983 cannot, for example, create a Linux kernel event rule, start tracing,
5984 and then create a user space event rule, because no user space channel
5985 exists yet and it's too late to create one.
5987 For this reason, make sure to configure your channels properly
5988 before starting the tracers for the first time!
5991 The following examples show how you can combine the previous
5992 command-line options to create simple to more complex channels.
5994 .Create a Linux kernel channel with default attributes.
5998 $ lttng enable-channel --kernel my-channel
6002 .Create a user space channel with 4 sub-buffers or 1{nbsp}MiB each, per CPU, per instrumented process.
6006 $ lttng enable-channel --userspace --num-subbuf=4 --subbuf-size=1M \
6007 --buffers-pid my-channel
6011 .[[blocking-timeout-example]]Create a default user space channel with an infinite blocking timeout:
6013 <<creating-destroying-tracing-sessions,Create a tracing-session>>,
6014 create the channel, <<enabling-disabling-events,create an event rule>>,
6015 and <<basic-tracing-session-control,start tracing>>:
6020 $ lttng enable-channel --userspace --blocking-timeout=-1 blocking-channel
6021 $ lttng enable-event --userspace --channel=blocking-channel --all
6025 Run an application instrumented with LTTng-UST and allow it to block:
6029 $ LTTNG_UST_ALLOW_BLOCKING=1 my-app
6033 .Create a Linux kernel channel which rotates 8 trace files of 4{nbsp}MiB each for each stream
6037 $ lttng enable-channel --kernel --tracefile-count=8 \
6038 --tracefile-size=4194304 my-channel
6042 .Create a user space channel in overwrite (or _flight recorder_) mode.
6046 $ lttng enable-channel --userspace --overwrite my-channel
6050 You can <<enabling-disabling-events,create>> the same event rule in
6051 two different channels:
6055 $ lttng enable-event --userspace --channel=my-channel app:tp
6056 $ lttng enable-event --userspace --channel=other-channel app:tp
6059 If both channels are enabled, when a tracepoint named `app:tp` is
6060 reached, LTTng records two events, one for each channel.
6064 === Disable a channel
6066 To disable a specific channel that you <<enabling-disabling-channels,created>>
6067 previously, use the man:lttng-disable-channel(1) command.
6069 .Disable a specific Linux kernel channel.
6073 $ lttng disable-channel --kernel my-channel
6077 The state of a channel precedes the individual states of event rules
6078 attached to it: event rules which belong to a disabled channel, even if
6079 they are enabled, are also considered disabled.
6083 === Add context fields to a channel
6085 Event record fields in trace files provide important information about
6086 events that occured previously, but sometimes some external context may
6087 help you solve a problem faster. Examples of context fields are:
6089 * The **process ID**, **thread ID**, **process name**, and
6090 **process priority** of the thread in which the event occurs.
6091 * The **hostname** of the system on which the event occurs.
6092 * The current values of many possible **performance counters** using
6094 ** CPU cycles, stalled cycles, idle cycles, and the other cycle types.
6096 ** Branch instructions, misses, and loads.
6098 * Any context defined at the application level (supported for the
6099 JUL and log4j <<domain,tracing domains>>).
6101 To get the full list of available context fields, see
6102 `lttng add-context --list`. Some context fields are reserved for a
6103 specific <<domain,tracing domain>> (Linux kernel or user space).
6105 You add context fields to <<channel,channels>>. All the events
6106 that a channel with added context fields records contain those fields.
6108 To add context fields to one or all the channels of a given tracing
6111 * Use the man:lttng-add-context(1) command.
6113 .Add context fields to all the channels of the current tracing session.
6115 The following command line adds the virtual process identifier and
6116 the per-thread CPU cycles count fields to all the user space channels
6117 of the current tracing session.
6121 $ lttng add-context --userspace --type=vpid --type=perf:thread:cpu-cycles
6125 .Add performance counter context fields by raw ID
6127 See man:lttng-add-context(1) for the exact format of the context field
6128 type, which is partly compatible with the format used in
6133 $ lttng add-context --userspace --type=perf:thread:raw:r0110:test
6134 $ lttng add-context --kernel --type=perf:cpu:raw:r0013c:x86unhalted
6138 .Add a context field to a specific channel.
6140 The following command line adds the thread identifier context field
6141 to the Linux kernel channel named `my-channel` in the current
6146 $ lttng add-context --kernel --channel=my-channel --type=tid
6150 .Add an application-specific context field to a specific channel.
6152 The following command line adds the `cur_msg_id` context field of the
6153 `retriever` context retriever for all the instrumented
6154 <<java-application,Java applications>> recording <<event,event records>>
6155 in the channel named `my-channel`:
6159 $ lttng add-context --kernel --channel=my-channel \
6160 --type='$app:retriever:cur_msg_id'
6163 IMPORTANT: Make sure to always quote the `$` character when you
6164 use man:lttng-add-context(1) from a shell.
6167 NOTE: You cannot remove context fields from a channel once you add it.
6172 === Track process IDs
6174 It's often useful to allow only specific process IDs (PIDs) to emit
6175 events. For example, you may wish to record all the system calls made by
6176 a given process (Ă la http://linux.die.net/man/1/strace[strace]).
6178 The man:lttng-track(1) and man:lttng-untrack(1) commands serve this
6179 purpose. Both commands operate on a whitelist of process IDs. You _add_
6180 entries to this whitelist with the man:lttng-track(1) command and remove
6181 entries with the man:lttng-untrack(1) command. Any process which has one
6182 of the PIDs in the whitelist is allowed to emit LTTng events which pass
6183 an enabled <<event,event rule>>.
6185 NOTE: The PID tracker tracks the _numeric process IDs_. Should a
6186 process with a given tracked ID exit and another process be given this
6187 ID, then the latter would also be allowed to emit events.
6189 .Track and untrack process IDs.
6191 For the sake of the following example, assume the target system has 16
6195 <<creating-destroying-tracing-sessions,create a tracing session>>,
6196 the whitelist contains all the possible PIDs:
6199 .All PIDs are tracked.
6200 image::track-all.png[]
6202 When the whitelist is full and you use the man:lttng-track(1) command to
6203 specify some PIDs to track, LTTng first clears the whitelist, then it
6204 tracks the specific PIDs. After:
6208 $ lttng track --pid=3,4,7,10,13
6214 .PIDs 3, 4, 7, 10, and 13 are tracked.
6215 image::track-3-4-7-10-13.png[]
6217 You can add more PIDs to the whitelist afterwards:
6221 $ lttng track --pid=1,15,16
6227 .PIDs 1, 15, and 16 are added to the whitelist.
6228 image::track-1-3-4-7-10-13-15-16.png[]
6230 The man:lttng-untrack(1) command removes entries from the PID tracker's
6231 whitelist. Given the previous example, the following command:
6235 $ lttng untrack --pid=3,7,10,13
6238 leads to this whitelist:
6241 .PIDs 3, 7, 10, and 13 are removed from the whitelist.
6242 image::track-1-4-15-16.png[]
6244 LTTng can track all possible PIDs again using the opt:track(1):--all
6249 $ lttng track --pid --all
6252 The result is, again:
6255 .All PIDs are tracked.
6256 image::track-all.png[]
6259 .Track only specific PIDs
6261 A very typical use case with PID tracking is to start with an empty
6262 whitelist, then <<basic-tracing-session-control,start the tracers>>, and
6263 then add PIDs manually while tracers are active. You can accomplish this
6264 by using the opt:lttng-untrack(1):--all option of the
6265 man:lttng-untrack(1) command to clear the whitelist after you
6266 <<creating-destroying-tracing-sessions,create a tracing session>>:
6270 $ lttng untrack --pid --all
6276 .No PIDs are tracked.
6277 image::untrack-all.png[]
6279 If you trace with this whitelist configuration, the tracer records no
6280 events for this <<domain,tracing domain>> because no processes are
6281 tracked. You can use the man:lttng-track(1) command as usual to track
6282 specific PIDs, for example:
6286 $ lttng track --pid=6,11
6292 .PIDs 6 and 11 are tracked.
6293 image::track-6-11.png[]
6298 [[saving-loading-tracing-session]]
6299 === Save and load tracing session configurations
6301 Configuring a <<tracing-session,tracing session>> can be long. Some of
6302 the tasks involved are:
6304 * <<enabling-disabling-channels,Create channels>> with
6305 specific attributes.
6306 * <<adding-context,Add context fields>> to specific channels.
6307 * <<enabling-disabling-events,Create event rules>> with specific log
6308 level and filter conditions.
6310 If you use LTTng to solve real world problems, chances are you have to
6311 record events using the same tracing session setup over and over,
6312 modifying a few variables each time in your instrumented program
6313 or environment. To avoid constant tracing session reconfiguration,
6314 the man:lttng(1) command-line tool can save and load tracing session
6315 configurations to/from XML files.
6317 To save a given tracing session configuration:
6319 * Use the man:lttng-save(1) command:
6324 $ lttng save my-session
6328 Replace `my-session` with the name of the tracing session to save.
6330 LTTng saves tracing session configurations to
6331 dir:{$LTTNG_HOME/.lttng/sessions} by default. Note that the
6332 env:LTTNG_HOME environment variable defaults to `$HOME` if not set. Use
6333 the opt:lttng-save(1):--output-path option to change this destination
6336 LTTng saves all configuration parameters, for example:
6338 * The tracing session name.
6339 * The trace data output path.
6340 * The channels with their state and all their attributes.
6341 * The context fields you added to channels.
6342 * The event rules with their state, log level and filter conditions.
6344 To load a tracing session:
6346 * Use the man:lttng-load(1) command:
6351 $ lttng load my-session
6355 Replace `my-session` with the name of the tracing session to load.
6357 When LTTng loads a configuration, it restores your saved tracing session
6358 as if you just configured it manually.
6360 See man:lttng(1) for the complete list of command-line options. You
6361 can also save and load all many sessions at a time, and decide in which
6362 directory to output the XML files.
6365 [[sending-trace-data-over-the-network]]
6366 === Send trace data over the network
6368 LTTng can send the recorded trace data to a remote system over the
6369 network instead of writing it to the local file system.
6371 To send the trace data over the network:
6373 . On the _remote_ system (which can also be the target system),
6374 start an LTTng <<lttng-relayd,relay daemon>> (man:lttng-relayd(8)):
6383 . On the _target_ system, create a tracing session configured to
6384 send trace data over the network:
6389 $ lttng create my-session --set-url=net://remote-system
6393 Replace `remote-system` by the host name or IP address of the
6394 remote system. See man:lttng-create(1) for the exact URL format.
6396 . On the target system, use the man:lttng(1) command-line tool as usual.
6397 When tracing is active, the target's consumer daemon sends sub-buffers
6398 to the relay daemon running on the remote system instead of flushing
6399 them to the local file system. The relay daemon writes the received
6400 packets to the local file system.
6402 The relay daemon writes trace files to
6403 +$LTTNG_HOME/lttng-traces/__hostname__/__session__+ by default, where
6404 +__hostname__+ is the host name of the target system and +__session__+
6405 is the tracing session name. Note that the env:LTTNG_HOME environment
6406 variable defaults to `$HOME` if not set. Use the
6407 opt:lttng-relayd(8):--output option of man:lttng-relayd(8) to write
6408 trace files to another base directory.
6413 === View events as LTTng emits them (noch:{LTTng} live)
6415 LTTng live is a network protocol implemented by the <<lttng-relayd,relay
6416 daemon>> (man:lttng-relayd(8)) to allow compatible trace viewers to
6417 display events as LTTng emits them on the target system while tracing is
6420 The relay daemon creates a _tee_: it forwards the trace data to both
6421 the local file system and to connected live viewers:
6424 .The relay daemon creates a _tee_, forwarding the trace data to both trace files and a connected live viewer.
6429 . On the _target system_, create a <<tracing-session,tracing session>>
6435 $ lttng create my-session --live
6439 This spawns a local relay daemon.
6441 . Start the live viewer and configure it to connect to the relay
6442 daemon. For example, with http://diamon.org/babeltrace[Babeltrace]:
6447 $ babeltrace --input-format=lttng-live \
6448 net://localhost/host/hostname/my-session
6455 * `hostname` with the host name of the target system.
6456 * `my-session` with the name of the tracing session to view.
6459 . Configure the tracing session as usual with the man:lttng(1)
6460 command-line tool, and <<basic-tracing-session-control,start tracing>>.
6462 You can list the available live tracing sessions with Babeltrace:
6466 $ babeltrace --input-format=lttng-live net://localhost
6469 You can start the relay daemon on another system. In this case, you need
6470 to specify the relay daemon's URL when you create the tracing session
6471 with the opt:lttng-create(1):--set-url option. You also need to replace
6472 `localhost` in the procedure above with the host name of the system on
6473 which the relay daemon is running.
6475 See man:lttng-create(1) and man:lttng-relayd(8) for the complete list of
6476 command-line options.
6480 [[taking-a-snapshot]]
6481 === Take a snapshot of the current sub-buffers of a tracing session
6483 The normal behavior of LTTng is to append full sub-buffers to growing
6484 trace data files. This is ideal to keep a full history of the events
6485 that occurred on the target system, but it can
6486 represent too much data in some situations. For example, you may wish
6487 to trace your application continuously until some critical situation
6488 happens, in which case you only need the latest few recorded
6489 events to perform the desired analysis, not multi-gigabyte trace files.
6491 With the man:lttng-snapshot(1) command, you can take a snapshot of the
6492 current sub-buffers of a given <<tracing-session,tracing session>>.
6493 LTTng can write the snapshot to the local file system or send it over
6498 . Create a tracing session in _snapshot mode_:
6503 $ lttng create my-session --snapshot
6507 The <<channel-overwrite-mode-vs-discard-mode,event loss mode>> of
6508 <<channel,channels>> created in this mode is automatically set to
6509 _overwrite_ (flight recorder mode).
6511 . Configure the tracing session as usual with the man:lttng(1)
6512 command-line tool, and <<basic-tracing-session-control,start tracing>>.
6514 . **Optional**: When you need to take a snapshot,
6515 <<basic-tracing-session-control,stop tracing>>.
6517 You can take a snapshot when the tracers are active, but if you stop
6518 them first, you are sure that the data in the sub-buffers does not
6519 change before you actually take the snapshot.
6526 $ lttng snapshot record --name=my-first-snapshot
6530 LTTng writes the current sub-buffers of all the current tracing
6531 session's channels to trace files on the local file system. Those trace
6532 files have `my-first-snapshot` in their name.
6534 There is no difference between the format of a normal trace file and the
6535 format of a snapshot: viewers of LTTng traces also support LTTng
6538 By default, LTTng writes snapshot files to the path shown by
6539 `lttng snapshot list-output`. You can change this path or decide to send
6540 snapshots over the network using either:
6542 . An output path or URL that you specify when you create the
6544 . An snapshot output path or URL that you add using
6545 `lttng snapshot add-output`
6546 . An output path or URL that you provide directly to the
6547 `lttng snapshot record` command.
6549 Method 3 overrides method 2, which overrides method 1. When you
6550 specify a URL, a relay daemon must listen on a remote system (see
6551 <<sending-trace-data-over-the-network,Send trace data over the network>>).
6556 === Use the machine interface
6558 With any command of the man:lttng(1) command-line tool, you can set the
6559 opt:lttng(1):--mi option to `xml` (before the command name) to get an
6560 XML machine interface output, for example:
6564 $ lttng --mi=xml enable-event --kernel --syscall open
6567 A schema definition (XSD) is
6568 https://github.com/lttng/lttng-tools/blob/stable-2.10/src/common/mi-lttng-3.0.xsd[available]
6569 to ease the integration with external tools as much as possible.
6573 [[metadata-regenerate]]
6574 === Regenerate the metadata of an LTTng trace
6576 An LTTng trace, which is a http://diamon.org/ctf[CTF] trace, has both
6577 data stream files and a metadata file. This metadata file contains,
6578 amongst other things, information about the offset of the clock sources
6579 used to timestamp <<event,event records>> when tracing.
6581 If, once a <<tracing-session,tracing session>> is
6582 <<basic-tracing-session-control,started>>, a major
6583 https://en.wikipedia.org/wiki/Network_Time_Protocol[NTP] correction
6584 happens, the trace's clock offset also needs to be updated. You
6585 can use the `metadata` item of the man:lttng-regenerate(1) command
6588 The main use case of this command is to allow a system to boot with
6589 an incorrect wall time and trace it with LTTng before its wall time
6590 is corrected. Once the system is known to be in a state where its
6591 wall time is correct, it can run `lttng regenerate metadata`.
6593 To regenerate the metadata of an LTTng trace:
6595 * Use the `metadata` item of the man:lttng-regenerate(1) command:
6600 $ lttng regenerate metadata
6606 `lttng regenerate metadata` has the following limitations:
6608 * Tracing session <<creating-destroying-tracing-sessions,created>>
6610 * User space <<channel,channels>>, if any, are using
6611 <<channel-buffering-schemes,per-user buffering>>.
6616 [[regenerate-statedump]]
6617 === Regenerate the state dump of a tracing session
6619 The LTTng kernel and user space tracers generate state dump
6620 <<event,event records>> when the application starts or when you
6621 <<basic-tracing-session-control,start a tracing session>>. An analysis
6622 can use the state dump event records to set an initial state before it
6623 builds the rest of the state from the following event records.
6624 http://tracecompass.org/[Trace Compass] is a notable example of an
6625 application which uses the state dump of an LTTng trace.
6627 When you <<taking-a-snapshot,take a snapshot>>, it's possible that the
6628 state dump event records are not included in the snapshot because they
6629 were recorded to a sub-buffer that has been consumed or overwritten
6632 You can use the `lttng regenerate statedump` command to emit the state
6633 dump event records again.
6635 To regenerate the state dump of the current tracing session, provided
6636 create it in snapshot mode, before you take a snapshot:
6638 . Use the `statedump` item of the man:lttng-regenerate(1) command:
6643 $ lttng regenerate statedump
6647 . <<basic-tracing-session-control,Stop the tracing session>>:
6656 . <<taking-a-snapshot,Take a snapshot>>:
6661 $ lttng snapshot record --name=my-snapshot
6665 Depending on the event throughput, you should run steps 1 and 2
6666 as closely as possible.
6668 NOTE: To record the state dump events, you need to
6669 <<enabling-disabling-events,create event rules>> which enable them.
6670 LTTng-UST state dump tracepoints start with `lttng_ust_statedump:`.
6671 LTTng-modules state dump tracepoints start with `lttng_statedump_`.
6675 [[persistent-memory-file-systems]]
6676 === Record trace data on persistent memory file systems
6678 https://en.wikipedia.org/wiki/Non-volatile_random-access_memory[Non-volatile random-access memory]
6679 (NVRAM) is random-access memory that retains its information when power
6680 is turned off (non-volatile). Systems with such memory can store data
6681 structures in RAM and retrieve them after a reboot, without flushing
6682 to typical _storage_.
6684 Linux supports NVRAM file systems thanks to either
6685 http://pramfs.sourceforge.net/[PRAMFS] or
6686 https://www.kernel.org/doc/Documentation/filesystems/dax.txt[DAX]{nbsp}+{nbsp}http://lkml.iu.edu/hypermail/linux/kernel/1504.1/03463.html[pmem]
6687 (requires Linux 4.1+).
6689 This section does not describe how to operate such file systems;
6690 we assume that you have a working persistent memory file system.
6692 When you create a <<tracing-session,tracing session>>, you can specify
6693 the path of the shared memory holding the sub-buffers. If you specify a
6694 location on an NVRAM file system, then you can retrieve the latest
6695 recorded trace data when the system reboots after a crash.
6697 To record trace data on a persistent memory file system and retrieve the
6698 trace data after a system crash:
6700 . Create a tracing session with a sub-buffer shared memory path located
6701 on an NVRAM file system:
6706 $ lttng create my-session --shm-path=/path/to/shm
6710 . Configure the tracing session as usual with the man:lttng(1)
6711 command-line tool, and <<basic-tracing-session-control,start tracing>>.
6713 . After a system crash, use the man:lttng-crash(1) command-line tool to
6714 view the trace data recorded on the NVRAM file system:
6719 $ lttng-crash /path/to/shm
6723 The binary layout of the ring buffer files is not exactly the same as
6724 the trace files layout. This is why you need to use man:lttng-crash(1)
6725 instead of your preferred trace viewer directly.
6727 To convert the ring buffer files to LTTng trace files:
6729 * Use the opt:lttng-crash(1):--extract option of man:lttng-crash(1):
6734 $ lttng-crash --extract=/path/to/trace /path/to/shm
6742 [[lttng-modules-ref]]
6743 === noch:{LTTng-modules}
6747 [[lttng-tracepoint-enum]]
6748 ==== `LTTNG_TRACEPOINT_ENUM()` usage
6750 Use the `LTTNG_TRACEPOINT_ENUM()` macro to define an enumeration:
6754 LTTNG_TRACEPOINT_ENUM(name, TP_ENUM_VALUES(entries))
6759 * `name` with the name of the enumeration (C identifier, unique
6760 amongst all the defined enumerations).
6761 * `entries` with a list of enumeration entries.
6763 The available enumeration entry macros are:
6765 +ctf_enum_value(__name__, __value__)+::
6766 Entry named +__name__+ mapped to the integral value +__value__+.
6768 +ctf_enum_range(__name__, __begin__, __end__)+::
6769 Entry named +__name__+ mapped to the range of integral values between
6770 +__begin__+ (included) and +__end__+ (included).
6772 +ctf_enum_auto(__name__)+::
6773 Entry named +__name__+ mapped to the integral value following the
6774 last mapping's value.
6776 The last value of a `ctf_enum_value()` entry is its +__value__+
6779 The last value of a `ctf_enum_range()` entry is its +__end__+ parameter.
6781 If `ctf_enum_auto()` is the first entry in the list, its integral
6784 Use the `ctf_enum()` <<lttng-modules-tp-fields,field definition macro>>
6785 to use a defined enumeration as a tracepoint field.
6787 .Define an enumeration with `LTTNG_TRACEPOINT_ENUM()`.
6791 LTTNG_TRACEPOINT_ENUM(
6794 ctf_enum_auto("AUTO: EXPECT 0")
6795 ctf_enum_value("VALUE: 23", 23)
6796 ctf_enum_value("VALUE: 27", 27)
6797 ctf_enum_auto("AUTO: EXPECT 28")
6798 ctf_enum_range("RANGE: 101 TO 303", 101, 303)
6799 ctf_enum_auto("AUTO: EXPECT 304")
6807 [[lttng-modules-tp-fields]]
6808 ==== Tracepoint fields macros (for `TP_FIELDS()`)
6810 [[tp-fast-assign]][[tp-struct-entry]]The available macros to define
6811 tracepoint fields, which must be listed within `TP_FIELDS()` in
6812 `LTTNG_TRACEPOINT_EVENT()`, are:
6814 [role="func-desc growable",cols="asciidoc,asciidoc"]
6815 .Available macros to define LTTng-modules tracepoint fields
6817 |Macro |Description and parameters
6820 +ctf_integer(__t__, __n__, __e__)+
6822 +ctf_integer_nowrite(__t__, __n__, __e__)+
6824 +ctf_user_integer(__t__, __n__, __e__)+
6826 +ctf_user_integer_nowrite(__t__, __n__, __e__)+
6828 Standard integer, displayed in base 10.
6831 Integer C type (`int`, `long`, `size_t`, ...).
6837 Argument expression.
6840 +ctf_integer_hex(__t__, __n__, __e__)+
6842 +ctf_user_integer_hex(__t__, __n__, __e__)+
6844 Standard integer, displayed in base 16.
6853 Argument expression.
6855 |+ctf_integer_oct(__t__, __n__, __e__)+
6857 Standard integer, displayed in base 8.
6866 Argument expression.
6869 +ctf_integer_network(__t__, __n__, __e__)+
6871 +ctf_user_integer_network(__t__, __n__, __e__)+
6873 Integer in network byte order (big-endian), displayed in base 10.
6882 Argument expression.
6885 +ctf_integer_network_hex(__t__, __n__, __e__)+
6887 +ctf_user_integer_network_hex(__t__, __n__, __e__)+
6889 Integer in network byte order, displayed in base 16.
6898 Argument expression.
6901 +ctf_enum(__N__, __t__, __n__, __e__)+
6903 +ctf_enum_nowrite(__N__, __t__, __n__, __e__)+
6905 +ctf_user_enum(__N__, __t__, __n__, __e__)+
6907 +ctf_user_enum_nowrite(__N__, __t__, __n__, __e__)+
6912 Name of a <<lttng-tracepoint-enum,previously defined enumeration>>.
6915 Integer C type (`int`, `long`, `size_t`, ...).
6921 Argument expression.
6924 +ctf_string(__n__, __e__)+
6926 +ctf_string_nowrite(__n__, __e__)+
6928 +ctf_user_string(__n__, __e__)+
6930 +ctf_user_string_nowrite(__n__, __e__)+
6932 Null-terminated string; undefined behavior if +__e__+ is `NULL`.
6938 Argument expression.
6941 +ctf_array(__t__, __n__, __e__, __s__)+
6943 +ctf_array_nowrite(__t__, __n__, __e__, __s__)+
6945 +ctf_user_array(__t__, __n__, __e__, __s__)+
6947 +ctf_user_array_nowrite(__t__, __n__, __e__, __s__)+
6949 Statically-sized array of integers.
6952 Array element C type.
6958 Argument expression.
6964 +ctf_array_bitfield(__t__, __n__, __e__, __s__)+
6966 +ctf_array_bitfield_nowrite(__t__, __n__, __e__, __s__)+
6968 +ctf_user_array_bitfield(__t__, __n__, __e__, __s__)+
6970 +ctf_user_array_bitfield_nowrite(__t__, __n__, __e__, __s__)+
6972 Statically-sized array of bits.
6974 The type of +__e__+ must be an integer type. +__s__+ is the number
6975 of elements of such type in +__e__+, not the number of bits.
6978 Array element C type.
6984 Argument expression.
6990 +ctf_array_text(__t__, __n__, __e__, __s__)+
6992 +ctf_array_text_nowrite(__t__, __n__, __e__, __s__)+
6994 +ctf_user_array_text(__t__, __n__, __e__, __s__)+
6996 +ctf_user_array_text_nowrite(__t__, __n__, __e__, __s__)+
6998 Statically-sized array, printed as text.
7000 The string does not need to be null-terminated.
7003 Array element C type (always `char`).
7009 Argument expression.
7015 +ctf_sequence(__t__, __n__, __e__, __T__, __E__)+
7017 +ctf_sequence_nowrite(__t__, __n__, __e__, __T__, __E__)+
7019 +ctf_user_sequence(__t__, __n__, __e__, __T__, __E__)+
7021 +ctf_user_sequence_nowrite(__t__, __n__, __e__, __T__, __E__)+
7023 Dynamically-sized array of integers.
7025 The type of +__E__+ must be unsigned.
7028 Array element C type.
7034 Argument expression.
7037 Length expression C type.
7043 +ctf_sequence_hex(__t__, __n__, __e__, __T__, __E__)+
7045 +ctf_user_sequence_hex(__t__, __n__, __e__, __T__, __E__)+
7047 Dynamically-sized array of integers, displayed in base 16.
7049 The type of +__E__+ must be unsigned.
7052 Array element C type.
7058 Argument expression.
7061 Length expression C type.
7066 |+ctf_sequence_network(__t__, __n__, __e__, __T__, __E__)+
7068 Dynamically-sized array of integers in network byte order (big-endian),
7069 displayed in base 10.
7071 The type of +__E__+ must be unsigned.
7074 Array element C type.
7080 Argument expression.
7083 Length expression C type.
7089 +ctf_sequence_bitfield(__t__, __n__, __e__, __T__, __E__)+
7091 +ctf_sequence_bitfield_nowrite(__t__, __n__, __e__, __T__, __E__)+
7093 +ctf_user_sequence_bitfield(__t__, __n__, __e__, __T__, __E__)+
7095 +ctf_user_sequence_bitfield_nowrite(__t__, __n__, __e__, __T__, __E__)+
7097 Dynamically-sized array of bits.
7099 The type of +__e__+ must be an integer type. +__s__+ is the number
7100 of elements of such type in +__e__+, not the number of bits.
7102 The type of +__E__+ must be unsigned.
7105 Array element C type.
7111 Argument expression.
7114 Length expression C type.
7120 +ctf_sequence_text(__t__, __n__, __e__, __T__, __E__)+
7122 +ctf_sequence_text_nowrite(__t__, __n__, __e__, __T__, __E__)+
7124 +ctf_user_sequence_text(__t__, __n__, __e__, __T__, __E__)+
7126 +ctf_user_sequence_text_nowrite(__t__, __n__, __e__, __T__, __E__)+
7128 Dynamically-sized array, displayed as text.
7130 The string does not need to be null-terminated.
7132 The type of +__E__+ must be unsigned.
7134 The behaviour is undefined if +__e__+ is `NULL`.
7137 Sequence element C type (always `char`).
7143 Argument expression.
7146 Length expression C type.
7152 Use the `_user` versions when the argument expression, `e`, is
7153 a user space address. In the cases of `ctf_user_integer*()` and
7154 `ctf_user_float*()`, `&e` must be a user space address, thus `e` must
7157 The `_nowrite` versions omit themselves from the session trace, but are
7158 otherwise identical. This means the `_nowrite` fields won't be written
7159 in the recorded trace. Their primary purpose is to make some
7160 of the event context available to the
7161 <<enabling-disabling-events,event filters>> without having to
7162 commit the data to sub-buffers.
7168 Terms related to LTTng and to tracing in general:
7171 The http://diamon.org/babeltrace[Babeltrace] project, which includes
7172 the cmd:babeltrace command, some libraries, and Python bindings.
7174 <<channel-buffering-schemes,buffering scheme>>::
7175 A layout of sub-buffers applied to a given channel.
7177 <<channel,channel>>::
7178 An entity which is responsible for a set of ring buffers.
7180 <<event,Event rules>> are always attached to a specific channel.
7183 A reference of time for a tracer.
7185 <<lttng-consumerd,consumer daemon>>::
7186 A process which is responsible for consuming the full sub-buffers
7187 and write them to a file system or send them over the network.
7189 <<channel-overwrite-mode-vs-discard-mode,discard mode>>:: The event loss
7190 mode in which the tracer _discards_ new event records when there's no
7191 sub-buffer space left to store them.
7194 The consequence of the execution of an instrumentation
7195 point, like a tracepoint that you manually place in some source code,
7196 or a Linux kernel KProbe.
7198 An event is said to _occur_ at a specific time. Different actions can
7199 be taken upon the occurrence of an event, like record the event's payload
7202 <<channel-overwrite-mode-vs-discard-mode,event loss mode>>::
7203 The mechanism by which event records of a given channel are lost
7204 (not recorded) when there is no sub-buffer space left to store them.
7206 [[def-event-name]]event name::
7207 The name of an event, which is also the name of the event record.
7208 This is also called the _instrumentation point name_.
7211 A record, in a trace, of the payload of an event which occured.
7213 <<event,event rule>>::
7214 Set of conditions which must be satisfied for one or more occuring
7215 events to be recorded.
7217 `java.util.logging`::
7219 https://docs.oracle.com/javase/7/docs/api/java/util/logging/package-summary.html[core logging facilities].
7221 <<instrumenting,instrumentation>>::
7222 The use of LTTng probes to make a piece of software traceable.
7224 instrumentation point::
7225 A point in the execution path of a piece of software that, when
7226 reached by this execution, can emit an event.
7228 instrumentation point name::
7229 See _<<def-event-name,event name>>_.
7232 A http://logging.apache.org/log4j/1.2/[logging library] for Java
7233 developed by the Apache Software Foundation.
7236 Level of severity of a log statement or user space
7237 instrumentation point.
7240 The _Linux Trace Toolkit: next generation_ project.
7242 <<lttng-cli,cmd:lttng>>::
7243 A command-line tool provided by the LTTng-tools project which you
7244 can use to send and receive control messages to and from a
7248 The https://github.com/lttng/lttng-analyses[LTTng analyses] project,
7249 which is a set of analyzing programs that are used to obtain a
7250 higher level view of an LTTng trace.
7252 cmd:lttng-consumerd::
7253 The name of the consumer daemon program.
7256 A utility provided by the LTTng-tools project which can convert
7257 ring buffer files (usually
7258 <<persistent-memory-file-systems,saved on a persistent memory file system>>)
7261 LTTng Documentation::
7264 <<lttng-live,LTTng live>>::
7265 A communication protocol between the relay daemon and live viewers
7266 which makes it possible to see events "live", as they are received by
7269 <<lttng-modules,LTTng-modules>>::
7270 The https://github.com/lttng/lttng-modules[LTTng-modules] project,
7271 which contains the Linux kernel modules to make the Linux kernel
7272 instrumentation points available for LTTng tracing.
7275 The name of the relay daemon program.
7277 cmd:lttng-sessiond::
7278 The name of the session daemon program.
7281 The https://github.com/lttng/lttng-tools[LTTng-tools] project, which
7282 contains the various programs and libraries used to
7283 <<controlling-tracing,control tracing>>.
7285 <<lttng-ust,LTTng-UST>>::
7286 The https://github.com/lttng/lttng-ust[LTTng-UST] project, which
7287 contains libraries to instrument user applications.
7289 <<lttng-ust-agents,LTTng-UST Java agent>>::
7290 A Java package provided by the LTTng-UST project to allow the
7291 LTTng instrumentation of `java.util.logging` and Apache log4j 1.2
7294 <<lttng-ust-agents,LTTng-UST Python agent>>::
7295 A Python package provided by the LTTng-UST project to allow the
7296 LTTng instrumentation of Python logging statements.
7298 <<channel-overwrite-mode-vs-discard-mode,overwrite mode>>::
7299 The event loss mode in which new event records overwrite older
7300 event records when there's no sub-buffer space left to store them.
7302 <<channel-buffering-schemes,per-process buffering>>::
7303 A buffering scheme in which each instrumented process has its own
7304 sub-buffers for a given user space channel.
7306 <<channel-buffering-schemes,per-user buffering>>::
7307 A buffering scheme in which all the processes of a Unix user share the
7308 same sub-buffer for a given user space channel.
7310 <<lttng-relayd,relay daemon>>::
7311 A process which is responsible for receiving the trace data sent by
7312 a distant consumer daemon.
7315 A set of sub-buffers.
7317 <<lttng-sessiond,session daemon>>::
7318 A process which receives control commands from you and orchestrates
7319 the tracers and various LTTng daemons.
7321 <<taking-a-snapshot,snapshot>>::
7322 A copy of the current data of all the sub-buffers of a given tracing
7323 session, saved as trace files.
7326 One part of an LTTng ring buffer which contains event records.
7329 The time information attached to an event when it is emitted.
7332 A set of files which are the concatenations of one or more
7333 flushed sub-buffers.
7336 The action of recording the events emitted by an application
7337 or by a system, or to initiate such recording by controlling
7341 The http://tracecompass.org[Trace Compass] project and application.
7344 An instrumentation point using the tracepoint mechanism of the Linux
7345 kernel or of LTTng-UST.
7347 tracepoint definition::
7348 The definition of a single tracepoint.
7351 The name of a tracepoint.
7353 tracepoint provider::
7354 A set of functions providing tracepoints to an instrumented user
7357 Not to be confused with a _tracepoint provider package_: many tracepoint
7358 providers can exist within a tracepoint provider package.
7360 tracepoint provider package::
7361 One or more tracepoint providers compiled as an object file or as
7365 A software which records emitted events.
7367 <<domain,tracing domain>>::
7368 A namespace for event sources.
7370 <<tracing-group,tracing group>>::
7371 The Unix group in which a Unix user can be to be allowed to trace the
7374 <<tracing-session,tracing session>>::
7375 A stateful dialogue between you and a <<lttng-sessiond,session
7379 An application running in user space, as opposed to a Linux kernel
7380 module, for example.