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`
115 <<notif-trigger-api,Get notified when a channel's buffer usage is too
118 ** You can now embed the whole textual LTTng-tools man pages into the
119 executables at build time with the `--enable-embedded-help`
120 configuration option. Thanks to this option, you don't need the
121 http://www.methods.co.nz/asciidoc/[AsciiDoc] and
122 https://directory.fsf.org/wiki/Xmlto[xmlto] tools at build time, and
123 a manual pager at run time, to get access to this documentation.
125 * **User space tracing**:
126 ** New blocking mode: an LTTng-UST tracepoint can now block until
127 <<channel,sub-buffer>> space is available instead of discarding event
128 records in <<channel-overwrite-mode-vs-discard-mode,discard mode>>.
129 With this feature, you can be sure that no event records are
130 discarded during your application's execution at the expense of
133 For example, the following command lines create a user space tracing
134 channel with an infinite blocking timeout and run an application
135 instrumented with LTTng-UST which is explicitly allowed to block:
141 $ lttng enable-channel --userspace --blocking-timeout=-1 blocking-channel
142 $ lttng enable-event --userspace --channel=blocking-channel --all
144 $ LTTNG_UST_ALLOW_BLOCKING=1 my-app
148 See the complete <<blocking-timeout-example,blocking timeout example>>.
150 * **Linux kernel tracing**:
151 ** Linux 4.10, 4.11, and 4.12 support.
152 ** The thread state dump events recorded by LTTng-modules now contain
153 the task's CPU identifier. This improves the precision of the
154 scheduler model for analyses.
155 ** Extended man:socketpair(2) system call tracing data.
161 What is LTTng? As its name suggests, the _Linux Trace Toolkit: next
162 generation_ is a modern toolkit for tracing Linux systems and
163 applications. So your first question might be:
170 As the history of software engineering progressed and led to what
171 we now take for granted--complex, numerous and
172 interdependent software applications running in parallel on
173 sophisticated operating systems like Linux--the authors of such
174 components, software developers, began feeling a natural
175 urge to have tools that would ensure the robustness and good performance
176 of their masterpieces.
178 One major achievement in this field is, inarguably, the
179 https://www.gnu.org/software/gdb/[GNU debugger (GDB)],
180 an essential tool for developers to find and fix bugs. But even the best
181 debugger won't help make your software run faster, and nowadays, faster
182 software means either more work done by the same hardware, or cheaper
183 hardware for the same work.
185 A _profiler_ is often the tool of choice to identify performance
186 bottlenecks. Profiling is suitable to identify _where_ performance is
187 lost in a given software. The profiler outputs a profile, a statistical
188 summary of observed events, which you may use to discover which
189 functions took the most time to execute. However, a profiler won't
190 report _why_ some identified functions are the bottleneck. Bottlenecks
191 might only occur when specific conditions are met, conditions that are
192 sometimes impossible to capture by a statistical profiler, or impossible
193 to reproduce with an application altered by the overhead of an
194 event-based profiler. For a thorough investigation of software
195 performance issues, a history of execution is essential, with the
196 recorded values of variables and context fields you choose, and
197 with as little influence as possible on the instrumented software. This
198 is where tracing comes in handy.
200 _Tracing_ is a technique used to understand what goes on in a running
201 software system. The software used for tracing is called a _tracer_,
202 which is conceptually similar to a tape recorder. When recording,
203 specific instrumentation points placed in the software source code
204 generate events that are saved on a giant tape: a _trace_ file. You
205 can trace user applications and the operating system at the same time,
206 opening the possibility of resolving a wide range of problems that would
207 otherwise be extremely challenging.
209 Tracing is often compared to _logging_. However, tracers and loggers are
210 two different tools, serving two different purposes. Tracers are
211 designed to record much lower-level events that occur much more
212 frequently than log messages, often in the range of thousands per
213 second, with very little execution overhead. Logging is more appropriate
214 for a very high-level analysis of less frequent events: user accesses,
215 exceptional conditions (errors and warnings, for example), database
216 transactions, instant messaging communications, and such. Simply put,
217 logging is one of the many use cases that can be satisfied with tracing.
219 The list of recorded events inside a trace file can be read manually
220 like a log file for the maximum level of detail, but it is generally
221 much more interesting to perform application-specific analyses to
222 produce reduced statistics and graphs that are useful to resolve a
223 given problem. Trace viewers and analyzers are specialized tools
226 In the end, this is what LTTng is: a powerful, open source set of
227 tools to trace the Linux kernel and user applications at the same time.
228 LTTng is composed of several components actively maintained and
229 developed by its link:/community/#where[community].
232 [[lttng-alternatives]]
233 === Alternatives to noch:{LTTng}
235 Excluding proprietary solutions, a few competing software tracers
238 * https://github.com/dtrace4linux/linux[dtrace4linux] is a port of
239 Sun Microsystems's DTrace to Linux. The cmd:dtrace tool interprets
240 user scripts and is responsible for loading code into the
241 Linux kernel for further execution and collecting the outputted data.
242 * https://en.wikipedia.org/wiki/Berkeley_Packet_Filter[eBPF] is a
243 subsystem in the Linux kernel in which a virtual machine can execute
244 programs passed from the user space to the kernel. You can attach
245 such programs to tracepoints and KProbes thanks to a system call, and
246 they can output data to the user space when executed thanks to
247 different mechanisms (pipe, VM register values, and eBPF maps, to name
249 * https://www.kernel.org/doc/Documentation/trace/ftrace.txt[ftrace]
250 is the de facto function tracer of the Linux kernel. Its user
251 interface is a set of special files in sysfs.
252 * https://perf.wiki.kernel.org/[perf] is
253 a performance analyzing tool for Linux which supports hardware
254 performance counters, tracepoints, as well as other counters and
255 types of probes. perf's controlling utility is the cmd:perf command
257 * http://linux.die.net/man/1/strace[strace]
258 is a command-line utility which records system calls made by a
259 user process, as well as signal deliveries and changes of process
260 state. strace makes use of https://en.wikipedia.org/wiki/Ptrace[ptrace]
261 to fulfill its function.
262 * http://www.sysdig.org/[sysdig], like SystemTap, uses scripts to
263 analyze Linux kernel events. You write scripts, or _chisels_ in
264 sysdig's jargon, in Lua and sysdig executes them while the system is
265 being traced or afterwards. sysdig's interface is the cmd:sysdig
266 command-line tool as well as the curses-based cmd:csysdig tool.
267 * https://sourceware.org/systemtap/[SystemTap] is a Linux kernel and
268 user space tracer which uses custom user scripts to produce plain text
269 traces. SystemTap converts the scripts to the C language, and then
270 compiles them as Linux kernel modules which are loaded to produce
271 trace data. SystemTap's primary user interface is the cmd:stap
274 The main distinctive features of LTTng is that it produces correlated
275 kernel and user space traces, as well as doing so with the lowest
276 overhead amongst other solutions. It produces trace files in the
277 http://diamon.org/ctf[CTF] format, a file format optimized
278 for the production and analyses of multi-gigabyte data.
280 LTTng is the result of more than 10 years of active open source
281 development by a community of passionate developers.
282 LTTng{nbsp}{revision} is currently available on major desktop and server
285 The main interface for tracing control is a single command-line tool
286 named cmd:lttng. The latter can create several tracing sessions, enable
287 and disable events on the fly, filter events efficiently with custom
288 user expressions, start and stop tracing, and much more. LTTng can
289 record the traces on the file system or send them over the network, and
290 keep them totally or partially. You can view the traces once tracing
291 becomes inactive or in real-time.
293 <<installing-lttng,Install LTTng now>> and
294 <<getting-started,start tracing>>!
300 **LTTng** is a set of software <<plumbing,components>> which interact to
301 <<instrumenting,instrument>> the Linux kernel and user applications, and
302 to <<controlling-tracing,control tracing>> (start and stop
303 tracing, enable and disable event rules, and the rest). Those
304 components are bundled into the following packages:
306 * **LTTng-tools**: Libraries and command-line interface to
308 * **LTTng-modules**: Linux kernel modules to instrument and
310 * **LTTng-UST**: Libraries and Java/Python packages to instrument and
311 trace user applications.
313 Most distributions mark the LTTng-modules and LTTng-UST packages as
314 optional when installing LTTng-tools (which is always required). In the
315 following sections, we always provide the steps to install all three,
318 * You only need to install LTTng-modules if you intend to trace the
320 * You only need to install LTTng-UST if you intend to trace user
324 .Availability of LTTng{nbsp}{revision} for major Linux distributions as of 25 July 2017.
326 |Distribution |Available in releases |Alternatives
328 |https://www.ubuntu.com/[Ubuntu]
329 |Ubuntu{nbsp}14.04 _Trusty Tahr_ and Ubuntu{nbsp}16.04 _Xenial Xerus_:
330 <<ubuntu-ppa,use the LTTng Stable{nbsp}{revision} PPA>>.
331 |link:/docs/v2.9#doc-ubuntu[LTTng{nbsp}2.9 for Ubuntu{nbsp}17.04 _Zesty Zapus_].
333 <<building-from-source,Build LTTng{nbsp}{revision} from source>> for
334 other Ubuntu releases.
336 |https://getfedora.org/[Fedora]
338 |link:/docs/v2.9#doc-fedora[LTTng{nbsp}2.9 for Fedora 26].
340 <<building-from-source,Build LTTng{nbsp}{revision} from source>>.
342 |https://www.debian.org/[Debian]
344 |link:/docs/v2.9#doc-debian[LTTng{nbsp}2.9 for Debian "stretch"
345 (stable), Debian "buster" (testing), and Debian "sid" (unstable)].
347 <<building-from-source,Build LTTng{nbsp}{revision} from source>>.
349 |https://www.archlinux.org/[Arch Linux]
351 |link:/docs/v2.9#doc-arch-linux[LTTng{nbsp}2.9 in the latest AUR packages].
353 |https://alpinelinux.org/[Alpine Linux]
355 |link:/docs/v2.9#doc-alpine-linux[LTTng{nbsp}2.9 for Alpine Linux "edge"].
357 <<building-from-source,Build LTTng{nbsp}{revision} from source>>.
359 |https://www.redhat.com/[RHEL] and https://www.suse.com/[SLES]
360 |See http://packages.efficios.com/[EfficiOS Enterprise Packages].
363 |https://buildroot.org/[Buildroot]
365 |link:/docs/v2.9#doc-buildroot[LTTng{nbsp}2.9 for Buildroot{nbsp}2017.02 and
366 Buildroot{nbsp}2017.05].
368 <<building-from-source,Build LTTng{nbsp}{revision} from source>>.
370 |http://www.openembedded.org/wiki/Main_Page[OpenEmbedded] and
371 https://www.yoctoproject.org/[Yocto]
373 |link:/docs/v2.9#doc-oe-yocto[LTTng{nbsp}2.9 for Yocto Project{nbsp}2.3 _Pyro_]
374 (`openembedded-core` layer).
376 <<building-from-source,Build LTTng{nbsp}{revision} from source>>.
381 === [[ubuntu-official-repositories]]Ubuntu
384 ==== noch:{LTTng} Stable {revision} PPA
386 The https://launchpad.net/~lttng/+archive/ubuntu/stable-{revision}[LTTng
387 Stable{nbsp}{revision} PPA] offers the latest stable
388 LTTng{nbsp}{revision} packages for:
390 * Ubuntu{nbsp}14.04 _Trusty Tahr_
391 * Ubuntu{nbsp}16.04 _Xenial Xerus_
393 To install LTTng{nbsp}{revision} from the LTTng Stable{nbsp}{revision} PPA:
395 . Add the LTTng Stable{nbsp}{revision} PPA repository and update the
401 # apt-add-repository ppa:lttng/stable-2.10
406 . Install the main LTTng{nbsp}{revision} packages:
411 # apt-get install lttng-tools
412 # apt-get install lttng-modules-dkms
413 # apt-get install liblttng-ust-dev
417 . **If you need to instrument and trace
418 <<java-application,Java applications>>**, install the LTTng-UST
424 # apt-get install liblttng-ust-agent-java
428 . **If you need to instrument and trace
429 <<python-application,Python{nbsp}3 applications>>**, install the
430 LTTng-UST Python agent:
435 # apt-get install python3-lttngust
440 [[enterprise-distributions]]
441 === RHEL, SUSE, and other enterprise distributions
443 To install LTTng on enterprise Linux distributions, such as Red Hat
444 Enterprise Linux (RHEL) and SUSE Linux Enterprise Server (SUSE), please
445 see http://packages.efficios.com/[EfficiOS Enterprise Packages].
448 [[building-from-source]]
449 === Build from source
451 To build and install LTTng{nbsp}{revision} from source:
453 . Using your distribution's package manager, or from source, install
454 the following dependencies of LTTng-tools and LTTng-UST:
457 * https://sourceforge.net/projects/libuuid/[libuuid]
458 * http://directory.fsf.org/wiki/Popt[popt]
459 * http://liburcu.org/[Userspace RCU]
460 * http://www.xmlsoft.org/[libxml2]
463 . Download, build, and install the latest LTTng-modules{nbsp}{revision}:
469 wget http://lttng.org/files/lttng-modules/lttng-modules-latest-2.10.tar.bz2 &&
470 tar -xf lttng-modules-latest-2.10.tar.bz2 &&
471 cd lttng-modules-2.10.* &&
473 sudo make modules_install &&
478 . Download, build, and install the latest LTTng-UST{nbsp}{revision}:
484 wget http://lttng.org/files/lttng-ust/lttng-ust-latest-2.10.tar.bz2 &&
485 tar -xf lttng-ust-latest-2.10.tar.bz2 &&
486 cd lttng-ust-2.10.* &&
496 .Java and Python application tracing
498 If you need to instrument and trace <<java-application,Java
499 applications>>, pass the `--enable-java-agent-jul`,
500 `--enable-java-agent-log4j`, or `--enable-java-agent-all` options to the
501 `configure` script, depending on which Java logging framework you use.
503 If you need to instrument and trace <<python-application,Python
504 applications>>, pass the `--enable-python-agent` option to the
505 `configure` script. You can set the `PYTHON` environment variable to the
506 path to the Python interpreter for which to install the LTTng-UST Python
514 By default, LTTng-UST libraries are installed to
515 dir:{/usr/local/lib}, which is the de facto directory in which to
516 keep self-compiled and third-party libraries.
518 When <<building-tracepoint-providers-and-user-application,linking an
519 instrumented user application with `liblttng-ust`>>:
521 * Append `/usr/local/lib` to the env:LD_LIBRARY_PATH environment
523 * Pass the `-L/usr/local/lib` and `-Wl,-rpath,/usr/local/lib` options to
524 man:gcc(1), man:g++(1), or man:clang(1).
528 . Download, build, and install the latest LTTng-tools{nbsp}{revision}:
534 wget http://lttng.org/files/lttng-tools/lttng-tools-latest-2.10.tar.bz2 &&
535 tar -xf lttng-tools-latest-2.10.tar.bz2 &&
536 cd lttng-tools-2.10.* &&
544 TIP: The https://github.com/eepp/vlttng[vlttng tool] can do all the
545 previous steps automatically for a given version of LTTng and confine
546 the installed files in a specific directory. This can be useful to test
547 LTTng without installing it on your system.
553 This is a short guide to get started quickly with LTTng kernel and user
556 Before you follow this guide, make sure to <<installing-lttng,install>>
559 This tutorial walks you through the steps to:
561 . <<tracing-the-linux-kernel,Trace the Linux kernel>>.
562 . <<tracing-your-own-user-application,Trace a user application>> written
564 . <<viewing-and-analyzing-your-traces,View and analyze the
568 [[tracing-the-linux-kernel]]
569 === Trace the Linux kernel
571 The following command lines start with the `#` prompt because you need
572 root privileges to trace the Linux kernel. You can also trace the kernel
573 as a regular user if your Unix user is a member of the
574 <<tracing-group,tracing group>>.
576 . Create a <<tracing-session,tracing session>> which writes its traces
577 to dir:{/tmp/my-kernel-trace}:
582 # lttng create my-kernel-session --output=/tmp/my-kernel-trace
586 . List the available kernel tracepoints and system calls:
591 # lttng list --kernel
592 # lttng list --kernel --syscall
596 . Create <<event,event rules>> which match the desired instrumentation
597 point names, for example the `sched_switch` and `sched_process_fork`
598 tracepoints, and the man:open(2) and man:close(2) system calls:
603 # lttng enable-event --kernel sched_switch,sched_process_fork
604 # lttng enable-event --kernel --syscall open,close
608 You can also create an event rule which matches _all_ the Linux kernel
609 tracepoints (this will generate a lot of data when tracing):
614 # lttng enable-event --kernel --all
618 . <<basic-tracing-session-control,Start tracing>>:
627 . Do some operation on your system for a few seconds. For example,
628 load a website, or list the files of a directory.
629 . <<basic-tracing-session-control,Stop tracing>> and destroy the
640 The man:lttng-destroy(1) command does not destroy the trace data; it
641 only destroys the state of the tracing session.
643 . For the sake of this example, make the recorded trace accessible to
649 # chown -R $(whoami) /tmp/my-kernel-trace
653 See <<viewing-and-analyzing-your-traces,View and analyze the
654 recorded events>> to view the recorded events.
657 [[tracing-your-own-user-application]]
658 === Trace a user application
660 This section steps you through a simple example to trace a
661 _Hello world_ program written in C.
663 To create the traceable user application:
665 . Create the tracepoint provider header file, which defines the
666 tracepoints and the events they can generate:
672 #undef TRACEPOINT_PROVIDER
673 #define TRACEPOINT_PROVIDER hello_world
675 #undef TRACEPOINT_INCLUDE
676 #define TRACEPOINT_INCLUDE "./hello-tp.h"
678 #if !defined(_HELLO_TP_H) || defined(TRACEPOINT_HEADER_MULTI_READ)
681 #include <lttng/tracepoint.h>
691 ctf_string(my_string_field, my_string_arg)
692 ctf_integer(int, my_integer_field, my_integer_arg)
696 #endif /* _HELLO_TP_H */
698 #include <lttng/tracepoint-event.h>
702 . Create the tracepoint provider package source file:
708 #define TRACEPOINT_CREATE_PROBES
709 #define TRACEPOINT_DEFINE
711 #include "hello-tp.h"
715 . Build the tracepoint provider package:
720 $ gcc -c -I. hello-tp.c
724 . Create the _Hello World_ application source file:
731 #include "hello-tp.h"
733 int main(int argc, char *argv[])
737 puts("Hello, World!\nPress Enter to continue...");
740 * The following getchar() call is only placed here for the purpose
741 * of this demonstration, to pause the application in order for
742 * you to have time to list its tracepoints. It is not
748 * A tracepoint() call.
750 * Arguments, as defined in hello-tp.h:
752 * 1. Tracepoint provider name (required)
753 * 2. Tracepoint name (required)
754 * 3. my_integer_arg (first user-defined argument)
755 * 4. my_string_arg (second user-defined argument)
757 * Notice the tracepoint provider and tracepoint names are
758 * NOT strings: they are in fact parts of variables that the
759 * macros in hello-tp.h create.
761 tracepoint(hello_world, my_first_tracepoint, 23, "hi there!");
763 for (x = 0; x < argc; ++x) {
764 tracepoint(hello_world, my_first_tracepoint, x, argv[x]);
767 puts("Quitting now!");
768 tracepoint(hello_world, my_first_tracepoint, x * x, "x^2");
775 . Build the application:
784 . Link the application with the tracepoint provider package,
785 `liblttng-ust`, and `libdl`:
790 $ gcc -o hello hello.o hello-tp.o -llttng-ust -ldl
794 Here's the whole build process:
797 .User space tracing tutorial's build steps.
798 image::ust-flow.png[]
800 To trace the user application:
802 . Run the application with a few arguments:
807 $ ./hello world and beyond
816 Press Enter to continue...
820 . Start an LTTng <<lttng-sessiond,session daemon>>:
825 $ lttng-sessiond --daemonize
829 Note that a session daemon might already be running, for example as
830 a service that the distribution's service manager started.
832 . List the available user space tracepoints:
837 $ lttng list --userspace
841 You see the `hello_world:my_first_tracepoint` tracepoint listed
842 under the `./hello` process.
844 . Create a <<tracing-session,tracing session>>:
849 $ lttng create my-user-space-session
853 . Create an <<event,event rule>> which matches the
854 `hello_world:my_first_tracepoint` event name:
859 $ lttng enable-event --userspace hello_world:my_first_tracepoint
863 . <<basic-tracing-session-control,Start tracing>>:
872 . Go back to the running `hello` application and press Enter. The
873 program executes all `tracepoint()` instrumentation points and exits.
874 . <<basic-tracing-session-control,Stop tracing>> and destroy the
885 The man:lttng-destroy(1) command does not destroy the trace data; it
886 only destroys the state of the tracing session.
888 By default, LTTng saves the traces in
889 +$LTTNG_HOME/lttng-traces/__name__-__date__-__time__+,
890 where +__name__+ is the tracing session name. The
891 env:LTTNG_HOME environment variable defaults to `$HOME` if not set.
893 See <<viewing-and-analyzing-your-traces,View and analyze the
894 recorded events>> to view the recorded events.
897 [[viewing-and-analyzing-your-traces]]
898 === View and analyze the recorded events
900 Once you have completed the <<tracing-the-linux-kernel,Trace the Linux
901 kernel>> and <<tracing-your-own-user-application,Trace a user
902 application>> tutorials, you can inspect the recorded events.
904 Many tools are available to read LTTng traces:
906 * **cmd:babeltrace** is a command-line utility which converts trace
907 formats; it supports the format that LTTng produces, CTF, as well as a
908 basic text output which can be ++grep++ed. The cmd:babeltrace command
909 is part of the http://diamon.org/babeltrace[Babeltrace] project.
910 * Babeltrace also includes
911 **https://www.python.org/[Python] bindings** so
912 that you can easily open and read an LTTng trace with your own script,
913 benefiting from the power of Python.
914 * http://tracecompass.org/[**Trace Compass**]
915 is a graphical user interface for viewing and analyzing any type of
916 logs or traces, including LTTng's.
917 * https://github.com/lttng/lttng-analyses[**LTTng analyses**] is a
918 project which includes many high-level analyses of LTTng kernel
919 traces, like scheduling statistics, interrupt frequency distribution,
920 top CPU usage, and more.
922 NOTE: This section assumes that the traces recorded during the previous
923 tutorials were saved to their default location, in the
924 dir:{$LTTNG_HOME/lttng-traces} directory. The env:LTTNG_HOME
925 environment variable defaults to `$HOME` if not set.
928 [[viewing-and-analyzing-your-traces-bt]]
929 ==== Use the cmd:babeltrace command-line tool
931 The simplest way to list all the recorded events of a trace is to pass
932 its path to cmd:babeltrace with no options:
936 $ babeltrace ~/lttng-traces/my-user-space-session*
939 cmd:babeltrace finds all traces recursively within the given path and
940 prints all their events, merging them in chronological order.
942 You can pipe the output of cmd:babeltrace into a tool like man:grep(1) for
947 $ babeltrace /tmp/my-kernel-trace | grep _switch
950 You can pipe the output of cmd:babeltrace into a tool like man:wc(1) to
951 count the recorded events:
955 $ babeltrace /tmp/my-kernel-trace | grep _open | wc --lines
959 [[viewing-and-analyzing-your-traces-bt-python]]
960 ==== Use the Babeltrace Python bindings
962 The <<viewing-and-analyzing-your-traces-bt,text output of cmd:babeltrace>>
963 is useful to isolate events by simple matching using man:grep(1) and
964 similar utilities. However, more elaborate filters, such as keeping only
965 event records with a field value falling within a specific range, are
966 not trivial to write using a shell. Moreover, reductions and even the
967 most basic computations involving multiple event records are virtually
968 impossible to implement.
970 Fortunately, Babeltrace ships with Python 3 bindings which makes it easy
971 to read the event records of an LTTng trace sequentially and compute the
974 The following script accepts an LTTng Linux kernel trace path as its
975 first argument and prints the short names of the top 5 running processes
976 on CPU 0 during the whole trace:
981 from collections import Counter
987 if len(sys.argv) != 2:
988 msg = 'Usage: python3 {} TRACEPATH'.format(sys.argv[0])
989 print(msg, file=sys.stderr)
992 # A trace collection contains one or more traces
993 col = babeltrace.TraceCollection()
995 # Add the trace provided by the user (LTTng traces always have
997 if col.add_trace(sys.argv[1], 'ctf') is None:
998 raise RuntimeError('Cannot add trace')
1000 # This counter dict contains execution times:
1002 # task command name -> total execution time (ns)
1003 exec_times = Counter()
1005 # This contains the last `sched_switch` timestamp
1009 for event in col.events:
1010 # Keep only `sched_switch` events
1011 if event.name != 'sched_switch':
1014 # Keep only events which happened on CPU 0
1015 if event['cpu_id'] != 0:
1019 cur_ts = event.timestamp
1025 # Previous task command (short) name
1026 prev_comm = event['prev_comm']
1028 # Initialize entry in our dict if not yet done
1029 if prev_comm not in exec_times:
1030 exec_times[prev_comm] = 0
1032 # Compute previous command execution time
1033 diff = cur_ts - last_ts
1035 # Update execution time of this command
1036 exec_times[prev_comm] += diff
1038 # Update last timestamp
1042 for name, ns in exec_times.most_common(5):
1044 print('{:20}{} s'.format(name, s))
1049 if __name__ == '__main__':
1050 sys.exit(0 if top5proc() else 1)
1057 $ python3 top5proc.py /tmp/my-kernel-trace/kernel
1063 swapper/0 48.607245889 s
1064 chromium 7.192738188 s
1065 pavucontrol 0.709894415 s
1066 Compositor 0.660867933 s
1067 Xorg.bin 0.616753786 s
1070 Note that `swapper/0` is the "idle" process of CPU 0 on Linux; since we
1071 weren't using the CPU that much when tracing, its first position in the
1076 == [[understanding-lttng]]Core concepts
1078 From a user's perspective, the LTTng system is built on a few concepts,
1079 or objects, on which the <<lttng-cli,cmd:lttng command-line tool>>
1080 operates by sending commands to the <<lttng-sessiond,session daemon>>.
1081 Understanding how those objects relate to eachother is key in mastering
1084 The core concepts are:
1086 * <<tracing-session,Tracing session>>
1087 * <<domain,Tracing domain>>
1088 * <<channel,Channel and ring buffer>>
1089 * <<"event","Instrumentation point, event rule, event, and event record">>
1095 A _tracing session_ is a stateful dialogue between you and
1096 a <<lttng-sessiond,session daemon>>. You can
1097 <<creating-destroying-tracing-sessions,create a new tracing
1098 session>> with the `lttng create` command.
1100 Anything that you do when you control LTTng tracers happens within a
1101 tracing session. In particular, a tracing session:
1104 * Has its own set of trace files.
1105 * Has its own state of activity (started or stopped).
1106 * Has its own <<tracing-session-mode,mode>> (local, network streaming,
1108 * Has its own <<channel,channels>> which have their own
1109 <<event,event rules>>.
1112 .A _tracing session_ contains <<channel,channels>> that are members of <<domain,tracing domains>> and contain <<event,event rules>>.
1113 image::concepts.png[]
1115 Those attributes and objects are completely isolated between different
1118 A tracing session is analogous to a cash machine session:
1119 the operations you do on the banking system through the cash machine do
1120 not alter the data of other users of the same system. In the case of
1121 the cash machine, a session lasts as long as your bank card is inside.
1122 In the case of LTTng, a tracing session lasts from the `lttng create`
1123 command to the `lttng destroy` command.
1126 .Each Unix user has its own set of tracing sessions.
1127 image::many-sessions.png[]
1130 [[tracing-session-mode]]
1131 ==== Tracing session mode
1133 LTTng can send the generated trace data to different locations. The
1134 _tracing session mode_ dictates where to send it. The following modes
1135 are available in LTTng{nbsp}{revision}:
1138 LTTng writes the traces to the file system of the machine being traced
1141 Network streaming mode::
1142 LTTng sends the traces over the network to a
1143 <<lttng-relayd,relay daemon>> running on a remote system.
1146 LTTng does not write the traces by default. Instead, you can request
1147 LTTng to <<taking-a-snapshot,take a snapshot>>, that is, a copy of the
1148 current tracing buffers, and to write it to the target's file system
1149 or to send it over the network to a <<lttng-relayd,relay daemon>>
1150 running on a remote system.
1153 This mode is similar to the network streaming mode, but a live
1154 trace viewer can connect to the distant relay daemon to
1155 <<lttng-live,view event records as LTTng generates them>> by
1162 A _tracing domain_ is a namespace for event sources. A tracing domain
1163 has its own properties and features.
1165 There are currently five available tracing domains:
1169 * `java.util.logging` (JUL)
1173 You must specify a tracing domain when using some commands to avoid
1174 ambiguity. For example, since all the domains support named tracepoints
1175 as event sources (instrumentation points that you manually insert in the
1176 source code), you need to specify a tracing domain when
1177 <<enabling-disabling-events,creating an event rule>> because all the
1178 tracing domains could have tracepoints with the same names.
1180 Some features are reserved to specific tracing domains. Dynamic function
1181 entry and return instrumentation points, for example, are currently only
1182 supported in the Linux kernel tracing domain, but support for other
1183 tracing domains could be added in the future.
1185 You can create <<channel,channels>> in the Linux kernel and user space
1186 tracing domains. The other tracing domains have a single default
1191 === Channel and ring buffer
1193 A _channel_ is an object which is responsible for a set of ring buffers.
1194 Each ring buffer is divided into multiple sub-buffers. When an LTTng
1195 tracer emits an event, it can record it to one or more
1196 sub-buffers. The attributes of a channel determine what to do when
1197 there's no space left for a new event record because all sub-buffers
1198 are full, where to send a full sub-buffer, and other behaviours.
1200 A channel is always associated to a <<domain,tracing domain>>. The
1201 `java.util.logging` (JUL), log4j, and Python tracing domains each have
1202 a default channel which you cannot configure.
1204 A channel also owns <<event,event rules>>. When an LTTng tracer emits
1205 an event, it records it to the sub-buffers of all
1206 the enabled channels with a satisfied event rule, as long as those
1207 channels are part of active <<tracing-session,tracing sessions>>.
1210 [[channel-buffering-schemes]]
1211 ==== Per-user vs. per-process buffering schemes
1213 A channel has at least one ring buffer _per CPU_. LTTng always
1214 records an event to the ring buffer associated to the CPU on which it
1217 Two _buffering schemes_ are available when you
1218 <<enabling-disabling-channels,create a channel>> in the
1219 user space <<domain,tracing domain>>:
1221 Per-user buffering::
1222 Allocate one set of ring buffers--one per CPU--shared by all the
1223 instrumented processes of each Unix user.
1227 .Per-user buffering scheme.
1228 image::per-user-buffering.png[]
1231 Per-process buffering::
1232 Allocate one set of ring buffers--one per CPU--for each
1233 instrumented process.
1237 .Per-process buffering scheme.
1238 image::per-process-buffering.png[]
1241 The per-process buffering scheme tends to consume more memory than the
1242 per-user option because systems generally have more instrumented
1243 processes than Unix users running instrumented processes. However, the
1244 per-process buffering scheme ensures that one process having a high
1245 event throughput won't fill all the shared sub-buffers of the same
1248 The Linux kernel tracing domain has only one available buffering scheme
1249 which is to allocate a single set of ring buffers for the whole system.
1250 This scheme is similar to the per-user option, but with a single, global
1251 user "running" the kernel.
1254 [[channel-overwrite-mode-vs-discard-mode]]
1255 ==== Overwrite vs. discard event loss modes
1257 When an event occurs, LTTng records it to a specific sub-buffer (yellow
1258 arc in the following animation) of a specific channel's ring buffer.
1259 When there's no space left in a sub-buffer, the tracer marks it as
1260 consumable (red) and another, empty sub-buffer starts receiving the
1261 following event records. A <<lttng-consumerd,consumer daemon>>
1262 eventually consumes the marked sub-buffer (returns to white).
1265 [role="docsvg-channel-subbuf-anim"]
1270 In an ideal world, sub-buffers are consumed faster than they are filled,
1271 as is the case in the previous animation. In the real world,
1272 however, all sub-buffers can be full at some point, leaving no space to
1273 record the following events.
1275 By default, LTTng-modules and LTTng-UST are _non-blocking_ tracers: when
1276 no empty sub-buffer is available, it is acceptable to lose event records
1277 when the alternative would be to cause substantial delays in the
1278 instrumented application's execution. LTTng privileges performance over
1279 integrity; it aims at perturbing the traced system as little as possible
1280 in order to make tracing of subtle race conditions and rare interrupt
1283 Starting from LTTng{nbsp}2.10, the LTTng user space tracer, LTTng-UST,
1284 supports a _blocking mode_. See the <<blocking-timeout-example,blocking
1285 timeout example>> to learn how to use the blocking mode.
1287 When it comes to losing event records because no empty sub-buffer is
1288 available, or because the <<opt-blocking-timeout,blocking timeout>> is
1289 reached, the channel's _event loss mode_ determines what to do. The
1290 available event loss modes are:
1293 Drop the newest event records until a the tracer
1294 releases a sub-buffer.
1297 Clear the sub-buffer containing the oldest event records and start
1298 writing the newest event records there.
1300 This mode is sometimes called _flight recorder mode_ because it's
1302 https://en.wikipedia.org/wiki/Flight_recorder[flight recorder]:
1303 always keep a fixed amount of the latest data.
1305 Which mechanism you should choose depends on your context: prioritize
1306 the newest or the oldest event records in the ring buffer?
1308 Beware that, in overwrite mode, the tracer abandons a whole sub-buffer
1309 as soon as a there's no space left for a new event record, whereas in
1310 discard mode, the tracer only discards the event record that doesn't
1313 In discard mode, LTTng increments a count of lost event records when
1314 an event record is lost and saves this count to the trace. In
1315 overwrite mode, LTTng keeps no information when it overwrites a
1316 sub-buffer before consuming it.
1318 There are a few ways to decrease your probability of losing event
1320 <<channel-subbuf-size-vs-subbuf-count,Sub-buffer count and size>> shows
1321 how you can fine-une the sub-buffer count and size of a channel to
1322 virtually stop losing event records, though at the cost of greater
1326 [[channel-subbuf-size-vs-subbuf-count]]
1327 ==== Sub-buffer count and size
1329 When you <<enabling-disabling-channels,create a channel>>, you can
1330 set its number of sub-buffers and their size.
1332 Note that there is noticeable CPU overhead introduced when
1333 switching sub-buffers (marking a full one as consumable and switching
1334 to an empty one for the following events to be recorded). Knowing this,
1335 the following list presents a few practical situations along with how
1336 to configure the sub-buffer count and size for them:
1338 * **High event throughput**: In general, prefer bigger sub-buffers to
1339 lower the risk of losing event records.
1341 Having bigger sub-buffers also ensures a lower
1342 <<channel-switch-timer,sub-buffer switching frequency>>.
1344 The number of sub-buffers is only meaningful if you create the channel
1345 in overwrite mode: in this case, if a sub-buffer overwrite happens, the
1346 other sub-buffers are left unaltered.
1348 * **Low event throughput**: In general, prefer smaller sub-buffers
1349 since the risk of losing event records is low.
1351 Because events occur less frequently, the sub-buffer switching frequency
1352 should remain low and thus the tracer's overhead should not be a
1355 * **Low memory system**: If your target system has a low memory
1356 limit, prefer fewer first, then smaller sub-buffers.
1358 Even if the system is limited in memory, you want to keep the
1359 sub-buffers as big as possible to avoid a high sub-buffer switching
1362 Note that LTTng uses http://diamon.org/ctf/[CTF] as its trace format,
1363 which means event data is very compact. For example, the average
1364 LTTng kernel event record weights about 32{nbsp}bytes. Thus, a
1365 sub-buffer size of 1{nbsp}MiB is considered big.
1367 The previous situations highlight the major trade-off between a few big
1368 sub-buffers and more, smaller sub-buffers: sub-buffer switching
1369 frequency vs. how much data is lost in overwrite mode. Assuming a
1370 constant event throughput and using the overwrite mode, the two
1371 following configurations have the same ring buffer total size:
1374 [role="docsvg-channel-subbuf-size-vs-count-anim"]
1379 * **2 sub-buffers of 4{nbsp}MiB each**: Expect a very low sub-buffer
1380 switching frequency, but if a sub-buffer overwrite happens, half of
1381 the event records so far (4{nbsp}MiB) are definitely lost.
1382 * **8 sub-buffers of 1{nbsp}MiB each**: Expect 4{nbsp}times the tracer's
1383 overhead as the previous configuration, but if a sub-buffer
1384 overwrite happens, only the eighth of event records so far are
1387 In discard mode, the sub-buffers count parameter is pointless: use two
1388 sub-buffers and set their size according to the requirements of your
1392 [[channel-switch-timer]]
1393 ==== Switch timer period
1395 The _switch timer period_ is an important configurable attribute of
1396 a channel to ensure periodic sub-buffer flushing.
1398 When the _switch timer_ expires, a sub-buffer switch happens. You can
1399 set the switch timer period attribute when you
1400 <<enabling-disabling-channels,create a channel>> to ensure that event
1401 data is consumed and committed to trace files or to a distant relay
1402 daemon periodically in case of a low event throughput.
1405 [role="docsvg-channel-switch-timer"]
1410 This attribute is also convenient when you use big sub-buffers to cope
1411 with a sporadic high event throughput, even if the throughput is
1415 [[channel-read-timer]]
1416 ==== Read timer period
1418 By default, the LTTng tracers use a notification mechanism to signal a
1419 full sub-buffer so that a consumer daemon can consume it. When such
1420 notifications must be avoided, for example in real-time applications,
1421 you can use the channel's _read timer_ instead. When the read timer
1422 fires, the <<lttng-consumerd,consumer daemon>> checks for full,
1423 consumable sub-buffers.
1426 [[tracefile-rotation]]
1427 ==== Trace file count and size
1429 By default, trace files can grow as large as needed. You can set the
1430 maximum size of each trace file that a channel writes when you
1431 <<enabling-disabling-channels,create a channel>>. When the size of
1432 a trace file reaches the channel's fixed maximum size, LTTng creates
1433 another file to contain the next event records. LTTng appends a file
1434 count to each trace file name in this case.
1436 If you set the trace file size attribute when you create a channel, the
1437 maximum number of trace files that LTTng creates is _unlimited_ by
1438 default. To limit them, you can also set a maximum number of trace
1439 files. When the number of trace files reaches the channel's fixed
1440 maximum count, the oldest trace file is overwritten. This mechanism is
1441 called _trace file rotation_.
1445 === Instrumentation point, event rule, event, and event record
1447 An _event rule_ is a set of conditions which must be **all** satisfied
1448 for LTTng to record an occuring event.
1450 You set the conditions when you <<enabling-disabling-events,create
1453 You always attach an event rule to <<channel,channel>> when you create
1456 When an event passes the conditions of an event rule, LTTng records it
1457 in one of the attached channel's sub-buffers.
1459 The available conditions, as of LTTng{nbsp}{revision}, are:
1461 * The event rule _is enabled_.
1462 * The instrumentation point's type _is{nbsp}T_.
1463 * The instrumentation point's name (sometimes called _event name_)
1464 _matches{nbsp}N_, but _is not{nbsp}E_.
1465 * The instrumentation point's log level _is as severe as{nbsp}L_, or
1466 _is exactly{nbsp}L_.
1467 * The fields of the event's payload _satisfy_ a filter
1468 expression{nbsp}__F__.
1470 As you can see, all the conditions but the dynamic filter are related to
1471 the event rule's status or to the instrumentation point, not to the
1472 occurring events. This is why, without a filter, checking if an event
1473 passes an event rule is not a dynamic task: when you create or modify an
1474 event rule, all the tracers of its tracing domain enable or disable the
1475 instrumentation points themselves once. This is possible because the
1476 attributes of an instrumentation point (type, name, and log level) are
1477 defined statically. In other words, without a dynamic filter, the tracer
1478 _does not evaluate_ the arguments of an instrumentation point unless it
1479 matches an enabled event rule.
1481 Note that, for LTTng to record an event, the <<channel,channel>> to
1482 which a matching event rule is attached must also be enabled, and the
1483 tracing session owning this channel must be active.
1486 .Logical path from an instrumentation point to an event record.
1487 image::event-rule.png[]
1489 .Event, event record, or event rule?
1491 With so many similar terms, it's easy to get confused.
1493 An **event** is the consequence of the execution of an _instrumentation
1494 point_, like a tracepoint that you manually place in some source code,
1495 or a Linux kernel KProbe. An event is said to _occur_ at a specific
1496 time. Different actions can be taken upon the occurrence of an event,
1497 like record the event's payload to a buffer.
1499 An **event record** is the representation of an event in a sub-buffer. A
1500 tracer is responsible for capturing the payload of an event, current
1501 context variables, the event's ID, and the event's timestamp. LTTng
1502 can append this sub-buffer to a trace file.
1504 An **event rule** is a set of conditions which must all be satisfied for
1505 LTTng to record an occuring event. Events still occur without
1506 satisfying event rules, but LTTng does not record them.
1511 == Components of noch:{LTTng}
1513 The second _T_ in _LTTng_ stands for _toolkit_: it would be wrong
1514 to call LTTng a simple _tool_ since it is composed of multiple
1515 interacting components. This section describes those components,
1516 explains their respective roles, and shows how they connect together to
1517 form the LTTng ecosystem.
1519 The following diagram shows how the most important components of LTTng
1520 interact with user applications, the Linux kernel, and you:
1523 .Control and trace data paths between LTTng components.
1524 image::plumbing.png[]
1526 The LTTng project incorporates:
1528 * **LTTng-tools**: Libraries and command-line interface to
1529 control tracing sessions.
1530 ** <<lttng-sessiond,Session daemon>> (man:lttng-sessiond(8)).
1531 ** <<lttng-consumerd,Consumer daemon>> (man:lttng-consumerd(8)).
1532 ** <<lttng-relayd,Relay daemon>> (man:lttng-relayd(8)).
1533 ** <<liblttng-ctl-lttng,Tracing control library>> (`liblttng-ctl`).
1534 ** <<lttng-cli,Tracing control command-line tool>> (man:lttng(1)).
1535 * **LTTng-UST**: Libraries and Java/Python packages to trace user
1537 ** <<lttng-ust,User space tracing library>> (`liblttng-ust`) and its
1538 headers to instrument and trace any native user application.
1539 ** <<prebuilt-ust-helpers,Preloadable user space tracing helpers>>:
1540 *** `liblttng-ust-libc-wrapper`
1541 *** `liblttng-ust-pthread-wrapper`
1542 *** `liblttng-ust-cyg-profile`
1543 *** `liblttng-ust-cyg-profile-fast`
1544 *** `liblttng-ust-dl`
1545 ** User space tracepoint provider source files generator command-line
1546 tool (man:lttng-gen-tp(1)).
1547 ** <<lttng-ust-agents,LTTng-UST Java agent>> to instrument and trace
1548 Java applications using `java.util.logging` or
1549 Apache log4j 1.2 logging.
1550 ** <<lttng-ust-agents,LTTng-UST Python agent>> to instrument
1551 Python applications using the standard `logging` package.
1552 * **LTTng-modules**: <<lttng-modules,Linux kernel modules>> to trace
1554 ** LTTng kernel tracer module.
1555 ** Tracing ring buffer kernel modules.
1556 ** Probe kernel modules.
1557 ** LTTng logger kernel module.
1561 === Tracing control command-line interface
1564 .The tracing control command-line interface.
1565 image::plumbing-lttng-cli.png[]
1567 The _man:lttng(1) command-line tool_ is the standard user interface to
1568 control LTTng <<tracing-session,tracing sessions>>. The cmd:lttng tool
1569 is part of LTTng-tools.
1571 The cmd:lttng tool is linked with
1572 <<liblttng-ctl-lttng,`liblttng-ctl`>> to communicate with
1573 one or more <<lttng-sessiond,session daemons>> behind the scenes.
1575 The cmd:lttng tool has a Git-like interface:
1579 $ lttng <GENERAL OPTIONS> <COMMAND> <COMMAND OPTIONS>
1582 The <<controlling-tracing,Tracing control>> section explores the
1583 available features of LTTng using the cmd:lttng tool.
1586 [[liblttng-ctl-lttng]]
1587 === Tracing control library
1590 .The tracing control library.
1591 image::plumbing-liblttng-ctl.png[]
1593 The _LTTng control library_, `liblttng-ctl`, is used to communicate
1594 with a <<lttng-sessiond,session daemon>> using a C API that hides the
1595 underlying protocol's details. `liblttng-ctl` is part of LTTng-tools.
1597 The <<lttng-cli,cmd:lttng command-line tool>>
1598 is linked with `liblttng-ctl`.
1600 You can use `liblttng-ctl` in C or $$C++$$ source code by including its
1605 #include <lttng/lttng.h>
1608 Some objects are referenced by name (C string), such as tracing
1609 sessions, but most of them require to create a handle first using
1610 `lttng_create_handle()`.
1612 The best available developer documentation for `liblttng-ctl` is, as of
1613 LTTng{nbsp}{revision}, its installed header files. Every function and
1614 structure is thoroughly documented.
1618 === User space tracing library
1621 .The user space tracing library.
1622 image::plumbing-liblttng-ust.png[]
1624 The _user space tracing library_, `liblttng-ust` (see man:lttng-ust(3)),
1625 is the LTTng user space tracer. It receives commands from a
1626 <<lttng-sessiond,session daemon>>, for example to
1627 enable and disable specific instrumentation points, and writes event
1628 records to ring buffers shared with a
1629 <<lttng-consumerd,consumer daemon>>.
1630 `liblttng-ust` is part of LTTng-UST.
1632 Public C header files are installed beside `liblttng-ust` to
1633 instrument any <<c-application,C or $$C++$$ application>>.
1635 <<lttng-ust-agents,LTTng-UST agents>>, which are regular Java and Python
1636 packages, use their own library providing tracepoints which is
1637 linked with `liblttng-ust`.
1639 An application or library does not have to initialize `liblttng-ust`
1640 manually: its constructor does the necessary tasks to properly register
1641 to a session daemon. The initialization phase also enables the
1642 instrumentation points matching the <<event,event rules>> that you
1646 [[lttng-ust-agents]]
1647 === User space tracing agents
1650 .The user space tracing agents.
1651 image::plumbing-lttng-ust-agents.png[]
1653 The _LTTng-UST Java and Python agents_ are regular Java and Python
1654 packages which add LTTng tracing capabilities to the
1655 native logging frameworks. The LTTng-UST agents are part of LTTng-UST.
1657 In the case of Java, the
1658 https://docs.oracle.com/javase/7/docs/api/java/util/logging/package-summary.html[`java.util.logging`
1659 core logging facilities] and
1660 https://logging.apache.org/log4j/1.2/[Apache log4j 1.2] are supported.
1661 Note that Apache Log4{nbsp}2 is not supported.
1663 In the case of Python, the standard
1664 https://docs.python.org/3/library/logging.html[`logging`] package
1665 is supported. Both Python 2 and Python 3 modules can import the
1666 LTTng-UST Python agent package.
1668 The applications using the LTTng-UST agents are in the
1669 `java.util.logging` (JUL),
1670 log4j, and Python <<domain,tracing domains>>.
1672 Both agents use the same mechanism to trace the log statements. When an
1673 agent is initialized, it creates a log handler that attaches to the root
1674 logger. The agent also registers to a <<lttng-sessiond,session daemon>>.
1675 When the application executes a log statement, it is passed to the
1676 agent's log handler by the root logger. The agent's log handler calls a
1677 native function in a tracepoint provider package shared library linked
1678 with <<lttng-ust,`liblttng-ust`>>, passing the formatted log message and
1679 other fields, like its logger name and its log level. This native
1680 function contains a user space instrumentation point, hence tracing the
1683 The log level condition of an
1684 <<event,event rule>> is considered when tracing
1685 a Java or a Python application, and it's compatible with the standard
1686 JUL, log4j, and Python log levels.
1690 === LTTng kernel modules
1693 .The LTTng kernel modules.
1694 image::plumbing-lttng-modules.png[]
1696 The _LTTng kernel modules_ are a set of Linux kernel modules
1697 which implement the kernel tracer of the LTTng project. The LTTng
1698 kernel modules are part of LTTng-modules.
1700 The LTTng kernel modules include:
1702 * A set of _probe_ modules.
1704 Each module attaches to a specific subsystem
1705 of the Linux kernel using its tracepoint instrument points. There are
1706 also modules to attach to the entry and return points of the Linux
1707 system call functions.
1709 * _Ring buffer_ modules.
1711 A ring buffer implementation is provided as kernel modules. The LTTng
1712 kernel tracer writes to the ring buffer; a
1713 <<lttng-consumerd,consumer daemon>> reads from the ring buffer.
1715 * The _LTTng kernel tracer_ module.
1716 * The _LTTng logger_ module.
1718 The LTTng logger module implements the special path:{/proc/lttng-logger}
1719 file so that any executable can generate LTTng events by opening and
1720 writing to this file.
1722 See <<proc-lttng-logger-abi,LTTng logger>>.
1724 Generally, you do not have to load the LTTng kernel modules manually
1725 (using man:modprobe(8), for example): a root <<lttng-sessiond,session
1726 daemon>> loads the necessary modules when starting. If you have extra
1727 probe modules, you can specify to load them to the session daemon on
1730 The LTTng kernel modules are installed in
1731 +/usr/lib/modules/__release__/extra+ by default, where +__release__+ is
1732 the kernel release (see `uname --kernel-release`).
1739 .The session daemon.
1740 image::plumbing-sessiond.png[]
1742 The _session daemon_, man:lttng-sessiond(8), is a daemon responsible for
1743 managing tracing sessions and for controlling the various components of
1744 LTTng. The session daemon is part of LTTng-tools.
1746 The session daemon sends control requests to and receives control
1749 * The <<lttng-ust,user space tracing library>>.
1751 Any instance of the user space tracing library first registers to
1752 a session daemon. Then, the session daemon can send requests to
1753 this instance, such as:
1756 ** Get the list of tracepoints.
1757 ** Share an <<event,event rule>> so that the user space tracing library
1758 can enable or disable tracepoints. Amongst the possible conditions
1759 of an event rule is a filter expression which `liblttng-ust` evalutes
1760 when an event occurs.
1761 ** Share <<channel,channel>> attributes and ring buffer locations.
1764 The session daemon and the user space tracing library use a Unix
1765 domain socket for their communication.
1767 * The <<lttng-ust-agents,user space tracing agents>>.
1769 Any instance of a user space tracing agent first registers to
1770 a session daemon. Then, the session daemon can send requests to
1771 this instance, such as:
1774 ** Get the list of loggers.
1775 ** Enable or disable a specific logger.
1778 The session daemon and the user space tracing agent use a TCP connection
1779 for their communication.
1781 * The <<lttng-modules,LTTng kernel tracer>>.
1782 * The <<lttng-consumerd,consumer daemon>>.
1784 The session daemon sends requests to the consumer daemon to instruct
1785 it where to send the trace data streams, amongst other information.
1787 * The <<lttng-relayd,relay daemon>>.
1789 The session daemon receives commands from the
1790 <<liblttng-ctl-lttng,tracing control library>>.
1792 The root session daemon loads the appropriate
1793 <<lttng-modules,LTTng kernel modules>> on startup. It also spawns
1794 a <<lttng-consumerd,consumer daemon>> as soon as you create
1795 an <<event,event rule>>.
1797 The session daemon does not send and receive trace data: this is the
1798 role of the <<lttng-consumerd,consumer daemon>> and
1799 <<lttng-relayd,relay daemon>>. It does, however, generate the
1800 http://diamon.org/ctf/[CTF] metadata stream.
1802 Each Unix user can have its own session daemon instance. The
1803 tracing sessions managed by different session daemons are completely
1806 The root user's session daemon is the only one which is
1807 allowed to control the LTTng kernel tracer, and its spawned consumer
1808 daemon is the only one which is allowed to consume trace data from the
1809 LTTng kernel tracer. Note, however, that any Unix user which is a member
1810 of the <<tracing-group,tracing group>> is allowed
1811 to create <<channel,channels>> in the
1812 Linux kernel <<domain,tracing domain>>, and thus to trace the Linux
1815 The <<lttng-cli,cmd:lttng command-line tool>> automatically starts a
1816 session daemon when using its `create` command if none is currently
1817 running. You can also start the session daemon manually.
1824 .The consumer daemon.
1825 image::plumbing-consumerd.png[]
1827 The _consumer daemon_, man:lttng-consumerd(8), is a daemon which shares
1828 ring buffers with user applications or with the LTTng kernel modules to
1829 collect trace data and send it to some location (on disk or to a
1830 <<lttng-relayd,relay daemon>> over the network). The consumer daemon
1831 is part of LTTng-tools.
1833 You do not start a consumer daemon manually: a consumer daemon is always
1834 spawned by a <<lttng-sessiond,session daemon>> as soon as you create an
1835 <<event,event rule>>, that is, before you start tracing. When you kill
1836 its owner session daemon, the consumer daemon also exits because it is
1837 the session daemon's child process. Command-line options of
1838 man:lttng-sessiond(8) target the consumer daemon process.
1840 There are up to two running consumer daemons per Unix user, whereas only
1841 one session daemon can run per user. This is because each process can be
1842 either 32-bit or 64-bit: if the target system runs a mixture of 32-bit
1843 and 64-bit processes, it is more efficient to have separate
1844 corresponding 32-bit and 64-bit consumer daemons. The root user is an
1845 exception: it can have up to _three_ running consumer daemons: 32-bit
1846 and 64-bit instances for its user applications, and one more
1847 reserved for collecting kernel trace data.
1855 image::plumbing-relayd.png[]
1857 The _relay daemon_, man:lttng-relayd(8), is a daemon acting as a bridge
1858 between remote session and consumer daemons, local trace files, and a
1859 remote live trace viewer. The relay daemon is part of LTTng-tools.
1861 The main purpose of the relay daemon is to implement a receiver of
1862 <<sending-trace-data-over-the-network,trace data over the network>>.
1863 This is useful when the target system does not have much file system
1864 space to record trace files locally.
1866 The relay daemon is also a server to which a
1867 <<lttng-live,live trace viewer>> can
1868 connect. The live trace viewer sends requests to the relay daemon to
1869 receive trace data as the target system emits events. The
1870 communication protocol is named _LTTng live_; it is used over TCP
1873 Note that you can start the relay daemon on the target system directly.
1874 This is the setup of choice when the use case is to view events as
1875 the target system emits them without the need of a remote system.
1879 == [[using-lttng]]Instrumentation
1881 There are many examples of tracing and monitoring in our everyday life:
1883 * You have access to real-time and historical weather reports and
1884 forecasts thanks to weather stations installed around the country.
1885 * You know your heart is safe thanks to an electrocardiogram.
1886 * You make sure not to drive your car too fast and to have enough fuel
1887 to reach your destination thanks to gauges visible on your dashboard.
1889 All the previous examples have something in common: they rely on
1890 **instruments**. Without the electrodes attached to the surface of your
1891 body's skin, cardiac monitoring is futile.
1893 LTTng, as a tracer, is no different from those real life examples. If
1894 you're about to trace a software system or, in other words, record its
1895 history of execution, you better have **instrumentation points** in the
1896 subject you're tracing, that is, the actual software.
1898 Various ways were developed to instrument a piece of software for LTTng
1899 tracing. The most straightforward one is to manually place
1900 instrumentation points, called _tracepoints_, in the software's source
1901 code. It is also possible to add instrumentation points dynamically in
1902 the Linux kernel <<domain,tracing domain>>.
1904 If you're only interested in tracing the Linux kernel, your
1905 instrumentation needs are probably already covered by LTTng's built-in
1906 <<lttng-modules,Linux kernel tracepoints>>. You may also wish to trace a
1907 user application which is already instrumented for LTTng tracing.
1908 In such cases, you can skip this whole section and read the topics of
1909 the <<controlling-tracing,Tracing control>> section.
1911 Many methods are available to instrument a piece of software for LTTng
1914 * <<c-application,User space instrumentation for C and $$C++$$
1916 * <<prebuilt-ust-helpers,Prebuilt user space tracing helpers>>.
1917 * <<java-application,User space Java agent>>.
1918 * <<python-application,User space Python agent>>.
1919 * <<proc-lttng-logger-abi,LTTng logger>>.
1920 * <<instrumenting-linux-kernel,LTTng kernel tracepoints>>.
1924 === [[cxx-application]]User space instrumentation for C and $$C++$$ applications
1926 The procedure to instrument a C or $$C++$$ user application with
1927 the <<lttng-ust,LTTng user space tracing library>>, `liblttng-ust`, is:
1929 . <<tracepoint-provider,Create the source files of a tracepoint provider
1931 . <<probing-the-application-source-code,Add tracepoints to
1932 the application's source code>>.
1933 . <<building-tracepoint-providers-and-user-application,Build and link
1934 a tracepoint provider package and the user application>>.
1936 If you need quick, man:printf(3)-like instrumentation, you can skip
1937 those steps and use <<tracef,`tracef()`>> or <<tracelog,`tracelog()`>>
1940 IMPORTANT: You need to <<installing-lttng,install>> LTTng-UST to
1941 instrument a user application with `liblttng-ust`.
1944 [[tracepoint-provider]]
1945 ==== Create the source files of a tracepoint provider package
1947 A _tracepoint provider_ is a set of compiled functions which provide
1948 **tracepoints** to an application, the type of instrumentation point
1949 supported by LTTng-UST. Those functions can emit events with
1950 user-defined fields and serialize those events as event records to one
1951 or more LTTng-UST <<channel,channel>> sub-buffers. The `tracepoint()`
1952 macro, which you <<probing-the-application-source-code,insert in a user
1953 application's source code>>, calls those functions.
1955 A _tracepoint provider package_ is an object file (`.o`) or a shared
1956 library (`.so`) which contains one or more tracepoint providers.
1957 Its source files are:
1959 * One or more <<tpp-header,tracepoint provider header>> (`.h`).
1960 * A <<tpp-source,tracepoint provider package source>> (`.c`).
1962 A tracepoint provider package is dynamically linked with `liblttng-ust`,
1963 the LTTng user space tracer, at run time.
1966 .User application linked with `liblttng-ust` and containing a tracepoint provider.
1967 image::ust-app.png[]
1969 NOTE: If you need quick, man:printf(3)-like instrumentation, you can
1970 skip creating and using a tracepoint provider and use
1971 <<tracef,`tracef()`>> or <<tracelog,`tracelog()`>> instead.
1975 ===== Create a tracepoint provider header file template
1977 A _tracepoint provider header file_ contains the tracepoint
1978 definitions of a tracepoint provider.
1980 To create a tracepoint provider header file:
1982 . Start from this template:
1986 .Tracepoint provider header file template (`.h` file extension).
1988 #undef TRACEPOINT_PROVIDER
1989 #define TRACEPOINT_PROVIDER provider_name
1991 #undef TRACEPOINT_INCLUDE
1992 #define TRACEPOINT_INCLUDE "./tp.h"
1994 #if !defined(_TP_H) || defined(TRACEPOINT_HEADER_MULTI_READ)
1997 #include <lttng/tracepoint.h>
2000 * Use TRACEPOINT_EVENT(), TRACEPOINT_EVENT_CLASS(),
2001 * TRACEPOINT_EVENT_INSTANCE(), and TRACEPOINT_LOGLEVEL() here.
2006 #include <lttng/tracepoint-event.h>
2012 * `provider_name` with the name of your tracepoint provider.
2013 * `"tp.h"` with the name of your tracepoint provider header file.
2015 . Below the `#include <lttng/tracepoint.h>` line, put your
2016 <<defining-tracepoints,tracepoint definitions>>.
2018 Your tracepoint provider name must be unique amongst all the possible
2019 tracepoint provider names used on the same target system. We
2020 suggest to include the name of your project or company in the name,
2021 for example, `org_lttng_my_project_tpp`.
2023 TIP: [[lttng-gen-tp]]You can use the man:lttng-gen-tp(1) tool to create
2024 this boilerplate for you. When using cmd:lttng-gen-tp, all you need to
2025 write are the <<defining-tracepoints,tracepoint definitions>>.
2028 [[defining-tracepoints]]
2029 ===== Create a tracepoint definition
2031 A _tracepoint definition_ defines, for a given tracepoint:
2033 * Its **input arguments**. They are the macro parameters that the
2034 `tracepoint()` macro accepts for this particular tracepoint
2035 in the user application's source code.
2036 * Its **output event fields**. They are the sources of event fields
2037 that form the payload of any event that the execution of the
2038 `tracepoint()` macro emits for this particular tracepoint.
2040 You can create a tracepoint definition by using the
2041 `TRACEPOINT_EVENT()` macro below the `#include <lttng/tracepoint.h>`
2043 <<tpp-header,tracepoint provider header file template>>.
2045 The syntax of the `TRACEPOINT_EVENT()` macro is:
2048 .`TRACEPOINT_EVENT()` macro syntax.
2051 /* Tracepoint provider name */
2054 /* Tracepoint name */
2057 /* Input arguments */
2062 /* Output event fields */
2071 * `provider_name` with your tracepoint provider name.
2072 * `tracepoint_name` with your tracepoint name.
2073 * `arguments` with the <<tpp-def-input-args,input arguments>>.
2074 * `fields` with the <<tpp-def-output-fields,output event field>>
2077 This tracepoint emits events named `provider_name:tracepoint_name`.
2080 .Event name's length limitation
2082 The concatenation of the tracepoint provider name and the
2083 tracepoint name must not exceed **254 characters**. If it does, the
2084 instrumented application compiles and runs, but LTTng throws multiple
2085 warnings and you could experience serious issues.
2088 [[tpp-def-input-args]]The syntax of the `TP_ARGS()` macro is:
2091 .`TP_ARGS()` macro syntax.
2100 * `type` with the C type of the argument.
2101 * `arg_name` with the argument name.
2103 You can repeat `type` and `arg_name` up to 10 times to have
2104 more than one argument.
2106 .`TP_ARGS()` usage with three arguments.
2118 The `TP_ARGS()` and `TP_ARGS(void)` forms are valid to create a
2119 tracepoint definition with no input arguments.
2121 [[tpp-def-output-fields]]The `TP_FIELDS()` macro contains a list of
2122 `ctf_*()` macros. Each `ctf_*()` macro defines one event field. See
2123 man:lttng-ust(3) for a complete description of the available `ctf_*()`
2124 macros. A `ctf_*()` macro specifies the type, size, and byte order of
2127 Each `ctf_*()` macro takes an _argument expression_ parameter. This is a
2128 C expression that the tracer evalutes at the `tracepoint()` macro site
2129 in the application's source code. This expression provides a field's
2130 source of data. The argument expression can include input argument names
2131 listed in the `TP_ARGS()` macro.
2133 Each `ctf_*()` macro also takes a _field name_ parameter. Field names
2134 must be unique within a given tracepoint definition.
2136 Here's a complete tracepoint definition example:
2138 .Tracepoint definition.
2140 The following tracepoint definition defines a tracepoint which takes
2141 three input arguments and has four output event fields.
2145 #include "my-custom-structure.h"
2151 const struct my_custom_structure*, my_custom_structure,
2156 ctf_string(query_field, query)
2157 ctf_float(double, ratio_field, ratio)
2158 ctf_integer(int, recv_size, my_custom_structure->recv_size)
2159 ctf_integer(int, send_size, my_custom_structure->send_size)
2164 You can refer to this tracepoint definition with the `tracepoint()`
2165 macro in your application's source code like this:
2169 tracepoint(my_provider, my_tracepoint,
2170 my_structure, some_ratio, the_query);
2174 NOTE: The LTTng tracer only evaluates tracepoint arguments at run time
2175 if they satisfy an enabled <<event,event rule>>.
2178 [[using-tracepoint-classes]]
2179 ===== Use a tracepoint class
2181 A _tracepoint class_ is a class of tracepoints which share the same
2182 output event field definitions. A _tracepoint instance_ is one
2183 instance of such a defined tracepoint class, with its own tracepoint
2186 The <<defining-tracepoints,`TRACEPOINT_EVENT()` macro>> is actually a
2187 shorthand which defines both a tracepoint class and a tracepoint
2188 instance at the same time.
2190 When you build a tracepoint provider package, the C or $$C++$$ compiler
2191 creates one serialization function for each **tracepoint class**. A
2192 serialization function is responsible for serializing the event fields
2193 of a tracepoint to a sub-buffer when tracing.
2195 For various performance reasons, when your situation requires multiple
2196 tracepoint definitions with different names, but with the same event
2197 fields, we recommend that you manually create a tracepoint class
2198 and instantiate as many tracepoint instances as needed. One positive
2199 effect of such a design, amongst other advantages, is that all
2200 tracepoint instances of the same tracepoint class reuse the same
2201 serialization function, thus reducing
2202 https://en.wikipedia.org/wiki/Cache_pollution[cache pollution].
2204 .Use a tracepoint class and tracepoint instances.
2206 Consider the following three tracepoint definitions:
2218 ctf_integer(int, userid, userid)
2219 ctf_integer(size_t, len, len)
2231 ctf_integer(int, userid, userid)
2232 ctf_integer(size_t, len, len)
2244 ctf_integer(int, userid, userid)
2245 ctf_integer(size_t, len, len)
2250 In this case, we create three tracepoint classes, with one implicit
2251 tracepoint instance for each of them: `get_account`, `get_settings`, and
2252 `get_transaction`. However, they all share the same event field names
2253 and types. Hence three identical, yet independent serialization
2254 functions are created when you build the tracepoint provider package.
2256 A better design choice is to define a single tracepoint class and three
2257 tracepoint instances:
2261 /* The tracepoint class */
2262 TRACEPOINT_EVENT_CLASS(
2263 /* Tracepoint provider name */
2266 /* Tracepoint class name */
2269 /* Input arguments */
2275 /* Output event fields */
2277 ctf_integer(int, userid, userid)
2278 ctf_integer(size_t, len, len)
2282 /* The tracepoint instances */
2283 TRACEPOINT_EVENT_INSTANCE(
2284 /* Tracepoint provider name */
2287 /* Tracepoint class name */
2290 /* Tracepoint name */
2293 /* Input arguments */
2299 TRACEPOINT_EVENT_INSTANCE(
2308 TRACEPOINT_EVENT_INSTANCE(
2321 [[assigning-log-levels]]
2322 ===== Assign a log level to a tracepoint definition
2324 You can assign an optional _log level_ to a
2325 <<defining-tracepoints,tracepoint definition>>.
2327 Assigning different levels of severity to tracepoint definitions can
2328 be useful: when you <<enabling-disabling-events,create an event rule>>,
2329 you can target tracepoints having a log level as severe as a specific
2332 The concept of LTTng-UST log levels is similar to the levels found
2333 in typical logging frameworks:
2335 * In a logging framework, the log level is given by the function
2336 or method name you use at the log statement site: `debug()`,
2337 `info()`, `warn()`, `error()`, and so on.
2338 * In LTTng-UST, you statically assign the log level to a tracepoint
2339 definition; any `tracepoint()` macro invocation which refers to
2340 this definition has this log level.
2342 You can assign a log level to a tracepoint definition with the
2343 `TRACEPOINT_LOGLEVEL()` macro. You must use this macro _after_ the
2344 <<defining-tracepoints,`TRACEPOINT_EVENT()`>> or
2345 <<using-tracepoint-classes,`TRACEPOINT_INSTANCE()`>> macro for a given
2348 The syntax of the `TRACEPOINT_LOGLEVEL()` macro is:
2351 .`TRACEPOINT_LOGLEVEL()` macro syntax.
2353 TRACEPOINT_LOGLEVEL(provider_name, tracepoint_name, log_level)
2358 * `provider_name` with the tracepoint provider name.
2359 * `tracepoint_name` with the tracepoint name.
2360 * `log_level` with the log level to assign to the tracepoint
2361 definition named `tracepoint_name` in the `provider_name`
2362 tracepoint provider.
2364 See man:lttng-ust(3) for a list of available log level names.
2366 .Assign the `TRACE_DEBUG_UNIT` log level to a tracepoint definition.
2370 /* Tracepoint definition */
2379 ctf_integer(int, userid, userid)
2380 ctf_integer(size_t, len, len)
2384 /* Log level assignment */
2385 TRACEPOINT_LOGLEVEL(my_app, get_transaction, TRACE_DEBUG_UNIT)
2391 ===== Create a tracepoint provider package source file
2393 A _tracepoint provider package source file_ is a C source file which
2394 includes a <<tpp-header,tracepoint provider header file>> to expand its
2395 macros into event serialization and other functions.
2397 You can always use the following tracepoint provider package source
2401 .Tracepoint provider package source file template.
2403 #define TRACEPOINT_CREATE_PROBES
2408 Replace `tp.h` with the name of your <<tpp-header,tracepoint provider
2409 header file>> name. You may also include more than one tracepoint
2410 provider header file here to create a tracepoint provider package
2411 holding more than one tracepoint providers.
2414 [[probing-the-application-source-code]]
2415 ==== Add tracepoints to an application's source code
2417 Once you <<tpp-header,create a tracepoint provider header file>>, you
2418 can use the `tracepoint()` macro in your application's
2419 source code to insert the tracepoints that this header
2420 <<defining-tracepoints,defines>>.
2422 The `tracepoint()` macro takes at least two parameters: the tracepoint
2423 provider name and the tracepoint name. The corresponding tracepoint
2424 definition defines the other parameters.
2426 .`tracepoint()` usage.
2428 The following <<defining-tracepoints,tracepoint definition>> defines a
2429 tracepoint which takes two input arguments and has two output event
2433 .Tracepoint provider header file.
2435 #include "my-custom-structure.h"
2442 const char*, cmd_name
2445 ctf_string(cmd_name, cmd_name)
2446 ctf_integer(int, number_of_args, argc)
2451 You can refer to this tracepoint definition with the `tracepoint()`
2452 macro in your application's source code like this:
2455 .Application's source file.
2459 int main(int argc, char* argv[])
2461 tracepoint(my_provider, my_tracepoint, argc, argv[0]);
2467 Note how the application's source code includes
2468 the tracepoint provider header file containing the tracepoint
2469 definitions to use, path:{tp.h}.
2472 .`tracepoint()` usage with a complex tracepoint definition.
2474 Consider this complex tracepoint definition, where multiple event
2475 fields refer to the same input arguments in their argument expression
2479 .Tracepoint provider header file.
2481 /* For `struct stat` */
2482 #include <sys/types.h>
2483 #include <sys/stat.h>
2495 ctf_integer(int, my_constant_field, 23 + 17)
2496 ctf_integer(int, my_int_arg_field, my_int_arg)
2497 ctf_integer(int, my_int_arg_field2, my_int_arg * my_int_arg)
2498 ctf_integer(int, sum4_field, my_str_arg[0] + my_str_arg[1] +
2499 my_str_arg[2] + my_str_arg[3])
2500 ctf_string(my_str_arg_field, my_str_arg)
2501 ctf_integer_hex(off_t, size_field, st->st_size)
2502 ctf_float(double, size_dbl_field, (double) st->st_size)
2503 ctf_sequence_text(char, half_my_str_arg_field, my_str_arg,
2504 size_t, strlen(my_str_arg) / 2)
2509 You can refer to this tracepoint definition with the `tracepoint()`
2510 macro in your application's source code like this:
2513 .Application's source file.
2515 #define TRACEPOINT_DEFINE
2522 stat("/etc/fstab", &s);
2523 tracepoint(my_provider, my_tracepoint, 23, "Hello, World!", &s);
2529 If you look at the event record that LTTng writes when tracing this
2530 program, assuming the file size of path:{/etc/fstab} is 301{nbsp}bytes,
2531 it should look like this:
2533 .Event record fields
2535 |Field's name |Field's value
2536 |`my_constant_field` |40
2537 |`my_int_arg_field` |23
2538 |`my_int_arg_field2` |529
2540 |`my_str_arg_field` |`Hello, World!`
2541 |`size_field` |0x12d
2542 |`size_dbl_field` |301.0
2543 |`half_my_str_arg_field` |`Hello,`
2547 Sometimes, the arguments you pass to `tracepoint()` are expensive to
2548 compute--they use the call stack, for example. To avoid this
2549 computation when the tracepoint is disabled, you can use the
2550 `tracepoint_enabled()` and `do_tracepoint()` macros.
2552 The syntax of the `tracepoint_enabled()` and `do_tracepoint()` macros
2556 .`tracepoint_enabled()` and `do_tracepoint()` macros syntax.
2558 tracepoint_enabled(provider_name, tracepoint_name)
2559 do_tracepoint(provider_name, tracepoint_name, ...)
2564 * `provider_name` with the tracepoint provider name.
2565 * `tracepoint_name` with the tracepoint name.
2567 `tracepoint_enabled()` returns a non-zero value if the tracepoint named
2568 `tracepoint_name` from the provider named `provider_name` is enabled
2571 `do_tracepoint()` is like `tracepoint()`, except that it doesn't check
2572 if the tracepoint is enabled. Using `tracepoint()` with
2573 `tracepoint_enabled()` is dangerous since `tracepoint()` also contains
2574 the `tracepoint_enabled()` check, thus a race condition is
2575 possible in this situation:
2578 .Possible race condition when using `tracepoint_enabled()` with `tracepoint()`.
2580 if (tracepoint_enabled(my_provider, my_tracepoint)) {
2581 stuff = prepare_stuff();
2584 tracepoint(my_provider, my_tracepoint, stuff);
2587 If the tracepoint is enabled after the condition, then `stuff` is not
2588 prepared: the emitted event will either contain wrong data, or the whole
2589 application could crash (segmentation fault, for example).
2591 NOTE: Neither `tracepoint_enabled()` nor `do_tracepoint()` have an
2592 `STAP_PROBEV()` call. If you need it, you must emit
2596 [[building-tracepoint-providers-and-user-application]]
2597 ==== Build and link a tracepoint provider package and an application
2599 Once you have one or more <<tpp-header,tracepoint provider header
2600 files>> and a <<tpp-source,tracepoint provider package source file>>,
2601 you can create the tracepoint provider package by compiling its source
2602 file. From here, multiple build and run scenarios are possible. The
2603 following table shows common application and library configurations
2604 along with the required command lines to achieve them.
2606 In the following diagrams, we use the following file names:
2609 Executable application.
2612 Application's object file.
2615 Tracepoint provider package object file.
2618 Tracepoint provider package archive file.
2621 Tracepoint provider package shared object file.
2624 User library object file.
2627 User library shared object file.
2629 We use the following symbols in the diagrams of table below:
2632 .Symbols used in the build scenario diagrams.
2633 image::ust-sit-symbols.png[]
2635 We assume that path:{.} is part of the env:LD_LIBRARY_PATH environment
2636 variable in the following instructions.
2638 [role="growable ust-scenarios",cols="asciidoc,asciidoc"]
2639 .Common tracepoint provider package scenarios.
2641 |Scenario |Instructions
2644 The instrumented application is statically linked with
2645 the tracepoint provider package object.
2647 image::ust-sit+app-linked-with-tp-o+app-instrumented.png[]
2650 include::../common/ust-sit-step-tp-o.txt[]
2652 To build the instrumented application:
2654 . In path:{app.c}, before including path:{tpp.h}, add the following line:
2659 #define TRACEPOINT_DEFINE
2663 . Compile the application source file:
2672 . Build the application:
2677 $ gcc -o app app.o tpp.o -llttng-ust -ldl
2681 To run the instrumented application:
2683 * Start the application:
2693 The instrumented application is statically linked with the
2694 tracepoint provider package archive file.
2696 image::ust-sit+app-linked-with-tp-a+app-instrumented.png[]
2699 To create the tracepoint provider package archive file:
2701 . Compile the <<tpp-source,tracepoint provider package source file>>:
2710 . Create the tracepoint provider package archive file:
2715 $ ar rcs tpp.a tpp.o
2719 To build the instrumented application:
2721 . In path:{app.c}, before including path:{tpp.h}, add the following line:
2726 #define TRACEPOINT_DEFINE
2730 . Compile the application source file:
2739 . Build the application:
2744 $ gcc -o app app.o tpp.a -llttng-ust -ldl
2748 To run the instrumented application:
2750 * Start the application:
2760 The instrumented application is linked with the tracepoint provider
2761 package shared object.
2763 image::ust-sit+app-linked-with-tp-so+app-instrumented.png[]
2766 include::../common/ust-sit-step-tp-so.txt[]
2768 To build the instrumented application:
2770 . In path:{app.c}, before including path:{tpp.h}, add the following line:
2775 #define TRACEPOINT_DEFINE
2779 . Compile the application source file:
2788 . Build the application:
2793 $ gcc -o app app.o -ldl -L. -ltpp
2797 To run the instrumented application:
2799 * Start the application:
2809 The tracepoint provider package shared object is preloaded before the
2810 instrumented application starts.
2812 image::ust-sit+tp-so-preloaded+app-instrumented.png[]
2815 include::../common/ust-sit-step-tp-so.txt[]
2817 To build the instrumented application:
2819 . In path:{app.c}, before including path:{tpp.h}, add the
2825 #define TRACEPOINT_DEFINE
2826 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
2830 . Compile the application source file:
2839 . Build the application:
2844 $ gcc -o app app.o -ldl
2848 To run the instrumented application with tracing support:
2850 * Preload the tracepoint provider package shared object and
2851 start the application:
2856 $ LD_PRELOAD=./libtpp.so ./app
2860 To run the instrumented application without tracing support:
2862 * Start the application:
2872 The instrumented application dynamically loads the tracepoint provider
2873 package shared object.
2875 See the <<dlclose-warning,warning about `dlclose()`>>.
2877 image::ust-sit+app-dlopens-tp-so+app-instrumented.png[]
2880 include::../common/ust-sit-step-tp-so.txt[]
2882 To build the instrumented application:
2884 . In path:{app.c}, before including path:{tpp.h}, add the
2890 #define TRACEPOINT_DEFINE
2891 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
2895 . Compile the application source file:
2904 . Build the application:
2909 $ gcc -o app app.o -ldl
2913 To run the instrumented application:
2915 * Start the application:
2925 The application is linked with the instrumented user library.
2927 The instrumented user library is statically linked with the tracepoint
2928 provider package object file.
2930 image::ust-sit+app-linked-with-lib+lib-linked-with-tp-o+lib-instrumented.png[]
2933 include::../common/ust-sit-step-tp-o-fpic.txt[]
2935 To build the instrumented user library:
2937 . In path:{emon.c}, before including path:{tpp.h}, add the
2943 #define TRACEPOINT_DEFINE
2947 . Compile the user library source file:
2952 $ gcc -I. -fpic -c emon.c
2956 . Build the user library shared object:
2961 $ gcc -shared -o libemon.so emon.o tpp.o -llttng-ust -ldl
2965 To build the application:
2967 . Compile the application source file:
2976 . Build the application:
2981 $ gcc -o app app.o -L. -lemon
2985 To run the application:
2987 * Start the application:
2997 The application is linked with the instrumented user library.
2999 The instrumented user library is linked with the tracepoint provider
3000 package shared object.
3002 image::ust-sit+app-linked-with-lib+lib-linked-with-tp-so+lib-instrumented.png[]
3005 include::../common/ust-sit-step-tp-so.txt[]
3007 To build the instrumented user library:
3009 . In path:{emon.c}, before including path:{tpp.h}, add the
3015 #define TRACEPOINT_DEFINE
3019 . Compile the user library source file:
3024 $ gcc -I. -fpic -c emon.c
3028 . Build the user library shared object:
3033 $ gcc -shared -o libemon.so emon.o -ldl -L. -ltpp
3037 To build the application:
3039 . Compile the application source file:
3048 . Build the application:
3053 $ gcc -o app app.o -L. -lemon
3057 To run the application:
3059 * Start the application:
3069 The tracepoint provider package shared object is preloaded before the
3072 The application is linked with the instrumented user library.
3074 image::ust-sit+tp-so-preloaded+app-linked-with-lib+lib-instrumented.png[]
3077 include::../common/ust-sit-step-tp-so.txt[]
3079 To build the instrumented user library:
3081 . In path:{emon.c}, before including path:{tpp.h}, add the
3087 #define TRACEPOINT_DEFINE
3088 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3092 . Compile the user library source file:
3097 $ gcc -I. -fpic -c emon.c
3101 . Build the user library shared object:
3106 $ gcc -shared -o libemon.so emon.o -ldl
3110 To build the application:
3112 . Compile the application source file:
3121 . Build the application:
3126 $ gcc -o app app.o -L. -lemon
3130 To run the application with tracing support:
3132 * Preload the tracepoint provider package shared object and
3133 start the application:
3138 $ LD_PRELOAD=./libtpp.so ./app
3142 To run the application without tracing support:
3144 * Start the application:
3154 The application is linked with the instrumented user library.
3156 The instrumented user library dynamically loads the tracepoint provider
3157 package shared object.
3159 See the <<dlclose-warning,warning about `dlclose()`>>.
3161 image::ust-sit+app-linked-with-lib+lib-dlopens-tp-so+lib-instrumented.png[]
3164 include::../common/ust-sit-step-tp-so.txt[]
3166 To build the instrumented user library:
3168 . In path:{emon.c}, before including path:{tpp.h}, add the
3174 #define TRACEPOINT_DEFINE
3175 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3179 . Compile the user library source file:
3184 $ gcc -I. -fpic -c emon.c
3188 . Build the user library shared object:
3193 $ gcc -shared -o libemon.so emon.o -ldl
3197 To build the application:
3199 . Compile the application source file:
3208 . Build the application:
3213 $ gcc -o app app.o -L. -lemon
3217 To run the application:
3219 * Start the application:
3229 The application dynamically loads the instrumented user library.
3231 The instrumented user library is linked with the tracepoint provider
3232 package shared object.
3234 See the <<dlclose-warning,warning about `dlclose()`>>.
3236 image::ust-sit+app-dlopens-lib+lib-linked-with-tp-so+lib-instrumented.png[]
3239 include::../common/ust-sit-step-tp-so.txt[]
3241 To build the instrumented user library:
3243 . In path:{emon.c}, before including path:{tpp.h}, add the
3249 #define TRACEPOINT_DEFINE
3253 . Compile the user library source file:
3258 $ gcc -I. -fpic -c emon.c
3262 . Build the user library shared object:
3267 $ gcc -shared -o libemon.so emon.o -ldl -L. -ltpp
3271 To build the application:
3273 . Compile the application source file:
3282 . Build the application:
3287 $ gcc -o app app.o -ldl -L. -lemon
3291 To run the application:
3293 * Start the application:
3303 The application dynamically loads the instrumented user library.
3305 The instrumented user library dynamically loads the tracepoint provider
3306 package shared object.
3308 See the <<dlclose-warning,warning about `dlclose()`>>.
3310 image::ust-sit+app-dlopens-lib+lib-dlopens-tp-so+lib-instrumented.png[]
3313 include::../common/ust-sit-step-tp-so.txt[]
3315 To build the instrumented user library:
3317 . In path:{emon.c}, before including path:{tpp.h}, add the
3323 #define TRACEPOINT_DEFINE
3324 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3328 . Compile the user library source file:
3333 $ gcc -I. -fpic -c emon.c
3337 . Build the user library shared object:
3342 $ gcc -shared -o libemon.so emon.o -ldl
3346 To build the application:
3348 . Compile the application source file:
3357 . Build the application:
3362 $ gcc -o app app.o -ldl -L. -lemon
3366 To run the application:
3368 * Start the application:
3378 The tracepoint provider package shared object is preloaded before the
3381 The application dynamically loads the instrumented user library.
3383 image::ust-sit+tp-so-preloaded+app-dlopens-lib+lib-instrumented.png[]
3386 include::../common/ust-sit-step-tp-so.txt[]
3388 To build the instrumented user library:
3390 . In path:{emon.c}, before including path:{tpp.h}, add the
3396 #define TRACEPOINT_DEFINE
3397 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3401 . Compile the user library source file:
3406 $ gcc -I. -fpic -c emon.c
3410 . Build the user library shared object:
3415 $ gcc -shared -o libemon.so emon.o -ldl
3419 To build the application:
3421 . Compile the application source file:
3430 . Build the application:
3435 $ gcc -o app app.o -L. -lemon
3439 To run the application with tracing support:
3441 * Preload the tracepoint provider package shared object and
3442 start the application:
3447 $ LD_PRELOAD=./libtpp.so ./app
3451 To run the application without tracing support:
3453 * Start the application:
3463 The application is statically linked with the tracepoint provider
3464 package object file.
3466 The application is linked with the instrumented user library.
3468 image::ust-sit+app-linked-with-tp-o+app-linked-with-lib+lib-instrumented.png[]
3471 include::../common/ust-sit-step-tp-o.txt[]
3473 To build the instrumented user library:
3475 . In path:{emon.c}, before including path:{tpp.h}, add the
3481 #define TRACEPOINT_DEFINE
3485 . Compile the user library source file:
3490 $ gcc -I. -fpic -c emon.c
3494 . Build the user library shared object:
3499 $ gcc -shared -o libemon.so emon.o
3503 To build the application:
3505 . Compile the application source file:
3514 . Build the application:
3519 $ gcc -o app app.o tpp.o -llttng-ust -ldl -L. -lemon
3523 To run the instrumented application:
3525 * Start the application:
3535 The application is statically linked with the tracepoint provider
3536 package object file.
3538 The application dynamically loads the instrumented user library.
3540 image::ust-sit+app-linked-with-tp-o+app-dlopens-lib+lib-instrumented.png[]
3543 include::../common/ust-sit-step-tp-o.txt[]
3545 To build the application:
3547 . In path:{app.c}, before including path:{tpp.h}, add the following line:
3552 #define TRACEPOINT_DEFINE
3556 . Compile the application source file:
3565 . Build the application:
3570 $ gcc -Wl,--export-dynamic -o app app.o tpp.o \
3575 The `--export-dynamic` option passed to the linker is necessary for the
3576 dynamically loaded library to ``see'' the tracepoint symbols defined in
3579 To build the instrumented user library:
3581 . Compile the user library source file:
3586 $ gcc -I. -fpic -c emon.c
3590 . Build the user library shared object:
3595 $ gcc -shared -o libemon.so emon.o
3599 To run the application:
3601 * Start the application:
3613 .Do not use man:dlclose(3) on a tracepoint provider package
3615 Never use man:dlclose(3) on any shared object which:
3617 * Is linked with, statically or dynamically, a tracepoint provider
3619 * Calls man:dlopen(3) itself to dynamically open a tracepoint provider
3620 package shared object.
3622 This is currently considered **unsafe** due to a lack of reference
3623 counting from LTTng-UST to the shared object.
3625 A known workaround (available since glibc 2.2) is to use the
3626 `RTLD_NODELETE` flag when calling man:dlopen(3) initially. This has the
3627 effect of not unloading the loaded shared object, even if man:dlclose(3)
3630 You can also preload the tracepoint provider package shared object with
3631 the env:LD_PRELOAD environment variable to overcome this limitation.
3635 [[using-lttng-ust-with-daemons]]
3636 ===== Use noch:{LTTng-UST} with daemons
3638 If your instrumented application calls man:fork(2), man:clone(2),
3639 or BSD's man:rfork(2), without a following man:exec(3)-family
3640 system call, you must preload the path:{liblttng-ust-fork.so} shared
3641 object when you start the application.
3645 $ LD_PRELOAD=liblttng-ust-fork.so ./my-app
3648 If your tracepoint provider package is
3649 a shared library which you also preload, you must put both
3650 shared objects in env:LD_PRELOAD:
3654 $ LD_PRELOAD=liblttng-ust-fork.so:/path/to/tp.so ./my-app
3660 ===== Use noch:{LTTng-UST} with applications which close file descriptors that don't belong to them
3662 If your instrumented application closes one or more file descriptors
3663 which it did not open itself, you must preload the
3664 path:{liblttng-ust-fd.so} shared object when you start the application:
3668 $ LD_PRELOAD=liblttng-ust-fd.so ./my-app
3671 Typical use cases include closing all the file descriptors after
3672 man:fork(2) or man:rfork(2) and buggy applications doing
3676 [[lttng-ust-pkg-config]]
3677 ===== Use noch:{pkg-config}
3679 On some distributions, LTTng-UST ships with a
3680 https://www.freedesktop.org/wiki/Software/pkg-config/[pkg-config]
3681 metadata file. If this is your case, then you can use cmd:pkg-config to
3682 build an application on the command line:
3686 $ gcc -o my-app my-app.o tp.o $(pkg-config --cflags --libs lttng-ust)
3690 [[instrumenting-32-bit-app-on-64-bit-system]]
3691 ===== [[advanced-instrumenting-techniques]]Build a 32-bit instrumented application for a 64-bit target system
3693 In order to trace a 32-bit application running on a 64-bit system,
3694 LTTng must use a dedicated 32-bit
3695 <<lttng-consumerd,consumer daemon>>.
3697 The following steps show how to build and install a 32-bit consumer
3698 daemon, which is _not_ part of the default 64-bit LTTng build, how to
3699 build and install the 32-bit LTTng-UST libraries, and how to build and
3700 link an instrumented 32-bit application in that context.
3702 To build a 32-bit instrumented application for a 64-bit target system,
3703 assuming you have a fresh target system with no installed Userspace RCU
3706 . Download, build, and install a 32-bit version of Userspace RCU:
3711 $ cd $(mktemp -d) &&
3712 wget http://lttng.org/files/urcu/userspace-rcu-latest-0.9.tar.bz2 &&
3713 tar -xf userspace-rcu-latest-0.9.tar.bz2 &&
3714 cd userspace-rcu-0.9.* &&
3715 ./configure --libdir=/usr/local/lib32 CFLAGS=-m32 &&
3717 sudo make install &&
3722 . Using your distribution's package manager, or from source, install
3723 the following 32-bit versions of the following dependencies of
3724 LTTng-tools and LTTng-UST:
3727 * https://sourceforge.net/projects/libuuid/[libuuid]
3728 * http://directory.fsf.org/wiki/Popt[popt]
3729 * http://www.xmlsoft.org/[libxml2]
3732 . Download, build, and install a 32-bit version of the latest
3733 LTTng-UST{nbsp}{revision}:
3738 $ cd $(mktemp -d) &&
3739 wget http://lttng.org/files/lttng-ust/lttng-ust-latest-2.10.tar.bz2 &&
3740 tar -xf lttng-ust-latest-2.10.tar.bz2 &&
3741 cd lttng-ust-2.10.* &&
3742 ./configure --libdir=/usr/local/lib32 \
3743 CFLAGS=-m32 CXXFLAGS=-m32 \
3744 LDFLAGS='-L/usr/local/lib32 -L/usr/lib32' &&
3746 sudo make install &&
3753 Depending on your distribution,
3754 32-bit libraries could be installed at a different location than
3755 `/usr/lib32`. For example, Debian is known to install
3756 some 32-bit libraries in `/usr/lib/i386-linux-gnu`.
3758 In this case, make sure to set `LDFLAGS` to all the
3759 relevant 32-bit library paths, for example:
3763 $ LDFLAGS='-L/usr/lib/i386-linux-gnu -L/usr/lib32'
3767 . Download the latest LTTng-tools{nbsp}{revision}, build, and install
3768 the 32-bit consumer daemon:
3773 $ cd $(mktemp -d) &&
3774 wget http://lttng.org/files/lttng-tools/lttng-tools-latest-2.10.tar.bz2 &&
3775 tar -xf lttng-tools-latest-2.10.tar.bz2 &&
3776 cd lttng-tools-2.10.* &&
3777 ./configure --libdir=/usr/local/lib32 CFLAGS=-m32 CXXFLAGS=-m32 \
3778 LDFLAGS='-L/usr/local/lib32 -L/usr/lib32' \
3779 --disable-bin-lttng --disable-bin-lttng-crash \
3780 --disable-bin-lttng-relayd --disable-bin-lttng-sessiond &&
3782 cd src/bin/lttng-consumerd &&
3783 sudo make install &&
3788 . From your distribution or from source,
3789 <<installing-lttng,install>> the 64-bit versions of
3790 LTTng-UST and Userspace RCU.
3791 . Download, build, and install the 64-bit version of the
3792 latest LTTng-tools{nbsp}{revision}:
3797 $ cd $(mktemp -d) &&
3798 wget http://lttng.org/files/lttng-tools/lttng-tools-latest-2.10.tar.bz2 &&
3799 tar -xf lttng-tools-latest-2.10.tar.bz2 &&
3800 cd lttng-tools-2.10.* &&
3801 ./configure --with-consumerd32-libdir=/usr/local/lib32 \
3802 --with-consumerd32-bin=/usr/local/lib32/lttng/libexec/lttng-consumerd &&
3804 sudo make install &&
3809 . Pass the following options to man:gcc(1), man:g++(1), or man:clang(1)
3810 when linking your 32-bit application:
3813 -m32 -L/usr/lib32 -L/usr/local/lib32 \
3814 -Wl,-rpath,/usr/lib32,-rpath,/usr/local/lib32
3817 For example, let's rebuild the quick start example in
3818 <<tracing-your-own-user-application,Trace a user application>> as an
3819 instrumented 32-bit application:
3824 $ gcc -m32 -c -I. hello-tp.c
3825 $ gcc -m32 -c hello.c
3826 $ gcc -m32 -o hello hello.o hello-tp.o \
3827 -L/usr/lib32 -L/usr/local/lib32 \
3828 -Wl,-rpath,/usr/lib32,-rpath,/usr/local/lib32 \
3833 No special action is required to execute the 32-bit application and
3834 to trace it: use the command-line man:lttng(1) tool as usual.
3841 man:tracef(3) is a small LTTng-UST API designed for quick,
3842 man:printf(3)-like instrumentation without the burden of
3843 <<tracepoint-provider,creating>> and
3844 <<building-tracepoint-providers-and-user-application,building>>
3845 a tracepoint provider package.
3847 To use `tracef()` in your application:
3849 . In the C or C++ source files where you need to use `tracef()`,
3850 include `<lttng/tracef.h>`:
3855 #include <lttng/tracef.h>
3859 . In the application's source code, use `tracef()` like you would use
3867 tracef("my message: %d (%s)", my_integer, my_string);
3873 . Link your application with `liblttng-ust`:
3878 $ gcc -o app app.c -llttng-ust
3882 To trace the events that `tracef()` calls emit:
3884 * <<enabling-disabling-events,Create an event rule>> which matches the
3885 `lttng_ust_tracef:*` event name:
3890 $ lttng enable-event --userspace 'lttng_ust_tracef:*'
3895 .Limitations of `tracef()`
3897 The `tracef()` utility function was developed to make user space tracing
3898 super simple, albeit with notable disadvantages compared to
3899 <<defining-tracepoints,user-defined tracepoints>>:
3901 * All the emitted events have the same tracepoint provider and
3902 tracepoint names, respectively `lttng_ust_tracef` and `event`.
3903 * There is no static type checking.
3904 * The only event record field you actually get, named `msg`, is a string
3905 potentially containing the values you passed to `tracef()`
3906 using your own format string. This also means that you cannot filter
3907 events with a custom expression at run time because there are no
3909 * Since `tracef()` uses the C standard library's man:vasprintf(3)
3910 function behind the scenes to format the strings at run time, its
3911 expected performance is lower than with user-defined tracepoints,
3912 which do not require a conversion to a string.
3914 Taking this into consideration, `tracef()` is useful for some quick
3915 prototyping and debugging, but you should not consider it for any
3916 permanent and serious applicative instrumentation.
3922 ==== Use `tracelog()`
3924 The man:tracelog(3) API is very similar to <<tracef,`tracef()`>>, with
3925 the difference that it accepts an additional log level parameter.
3927 The goal of `tracelog()` is to ease the migration from logging to
3930 To use `tracelog()` in your application:
3932 . In the C or C++ source files where you need to use `tracelog()`,
3933 include `<lttng/tracelog.h>`:
3938 #include <lttng/tracelog.h>
3942 . In the application's source code, use `tracelog()` like you would use
3943 man:printf(3), except for the first parameter which is the log
3951 tracelog(TRACE_WARNING, "my message: %d (%s)",
3952 my_integer, my_string);
3958 See man:lttng-ust(3) for a list of available log level names.
3960 . Link your application with `liblttng-ust`:
3965 $ gcc -o app app.c -llttng-ust
3969 To trace the events that `tracelog()` calls emit with a log level
3970 _as severe as_ a specific log level:
3972 * <<enabling-disabling-events,Create an event rule>> which matches the
3973 `lttng_ust_tracelog:*` event name and a minimum level
3979 $ lttng enable-event --userspace 'lttng_ust_tracelog:*'
3980 --loglevel=TRACE_WARNING
3984 To trace the events that `tracelog()` calls emit with a
3985 _specific log level_:
3987 * Create an event rule which matches the `lttng_ust_tracelog:*`
3988 event name and a specific log level:
3993 $ lttng enable-event --userspace 'lttng_ust_tracelog:*'
3994 --loglevel-only=TRACE_INFO
3999 [[prebuilt-ust-helpers]]
4000 === Prebuilt user space tracing helpers
4002 The LTTng-UST package provides a few helpers in the form or preloadable
4003 shared objects which automatically instrument system functions and
4006 The helper shared objects are normally found in dir:{/usr/lib}. If you
4007 built LTTng-UST <<building-from-source,from source>>, they are probably
4008 located in dir:{/usr/local/lib}.
4010 The installed user space tracing helpers in LTTng-UST{nbsp}{revision}
4013 path:{liblttng-ust-libc-wrapper.so}::
4014 path:{liblttng-ust-pthread-wrapper.so}::
4015 <<liblttng-ust-libc-pthread-wrapper,C{nbsp}standard library
4016 memory and POSIX threads function tracing>>.
4018 path:{liblttng-ust-cyg-profile.so}::
4019 path:{liblttng-ust-cyg-profile-fast.so}::
4020 <<liblttng-ust-cyg-profile,Function entry and exit tracing>>.
4022 path:{liblttng-ust-dl.so}::
4023 <<liblttng-ust-dl,Dynamic linker tracing>>.
4025 To use a user space tracing helper with any user application:
4027 * Preload the helper shared object when you start the application:
4032 $ LD_PRELOAD=liblttng-ust-libc-wrapper.so my-app
4036 You can preload more than one helper:
4041 $ LD_PRELOAD=liblttng-ust-libc-wrapper.so:liblttng-ust-dl.so my-app
4047 [[liblttng-ust-libc-pthread-wrapper]]
4048 ==== Instrument C standard library memory and POSIX threads functions
4050 The path:{liblttng-ust-libc-wrapper.so} and
4051 path:{liblttng-ust-pthread-wrapper.so} helpers
4052 add instrumentation to some C standard library and POSIX
4056 .Functions instrumented by preloading path:{liblttng-ust-libc-wrapper.so}.
4058 |TP provider name |TP name |Instrumented function
4060 .6+|`lttng_ust_libc` |`malloc` |man:malloc(3)
4061 |`calloc` |man:calloc(3)
4062 |`realloc` |man:realloc(3)
4063 |`free` |man:free(3)
4064 |`memalign` |man:memalign(3)
4065 |`posix_memalign` |man:posix_memalign(3)
4069 .Functions instrumented by preloading path:{liblttng-ust-pthread-wrapper.so}.
4071 |TP provider name |TP name |Instrumented function
4073 .4+|`lttng_ust_pthread` |`pthread_mutex_lock_req` |man:pthread_mutex_lock(3p) (request time)
4074 |`pthread_mutex_lock_acq` |man:pthread_mutex_lock(3p) (acquire time)
4075 |`pthread_mutex_trylock` |man:pthread_mutex_trylock(3p)
4076 |`pthread_mutex_unlock` |man:pthread_mutex_unlock(3p)
4079 When you preload the shared object, it replaces the functions listed
4080 in the previous tables by wrappers which contain tracepoints and call
4081 the replaced functions.
4084 [[liblttng-ust-cyg-profile]]
4085 ==== Instrument function entry and exit
4087 The path:{liblttng-ust-cyg-profile*.so} helpers can add instrumentation
4088 to the entry and exit points of functions.
4090 man:gcc(1) and man:clang(1) have an option named
4091 https://gcc.gnu.org/onlinedocs/gcc/Instrumentation-Options.html[`-finstrument-functions`]
4092 which generates instrumentation calls for entry and exit to functions.
4093 The LTTng-UST function tracing helpers,
4094 path:{liblttng-ust-cyg-profile.so} and
4095 path:{liblttng-ust-cyg-profile-fast.so}, take advantage of this feature
4096 to add tracepoints to the two generated functions (which contain
4097 `cyg_profile` in their names, hence the helper's name).
4099 To use the LTTng-UST function tracing helper, the source files to
4100 instrument must be built using the `-finstrument-functions` compiler
4103 There are two versions of the LTTng-UST function tracing helper:
4105 * **path:{liblttng-ust-cyg-profile-fast.so}** is a lightweight variant
4106 that you should only use when it can be _guaranteed_ that the
4107 complete event stream is recorded without any lost event record.
4108 Any kind of duplicate information is left out.
4110 Assuming no event record is lost, having only the function addresses on
4111 entry is enough to create a call graph, since an event record always
4112 contains the ID of the CPU that generated it.
4114 You can use a tool like man:addr2line(1) to convert function addresses
4115 back to source file names and line numbers.
4117 * **path:{liblttng-ust-cyg-profile.so}** is a more robust variant
4118 which also works in use cases where event records might get discarded or
4119 not recorded from application startup.
4120 In these cases, the trace analyzer needs more information to be
4121 able to reconstruct the program flow.
4123 See man:lttng-ust-cyg-profile(3) to learn more about the instrumentation
4124 points of this helper.
4126 All the tracepoints that this helper provides have the
4127 log level `TRACE_DEBUG_FUNCTION` (see man:lttng-ust(3)).
4129 TIP: It's sometimes a good idea to limit the number of source files that
4130 you compile with the `-finstrument-functions` option to prevent LTTng
4131 from writing an excessive amount of trace data at run time. When using
4132 man:gcc(1), you can use the
4133 `-finstrument-functions-exclude-function-list` option to avoid
4134 instrument entries and exits of specific function names.
4139 ==== Instrument the dynamic linker
4141 The path:{liblttng-ust-dl.so} helper adds instrumentation to the
4142 man:dlopen(3) and man:dlclose(3) function calls.
4144 See man:lttng-ust-dl(3) to learn more about the instrumentation points
4149 [[java-application]]
4150 === User space Java agent
4152 You can instrument any Java application which uses one of the following
4155 * The https://docs.oracle.com/javase/7/docs/api/java/util/logging/package-summary.html[**`java.util.logging`**]
4156 (JUL) core logging facilities.
4157 * http://logging.apache.org/log4j/1.2/[**Apache log4j 1.2**], since
4158 LTTng 2.6. Note that Apache Log4j{nbsp}2 is not supported.
4161 .LTTng-UST Java agent imported by a Java application.
4162 image::java-app.png[]
4164 Note that the methods described below are new in LTTng{nbsp}{revision}.
4165 Previous LTTng versions use another technique.
4167 NOTE: We use http://openjdk.java.net/[OpenJDK]{nbsp}8 for development
4168 and https://ci.lttng.org/[continuous integration], thus this version is
4169 directly supported. However, the LTTng-UST Java agent is also tested
4170 with OpenJDK{nbsp}7.
4175 ==== Use the LTTng-UST Java agent for `java.util.logging`
4177 To use the LTTng-UST Java agent in a Java application which uses
4178 `java.util.logging` (JUL):
4180 . In the Java application's source code, import the LTTng-UST
4181 log handler package for `java.util.logging`:
4186 import org.lttng.ust.agent.jul.LttngLogHandler;
4190 . Create an LTTng-UST JUL log handler:
4195 Handler lttngUstLogHandler = new LttngLogHandler();
4199 . Add this handler to the JUL loggers which should emit LTTng events:
4204 Logger myLogger = Logger.getLogger("some-logger");
4206 myLogger.addHandler(lttngUstLogHandler);
4210 . Use `java.util.logging` log statements and configuration as usual.
4211 The loggers with an attached LTTng-UST log handler can emit
4214 . Before exiting the application, remove the LTTng-UST log handler from
4215 the loggers attached to it and call its `close()` method:
4220 myLogger.removeHandler(lttngUstLogHandler);
4221 lttngUstLogHandler.close();
4225 This is not strictly necessary, but it is recommended for a clean
4226 disposal of the handler's resources.
4228 . Include the LTTng-UST Java agent's common and JUL-specific JAR files,
4229 path:{lttng-ust-agent-common.jar} and path:{lttng-ust-agent-jul.jar},
4231 https://docs.oracle.com/javase/tutorial/essential/environment/paths.html[class
4232 path] when you build the Java application.
4234 The JAR files are typically located in dir:{/usr/share/java}.
4236 IMPORTANT: The LTTng-UST Java agent must be
4237 <<installing-lttng,installed>> for the logging framework your
4240 .Use the LTTng-UST Java agent for `java.util.logging`.
4245 import java.io.IOException;
4246 import java.util.logging.Handler;
4247 import java.util.logging.Logger;
4248 import org.lttng.ust.agent.jul.LttngLogHandler;
4252 private static final int answer = 42;
4254 public static void main(String[] argv) throws Exception
4257 Logger logger = Logger.getLogger("jello");
4259 // Create an LTTng-UST log handler
4260 Handler lttngUstLogHandler = new LttngLogHandler();
4262 // Add the LTTng-UST log handler to our logger
4263 logger.addHandler(lttngUstLogHandler);
4266 logger.info("some info");
4267 logger.warning("some warning");
4269 logger.finer("finer information; the answer is " + answer);
4271 logger.severe("error!");
4273 // Not mandatory, but cleaner
4274 logger.removeHandler(lttngUstLogHandler);
4275 lttngUstLogHandler.close();
4284 $ javac -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-jul.jar Test.java
4287 <<creating-destroying-tracing-sessions,Create a tracing session>>,
4288 <<enabling-disabling-events,create an event rule>> matching the
4289 `jello` JUL logger, and <<basic-tracing-session-control,start tracing>>:
4294 $ lttng enable-event --jul jello
4298 Run the compiled class:
4302 $ java -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-jul.jar:. Test
4305 <<basic-tracing-session-control,Stop tracing>> and inspect the
4315 In the resulting trace, an <<event,event record>> generated by a Java
4316 application using `java.util.logging` is named `lttng_jul:event` and
4317 has the following fields:
4320 Log record's message.
4326 Name of the class in which the log statement was executed.
4329 Name of the method in which the log statement was executed.
4332 Logging time (timestamp in milliseconds).
4335 Log level integer value.
4338 ID of the thread in which the log statement was executed.
4340 You can use the opt:lttng-enable-event(1):--loglevel or
4341 opt:lttng-enable-event(1):--loglevel-only option of the
4342 man:lttng-enable-event(1) command to target a range of JUL log levels
4343 or a specific JUL log level.
4348 ==== Use the LTTng-UST Java agent for Apache log4j
4350 To use the LTTng-UST Java agent in a Java application which uses
4353 . In the Java application's source code, import the LTTng-UST
4354 log appender package for Apache log4j:
4359 import org.lttng.ust.agent.log4j.LttngLogAppender;
4363 . Create an LTTng-UST log4j log appender:
4368 Appender lttngUstLogAppender = new LttngLogAppender();
4372 . Add this appender to the log4j loggers which should emit LTTng events:
4377 Logger myLogger = Logger.getLogger("some-logger");
4379 myLogger.addAppender(lttngUstLogAppender);
4383 . Use Apache log4j log statements and configuration as usual. The
4384 loggers with an attached LTTng-UST log appender can emit LTTng events.
4386 . Before exiting the application, remove the LTTng-UST log appender from
4387 the loggers attached to it and call its `close()` method:
4392 myLogger.removeAppender(lttngUstLogAppender);
4393 lttngUstLogAppender.close();
4397 This is not strictly necessary, but it is recommended for a clean
4398 disposal of the appender's resources.
4400 . Include the LTTng-UST Java agent's common and log4j-specific JAR
4401 files, path:{lttng-ust-agent-common.jar} and
4402 path:{lttng-ust-agent-log4j.jar}, in the
4403 https://docs.oracle.com/javase/tutorial/essential/environment/paths.html[class
4404 path] when you build the Java application.
4406 The JAR files are typically located in dir:{/usr/share/java}.
4408 IMPORTANT: The LTTng-UST Java agent must be
4409 <<installing-lttng,installed>> for the logging framework your
4412 .Use the LTTng-UST Java agent for Apache log4j.
4417 import org.apache.log4j.Appender;
4418 import org.apache.log4j.Logger;
4419 import org.lttng.ust.agent.log4j.LttngLogAppender;
4423 private static final int answer = 42;
4425 public static void main(String[] argv) throws Exception
4428 Logger logger = Logger.getLogger("jello");
4430 // Create an LTTng-UST log appender
4431 Appender lttngUstLogAppender = new LttngLogAppender();
4433 // Add the LTTng-UST log appender to our logger
4434 logger.addAppender(lttngUstLogAppender);
4437 logger.info("some info");
4438 logger.warn("some warning");
4440 logger.debug("debug information; the answer is " + answer);
4442 logger.fatal("error!");
4444 // Not mandatory, but cleaner
4445 logger.removeAppender(lttngUstLogAppender);
4446 lttngUstLogAppender.close();
4452 Build this example (`$LOG4JPATH` is the path to the Apache log4j JAR
4457 $ javac -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-log4j.jar:$LOG4JPATH Test.java
4460 <<creating-destroying-tracing-sessions,Create a tracing session>>,
4461 <<enabling-disabling-events,create an event rule>> matching the
4462 `jello` log4j logger, and <<basic-tracing-session-control,start tracing>>:
4467 $ lttng enable-event --log4j jello
4471 Run the compiled class:
4475 $ java -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-log4j.jar:$LOG4JPATH:. Test
4478 <<basic-tracing-session-control,Stop tracing>> and inspect the
4488 In the resulting trace, an <<event,event record>> generated by a Java
4489 application using log4j is named `lttng_log4j:event` and
4490 has the following fields:
4493 Log record's message.
4499 Name of the class in which the log statement was executed.
4502 Name of the method in which the log statement was executed.
4505 Name of the file in which the executed log statement is located.
4508 Line number at which the log statement was executed.
4514 Log level integer value.
4517 Name of the Java thread in which the log statement was executed.
4519 You can use the opt:lttng-enable-event(1):--loglevel or
4520 opt:lttng-enable-event(1):--loglevel-only option of the
4521 man:lttng-enable-event(1) command to target a range of Apache log4j log levels
4522 or a specific log4j log level.
4526 [[java-application-context]]
4527 ==== Provide application-specific context fields in a Java application
4529 A Java application-specific context field is a piece of state provided
4530 by the application which <<adding-context,you can add>>, using the
4531 man:lttng-add-context(1) command, to each <<event,event record>>
4532 produced by the log statements of this application.
4534 For example, a given object might have a current request ID variable.
4535 You can create a context information retriever for this object and
4536 assign a name to this current request ID. You can then, using the
4537 man:lttng-add-context(1) command, add this context field by name to
4538 the JUL or log4j <<channel,channel>>.
4540 To provide application-specific context fields in a Java application:
4542 . In the Java application's source code, import the LTTng-UST
4543 Java agent context classes and interfaces:
4548 import org.lttng.ust.agent.context.ContextInfoManager;
4549 import org.lttng.ust.agent.context.IContextInfoRetriever;
4553 . Create a context information retriever class, that is, a class which
4554 implements the `IContextInfoRetriever` interface:
4559 class MyContextInfoRetriever implements IContextInfoRetriever
4562 public Object retrieveContextInfo(String key)
4564 if (key.equals("intCtx")) {
4566 } else if (key.equals("strContext")) {
4567 return "context value!";
4576 This `retrieveContextInfo()` method is the only member of the
4577 `IContextInfoRetriever` interface. Its role is to return the current
4578 value of a state by name to create a context field. The names of the
4579 context fields and which state variables they return depends on your
4582 All primitive types and objects are supported as context fields.
4583 When `retrieveContextInfo()` returns an object, the context field
4584 serializer calls its `toString()` method to add a string field to
4585 event records. The method can also return `null`, which means that
4586 no context field is available for the required name.
4588 . Register an instance of your context information retriever class to
4589 the context information manager singleton:
4594 IContextInfoRetriever cir = new MyContextInfoRetriever();
4595 ContextInfoManager cim = ContextInfoManager.getInstance();
4596 cim.registerContextInfoRetriever("retrieverName", cir);
4600 . Before exiting the application, remove your context information
4601 retriever from the context information manager singleton:
4606 ContextInfoManager cim = ContextInfoManager.getInstance();
4607 cim.unregisterContextInfoRetriever("retrieverName");
4611 This is not strictly necessary, but it is recommended for a clean
4612 disposal of some manager's resources.
4614 . Build your Java application with LTTng-UST Java agent support as
4615 usual, following the procedure for either the <<jul,JUL>> or
4616 <<log4j,Apache log4j>> framework.
4619 .Provide application-specific context fields in a Java application.
4624 import java.util.logging.Handler;
4625 import java.util.logging.Logger;
4626 import org.lttng.ust.agent.jul.LttngLogHandler;
4627 import org.lttng.ust.agent.context.ContextInfoManager;
4628 import org.lttng.ust.agent.context.IContextInfoRetriever;
4632 // Our context information retriever class
4633 private static class MyContextInfoRetriever
4634 implements IContextInfoRetriever
4637 public Object retrieveContextInfo(String key) {
4638 if (key.equals("intCtx")) {
4640 } else if (key.equals("strContext")) {
4641 return "context value!";
4648 private static final int answer = 42;
4650 public static void main(String args[]) throws Exception
4652 // Get the context information manager instance
4653 ContextInfoManager cim = ContextInfoManager.getInstance();
4655 // Create and register our context information retriever
4656 IContextInfoRetriever cir = new MyContextInfoRetriever();
4657 cim.registerContextInfoRetriever("myRetriever", cir);
4660 Logger logger = Logger.getLogger("jello");
4662 // Create an LTTng-UST log handler
4663 Handler lttngUstLogHandler = new LttngLogHandler();
4665 // Add the LTTng-UST log handler to our logger
4666 logger.addHandler(lttngUstLogHandler);
4669 logger.info("some info");
4670 logger.warning("some warning");
4672 logger.finer("finer information; the answer is " + answer);
4674 logger.severe("error!");
4676 // Not mandatory, but cleaner
4677 logger.removeHandler(lttngUstLogHandler);
4678 lttngUstLogHandler.close();
4679 cim.unregisterContextInfoRetriever("myRetriever");
4688 $ javac -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-jul.jar Test.java
4691 <<creating-destroying-tracing-sessions,Create a tracing session>>
4692 and <<enabling-disabling-events,create an event rule>> matching the
4698 $ lttng enable-event --jul jello
4701 <<adding-context,Add the application-specific context fields>> to the
4706 $ lttng add-context --jul --type='$app.myRetriever:intCtx'
4707 $ lttng add-context --jul --type='$app.myRetriever:strContext'
4710 <<basic-tracing-session-control,Start tracing>>:
4717 Run the compiled class:
4721 $ java -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-jul.jar:. Test
4724 <<basic-tracing-session-control,Stop tracing>> and inspect the
4736 [[python-application]]
4737 === User space Python agent
4739 You can instrument a Python 2 or Python 3 application which uses the
4740 standard https://docs.python.org/3/library/logging.html[`logging`]
4743 Each log statement emits an LTTng event once the
4744 application module imports the
4745 <<lttng-ust-agents,LTTng-UST Python agent>> package.
4748 .A Python application importing the LTTng-UST Python agent.
4749 image::python-app.png[]
4751 To use the LTTng-UST Python agent:
4753 . In the Python application's source code, import the LTTng-UST Python
4763 The LTTng-UST Python agent automatically adds its logging handler to the
4764 root logger at import time.
4766 Any log statement that the application executes before this import does
4767 not emit an LTTng event.
4769 IMPORTANT: The LTTng-UST Python agent must be
4770 <<installing-lttng,installed>>.
4772 . Use log statements and logging configuration as usual.
4773 Since the LTTng-UST Python agent adds a handler to the _root_
4774 logger, you can trace any log statement from any logger.
4776 .Use the LTTng-UST Python agent.
4787 logging.basicConfig()
4788 logger = logging.getLogger('my-logger')
4791 logger.debug('debug message')
4792 logger.info('info message')
4793 logger.warn('warn message')
4794 logger.error('error message')
4795 logger.critical('critical message')
4799 if __name__ == '__main__':
4803 NOTE: `logging.basicConfig()`, which adds to the root logger a basic
4804 logging handler which prints to the standard error stream, is not
4805 strictly required for LTTng-UST tracing to work, but in versions of
4806 Python preceding 3.2, you could see a warning message which indicates
4807 that no handler exists for the logger `my-logger`.
4809 <<creating-destroying-tracing-sessions,Create a tracing session>>,
4810 <<enabling-disabling-events,create an event rule>> matching the
4811 `my-logger` Python logger, and <<basic-tracing-session-control,start
4817 $ lttng enable-event --python my-logger
4821 Run the Python script:
4828 <<basic-tracing-session-control,Stop tracing>> and inspect the recorded
4838 In the resulting trace, an <<event,event record>> generated by a Python
4839 application is named `lttng_python:event` and has the following fields:
4842 Logging time (string).
4845 Log record's message.
4851 Name of the function in which the log statement was executed.
4854 Line number at which the log statement was executed.
4857 Log level integer value.
4860 ID of the Python thread in which the log statement was executed.
4863 Name of the Python thread in which the log statement was executed.
4865 You can use the opt:lttng-enable-event(1):--loglevel or
4866 opt:lttng-enable-event(1):--loglevel-only option of the
4867 man:lttng-enable-event(1) command to target a range of Python log levels
4868 or a specific Python log level.
4870 When an application imports the LTTng-UST Python agent, the agent tries
4871 to register to a <<lttng-sessiond,session daemon>>. Note that you must
4872 <<start-sessiond,start the session daemon>> _before_ you run the Python
4873 application. If a session daemon is found, the agent tries to register
4874 to it during 5{nbsp}seconds, after which the application continues
4875 without LTTng tracing support. You can override this timeout value with
4876 the env:LTTNG_UST_PYTHON_REGISTER_TIMEOUT environment variable
4879 If the session daemon stops while a Python application with an imported
4880 LTTng-UST Python agent runs, the agent retries to connect and to
4881 register to a session daemon every 3{nbsp}seconds. You can override this
4882 delay with the env:LTTNG_UST_PYTHON_REGISTER_RETRY_DELAY environment
4887 [[proc-lttng-logger-abi]]
4890 The `lttng-tracer` Linux kernel module, part of
4891 <<lttng-modules,LTTng-modules>>, creates the special LTTng logger file
4892 path:{/proc/lttng-logger} when it's loaded. Any application can write
4893 text data to this file to emit an LTTng event.
4896 .An application writes to the LTTng logger file to emit an LTTng event.
4897 image::lttng-logger.png[]
4899 The LTTng logger is the quickest method--not the most efficient,
4900 however--to add instrumentation to an application. It is designed
4901 mostly to instrument shell scripts:
4905 $ echo "Some message, some $variable" > /proc/lttng-logger
4908 Any event that the LTTng logger emits is named `lttng_logger` and
4909 belongs to the Linux kernel <<domain,tracing domain>>. However, unlike
4910 other instrumentation points in the kernel tracing domain, **any Unix
4911 user** can <<enabling-disabling-events,create an event rule>> which
4912 matches its event name, not only the root user or users in the
4913 <<tracing-group,tracing group>>.
4915 To use the LTTng logger:
4917 * From any application, write text data to the path:{/proc/lttng-logger}
4920 The `msg` field of `lttng_logger` event records contains the
4923 NOTE: The maximum message length of an LTTng logger event is
4924 1024{nbsp}bytes. Writing more than this makes the LTTng logger emit more
4925 than one event to contain the remaining data.
4927 You should not use the LTTng logger to trace a user application which
4928 can be instrumented in a more efficient way, namely:
4930 * <<c-application,C and $$C++$$ applications>>.
4931 * <<java-application,Java applications>>.
4932 * <<python-application,Python applications>>.
4934 .Use the LTTng logger.
4939 echo 'Hello, World!' > /proc/lttng-logger
4941 df --human-readable --print-type / > /proc/lttng-logger
4944 <<creating-destroying-tracing-sessions,Create a tracing session>>,
4945 <<enabling-disabling-events,create an event rule>> matching the
4946 `lttng_logger` Linux kernel tracepoint, and
4947 <<basic-tracing-session-control,start tracing>>:
4952 $ lttng enable-event --kernel lttng_logger
4956 Run the Bash script:
4963 <<basic-tracing-session-control,Stop tracing>> and inspect the recorded
4974 [[instrumenting-linux-kernel]]
4975 === LTTng kernel tracepoints
4977 NOTE: This section shows how to _add_ instrumentation points to the
4978 Linux kernel. The kernel's subsystems are already thoroughly
4979 instrumented at strategic places for LTTng when you
4980 <<installing-lttng,install>> the <<lttng-modules,LTTng-modules>>
4984 There are two methods to instrument the Linux kernel:
4986 . <<linux-add-lttng-layer,Add an LTTng layer>> over an existing ftrace
4987 tracepoint which uses the `TRACE_EVENT()` API.
4989 Choose this if you want to instrumentation a Linux kernel tree with an
4990 instrumentation point compatible with ftrace, perf, and SystemTap.
4992 . Use an <<linux-lttng-tracepoint-event,LTTng-only approach>> to
4993 instrument an out-of-tree kernel module.
4995 Choose this if you don't need ftrace, perf, or SystemTap support.
4999 [[linux-add-lttng-layer]]
5000 ==== [[instrumenting-linux-kernel-itself]][[mainline-trace-event]][[lttng-adaptation-layer]]Add an LTTng layer to an existing ftrace tracepoint
5002 This section shows how to add an LTTng layer to existing ftrace
5003 instrumentation using the `TRACE_EVENT()` API.
5005 This section does not document the `TRACE_EVENT()` macro. You can
5006 read the following articles to learn more about this API:
5008 * http://lwn.net/Articles/379903/[Using the TRACE_EVENT() macro (Part 1)]
5009 * http://lwn.net/Articles/381064/[Using the TRACE_EVENT() macro (Part 2)]
5010 * http://lwn.net/Articles/383362/[Using the TRACE_EVENT() macro (Part 3)]
5012 The following procedure assumes that your ftrace tracepoints are
5013 correctly defined in their own header and that they are created in
5014 one source file using the `CREATE_TRACE_POINTS` definition.
5016 To add an LTTng layer over an existing ftrace tracepoint:
5018 . Make sure the following kernel configuration options are
5024 * `CONFIG_HIGH_RES_TIMERS`
5025 * `CONFIG_TRACEPOINTS`
5028 . Build the Linux source tree with your custom ftrace tracepoints.
5029 . Boot the resulting Linux image on your target system.
5031 Confirm that the tracepoints exist by looking for their names in the
5032 dir:{/sys/kernel/debug/tracing/events/subsys} directory, where `subsys`
5033 is your subsystem's name.
5035 . Get a copy of the latest LTTng-modules{nbsp}{revision}:
5040 $ cd $(mktemp -d) &&
5041 wget http://lttng.org/files/lttng-modules/lttng-modules-latest-2.10.tar.bz2 &&
5042 tar -xf lttng-modules-latest-2.10.tar.bz2 &&
5043 cd lttng-modules-2.10.*
5047 . In dir:{instrumentation/events/lttng-module}, relative to the root
5048 of the LTTng-modules source tree, create a header file named
5049 +__subsys__.h+ for your custom subsystem +__subsys__+ and write your
5050 LTTng-modules tracepoint definitions using the LTTng-modules
5053 Start with this template:
5057 .path:{instrumentation/events/lttng-module/my_subsys.h}
5060 #define TRACE_SYSTEM my_subsys
5062 #if !defined(_LTTNG_MY_SUBSYS_H) || defined(TRACE_HEADER_MULTI_READ)
5063 #define _LTTNG_MY_SUBSYS_H
5065 #include "../../../probes/lttng-tracepoint-event.h"
5066 #include <linux/tracepoint.h>
5068 LTTNG_TRACEPOINT_EVENT(
5070 * Format is identical to TRACE_EVENT()'s version for the three
5071 * following macro parameters:
5074 TP_PROTO(int my_int, const char *my_string),
5075 TP_ARGS(my_int, my_string),
5077 /* LTTng-modules specific macros */
5079 ctf_integer(int, my_int_field, my_int)
5080 ctf_string(my_bar_field, my_bar)
5084 #endif /* !defined(_LTTNG_MY_SUBSYS_H) || defined(TRACE_HEADER_MULTI_READ) */
5086 #include "../../../probes/define_trace.h"
5090 The entries in the `TP_FIELDS()` section are the list of fields for the
5091 LTTng tracepoint. This is similar to the `TP_STRUCT__entry()` part of
5092 ftrace's `TRACE_EVENT()` macro.
5094 See <<lttng-modules-tp-fields,Tracepoint fields macros>> for a
5095 complete description of the available `ctf_*()` macros.
5097 . Create the LTTng-modules probe's kernel module C source file,
5098 +probes/lttng-probe-__subsys__.c+, where +__subsys__+ is your
5103 .path:{probes/lttng-probe-my-subsys.c}
5105 #include <linux/module.h>
5106 #include "../lttng-tracer.h"
5109 * Build-time verification of mismatch between mainline
5110 * TRACE_EVENT() arguments and the LTTng-modules adaptation
5111 * layer LTTNG_TRACEPOINT_EVENT() arguments.
5113 #include <trace/events/my_subsys.h>
5115 /* Create LTTng tracepoint probes */
5116 #define LTTNG_PACKAGE_BUILD
5117 #define CREATE_TRACE_POINTS
5118 #define TRACE_INCLUDE_PATH ../instrumentation/events/lttng-module
5120 #include "../instrumentation/events/lttng-module/my_subsys.h"
5122 MODULE_LICENSE("GPL and additional rights");
5123 MODULE_AUTHOR("Your name <your-email>");
5124 MODULE_DESCRIPTION("LTTng my_subsys probes");
5125 MODULE_VERSION(__stringify(LTTNG_MODULES_MAJOR_VERSION) "."
5126 __stringify(LTTNG_MODULES_MINOR_VERSION) "."
5127 __stringify(LTTNG_MODULES_PATCHLEVEL_VERSION)
5128 LTTNG_MODULES_EXTRAVERSION);
5132 . Edit path:{probes/KBuild} and add your new kernel module object
5133 next to the existing ones:
5137 .path:{probes/KBuild}
5141 obj-m += lttng-probe-module.o
5142 obj-m += lttng-probe-power.o
5144 obj-m += lttng-probe-my-subsys.o
5150 . Build and install the LTTng kernel modules:
5155 $ make KERNELDIR=/path/to/linux
5156 # make modules_install && depmod -a
5160 Replace `/path/to/linux` with the path to the Linux source tree where
5161 you defined and used tracepoints with ftrace's `TRACE_EVENT()` macro.
5163 Note that you can also use the
5164 <<lttng-tracepoint-event-code,`LTTNG_TRACEPOINT_EVENT_CODE()` macro>>
5165 instead of `LTTNG_TRACEPOINT_EVENT()` to use custom local variables and
5166 C code that need to be executed before the event fields are recorded.
5168 The best way to learn how to use the previous LTTng-modules macros is to
5169 inspect the existing LTTng-modules tracepoint definitions in the
5170 dir:{instrumentation/events/lttng-module} header files. Compare them
5171 with the Linux kernel mainline versions in the
5172 dir:{include/trace/events} directory of the Linux source tree.
5176 [[lttng-tracepoint-event-code]]
5177 ===== Use custom C code to access the data for tracepoint fields
5179 Although we recommended to always use the
5180 <<lttng-adaptation-layer,`LTTNG_TRACEPOINT_EVENT()`>> macro to describe
5181 the arguments and fields of an LTTng-modules tracepoint when possible,
5182 sometimes you need a more complex process to access the data that the
5183 tracer records as event record fields. In other words, you need local
5184 variables and multiple C{nbsp}statements instead of simple
5185 argument-based expressions that you pass to the
5186 <<lttng-modules-tp-fields,`ctf_*()` macros of `TP_FIELDS()`>>.
5188 You can use the `LTTNG_TRACEPOINT_EVENT_CODE()` macro instead of
5189 `LTTNG_TRACEPOINT_EVENT()` to declare custom local variables and define
5190 a block of C{nbsp}code to be executed before LTTng records the fields.
5191 The structure of this macro is:
5194 .`LTTNG_TRACEPOINT_EVENT_CODE()` macro syntax.
5196 LTTNG_TRACEPOINT_EVENT_CODE(
5198 * Format identical to the LTTNG_TRACEPOINT_EVENT()
5199 * version for the following three macro parameters:
5202 TP_PROTO(int my_int, const char *my_string),
5203 TP_ARGS(my_int, my_string),
5205 /* Declarations of custom local variables */
5208 unsigned long b = 0;
5209 const char *name = "(undefined)";
5210 struct my_struct *my_struct;
5214 * Custom code which uses both tracepoint arguments
5215 * (in TP_ARGS()) and local variables (in TP_locvar()).
5217 * Local variables are actually members of a structure pointed
5218 * to by the special variable tp_locvar.
5222 tp_locvar->a = my_int + 17;
5223 tp_locvar->my_struct = get_my_struct_at(tp_locvar->a);
5224 tp_locvar->b = my_struct_compute_b(tp_locvar->my_struct);
5225 tp_locvar->name = my_struct_get_name(tp_locvar->my_struct);
5226 put_my_struct(tp_locvar->my_struct);
5235 * Format identical to the LTTNG_TRACEPOINT_EVENT()
5236 * version for this, except that tp_locvar members can be
5237 * used in the argument expression parameters of
5238 * the ctf_*() macros.
5241 ctf_integer(unsigned long, my_struct_b, tp_locvar->b)
5242 ctf_integer(int, my_struct_a, tp_locvar->a)
5243 ctf_string(my_string_field, my_string)
5244 ctf_string(my_struct_name, tp_locvar->name)
5249 IMPORTANT: The C code defined in `TP_code()` must not have any side
5250 effects when executed. In particular, the code must not allocate
5251 memory or get resources without deallocating this memory or putting
5252 those resources afterwards.
5255 [[instrumenting-linux-kernel-tracing]]
5256 ==== Load and unload a custom probe kernel module
5258 You must load a <<lttng-adaptation-layer,created LTTng-modules probe
5259 kernel module>> in the kernel before it can emit LTTng events.
5261 To load the default probe kernel modules and a custom probe kernel
5264 * Use the opt:lttng-sessiond(8):--extra-kmod-probes option to give extra
5265 probe modules to load when starting a root <<lttng-sessiond,session
5269 .Load the `my_subsys`, `usb`, and the default probe modules.
5273 # lttng-sessiond --extra-kmod-probes=my_subsys,usb
5278 You only need to pass the subsystem name, not the whole kernel module
5281 To load _only_ a given custom probe kernel module:
5283 * Use the opt:lttng-sessiond(8):--kmod-probes option to give the probe
5284 modules to load when starting a root session daemon:
5287 .Load only the `my_subsys` and `usb` probe modules.
5291 # lttng-sessiond --kmod-probes=my_subsys,usb
5296 To confirm that a probe module is loaded:
5303 $ lsmod | grep lttng_probe_usb
5307 To unload the loaded probe modules:
5309 * Kill the session daemon with `SIGTERM`:
5314 # pkill lttng-sessiond
5318 You can also use man:modprobe(8)'s `--remove` option if the session
5319 daemon terminates abnormally.
5322 [[controlling-tracing]]
5325 Once an application or a Linux kernel is
5326 <<instrumenting,instrumented>> for LTTng tracing,
5329 This section is divided in topics on how to use the various
5330 <<plumbing,components of LTTng>>, in particular the <<lttng-cli,cmd:lttng
5331 command-line tool>>, to _control_ the LTTng daemons and tracers.
5333 NOTE: In the following subsections, we refer to an man:lttng(1) command
5334 using its man page name. For example, instead of _Run the `create`
5335 command to..._, we use _Run the man:lttng-create(1) command to..._.
5339 === Start a session daemon
5341 In some situations, you need to run a <<lttng-sessiond,session daemon>>
5342 (man:lttng-sessiond(8)) _before_ you can use the man:lttng(1)
5345 You will see the following error when you run a command while no session
5349 Error: No session daemon is available
5352 The only command that automatically runs a session daemon is
5353 man:lttng-create(1), which you use to
5354 <<creating-destroying-tracing-sessions,create a tracing session>>. While
5355 this is most of the time the first operation that you do, sometimes it's
5356 not. Some examples are:
5358 * <<list-instrumentation-points,List the available instrumentation points>>.
5359 * <<saving-loading-tracing-session,Load a tracing session configuration>>.
5361 [[tracing-group]] Each Unix user must have its own running session
5362 daemon to trace user applications. The session daemon that the root user
5363 starts is the only one allowed to control the LTTng kernel tracer. Users
5364 that are part of the _tracing group_ can control the root session
5365 daemon. The default tracing group name is `tracing`; you can set it to
5366 something else with the opt:lttng-sessiond(8):--group option when you
5367 start the root session daemon.
5369 To start a user session daemon:
5371 * Run man:lttng-sessiond(8):
5376 $ lttng-sessiond --daemonize
5380 To start the root session daemon:
5382 * Run man:lttng-sessiond(8) as the root user:
5387 # lttng-sessiond --daemonize
5391 In both cases, remove the opt:lttng-sessiond(8):--daemonize option to
5392 start the session daemon in foreground.
5394 To stop a session daemon, use man:kill(1) on its process ID (standard
5397 Note that some Linux distributions could manage the LTTng session daemon
5398 as a service. In this case, you should use the service manager to
5399 start, restart, and stop session daemons.
5402 [[creating-destroying-tracing-sessions]]
5403 === Create and destroy a tracing session
5405 Almost all the LTTng control operations happen in the scope of
5406 a <<tracing-session,tracing session>>, which is the dialogue between the
5407 <<lttng-sessiond,session daemon>> and you.
5409 To create a tracing session with a generated name:
5411 * Use the man:lttng-create(1) command:
5420 The created tracing session's name is `auto` followed by the
5423 To create a tracing session with a specific name:
5425 * Use the optional argument of the man:lttng-create(1) command:
5430 $ lttng create my-session
5434 Replace `my-session` with the specific tracing session name.
5436 LTTng appends the creation date to the created tracing session's name.
5438 LTTng writes the traces of a tracing session in
5439 +$LTTNG_HOME/lttng-trace/__name__+ by default, where +__name__+ is the
5440 name of the tracing session. Note that the env:LTTNG_HOME environment
5441 variable defaults to `$HOME` if not set.
5443 To output LTTng traces to a non-default location:
5445 * Use the opt:lttng-create(1):--output option of the man:lttng-create(1) command:
5450 $ lttng create my-session --output=/tmp/some-directory
5454 You may create as many tracing sessions as you wish.
5456 To list all the existing tracing sessions for your Unix user:
5458 * Use the man:lttng-list(1) command:
5467 When you create a tracing session, it is set as the _current tracing
5468 session_. The following man:lttng(1) commands operate on the current
5469 tracing session when you don't specify one:
5471 [role="list-3-cols"]
5488 To change the current tracing session:
5490 * Use the man:lttng-set-session(1) command:
5495 $ lttng set-session new-session
5499 Replace `new-session` by the name of the new current tracing session.
5501 When you are done tracing in a given tracing session, you can destroy
5502 it. This operation frees the resources taken by the tracing session
5503 to destroy; it does not destroy the trace data that LTTng wrote for
5504 this tracing session.
5506 To destroy the current tracing session:
5508 * Use the man:lttng-destroy(1) command:
5518 [[list-instrumentation-points]]
5519 === List the available instrumentation points
5521 The <<lttng-sessiond,session daemon>> can query the running instrumented
5522 user applications and the Linux kernel to get a list of available
5523 instrumentation points. For the Linux kernel <<domain,tracing domain>>,
5524 they are tracepoints and system calls. For the user space tracing
5525 domain, they are tracepoints. For the other tracing domains, they are
5528 To list the available instrumentation points:
5530 * Use the man:lttng-list(1) command with the requested tracing domain's
5534 * opt:lttng-list(1):--kernel: Linux kernel tracepoints (your Unix user
5535 must be a root user, or it must be a member of the
5536 <<tracing-group,tracing group>>).
5537 * opt:lttng-list(1):--kernel with opt:lttng-list(1):--syscall: Linux
5538 kernel system calls (your Unix user must be a root user, or it must be
5539 a member of the tracing group).
5540 * opt:lttng-list(1):--userspace: user space tracepoints.
5541 * opt:lttng-list(1):--jul: `java.util.logging` loggers.
5542 * opt:lttng-list(1):--log4j: Apache log4j loggers.
5543 * opt:lttng-list(1):--python: Python loggers.
5546 .List the available user space tracepoints.
5550 $ lttng list --userspace
5554 .List the available Linux kernel system call tracepoints.
5558 $ lttng list --kernel --syscall
5563 [[enabling-disabling-events]]
5564 === Create and enable an event rule
5566 Once you <<creating-destroying-tracing-sessions,create a tracing
5567 session>>, you can create <<event,event rules>> with the
5568 man:lttng-enable-event(1) command.
5570 You specify each condition with a command-line option. The available
5571 condition options are shown in the following table.
5573 [role="growable",cols="asciidoc,asciidoc,default"]
5574 .Condition command-line options for the man:lttng-enable-event(1) command.
5576 |Option |Description |Applicable tracing domains
5582 . +--probe=__ADDR__+
5583 . +--function=__ADDR__+
5586 Instead of using the default _tracepoint_ instrumentation type, use:
5588 . A Linux system call.
5589 . A Linux https://lwn.net/Articles/132196/[KProbe] (symbol or address).
5590 . The entry and return points of a Linux function (symbol or address).
5594 |First positional argument.
5597 Tracepoint or system call name. In the case of a Linux KProbe or
5598 function, this is a custom name given to the event rule. With the
5599 JUL, log4j, and Python domains, this is a logger name.
5601 With a tracepoint, logger, or system call name, the last character
5602 can be `*` to match anything that remains.
5609 . +--loglevel=__LEVEL__+
5610 . +--loglevel-only=__LEVEL__+
5613 . Match only tracepoints or log statements with a logging level at
5614 least as severe as +__LEVEL__+.
5615 . Match only tracepoints or log statements with a logging level
5616 equal to +__LEVEL__+.
5618 See man:lttng-enable-event(1) for the list of available logging level
5621 |User space, JUL, log4j, and Python.
5623 |+--exclude=__EXCLUSIONS__+
5626 When you use a `*` character at the end of the tracepoint or logger
5627 name (first positional argument), exclude the specific names in the
5628 comma-delimited list +__EXCLUSIONS__+.
5631 User space, JUL, log4j, and Python.
5633 |+--filter=__EXPR__+
5636 Match only events which satisfy the expression +__EXPR__+.
5638 See man:lttng-enable-event(1) to learn more about the syntax of a
5645 You attach an event rule to a <<channel,channel>> on creation. If you do
5646 not specify the channel with the opt:lttng-enable-event(1):--channel
5647 option, and if the event rule to create is the first in its
5648 <<domain,tracing domain>> for a given tracing session, then LTTng
5649 creates a _default channel_ for you. This default channel is reused in
5650 subsequent invocations of the man:lttng-enable-event(1) command for the
5651 same tracing domain.
5653 An event rule is always enabled at creation time.
5655 The following examples show how you can combine the previous
5656 command-line options to create simple to more complex event rules.
5658 .Create an event rule targetting a Linux kernel tracepoint (default channel).
5662 $ lttng enable-event --kernel sched_switch
5666 .Create an event rule matching four Linux kernel system calls (default channel).
5670 $ lttng enable-event --kernel --syscall open,write,read,close
5674 .Create event rules matching tracepoints with filter expressions (default channel).
5678 $ lttng enable-event --kernel sched_switch --filter='prev_comm == "bash"'
5683 $ lttng enable-event --kernel --all \
5684 --filter='$ctx.tid == 1988 || $ctx.tid == 1534'
5689 $ lttng enable-event --jul my_logger \
5690 --filter='$app.retriever:cur_msg_id > 3'
5693 IMPORTANT: Make sure to always quote the filter string when you
5694 use man:lttng(1) from a shell.
5697 .Create an event rule matching any user space tracepoint of a given tracepoint provider with a log level range (default channel).
5701 $ lttng enable-event --userspace my_app:'*' --loglevel=TRACE_INFO
5704 IMPORTANT: Make sure to always quote the wildcard character when you
5705 use man:lttng(1) from a shell.
5708 .Create an event rule matching multiple Python loggers with a wildcard and with exclusions (default channel).
5712 $ lttng enable-event --python my-app.'*' \
5713 --exclude='my-app.module,my-app.hello'
5717 .Create an event rule matching any Apache log4j logger with a specific log level (default channel).
5721 $ lttng enable-event --log4j --all --loglevel-only=LOG4J_WARN
5725 .Create an event rule attached to a specific channel matching a specific user space tracepoint provider and tracepoint.
5729 $ lttng enable-event --userspace my_app:my_tracepoint --channel=my-channel
5733 The event rules of a given channel form a whitelist: as soon as an
5734 emitted event passes one of them, LTTng can record the event. For
5735 example, an event named `my_app:my_tracepoint` emitted from a user space
5736 tracepoint with a `TRACE_ERROR` log level passes both of the following
5741 $ lttng enable-event --userspace my_app:my_tracepoint
5742 $ lttng enable-event --userspace my_app:my_tracepoint \
5743 --loglevel=TRACE_INFO
5746 The second event rule is redundant: the first one includes
5750 [[disable-event-rule]]
5751 === Disable an event rule
5753 To disable an event rule that you <<enabling-disabling-events,created>>
5754 previously, use the man:lttng-disable-event(1) command. This command
5755 disables _all_ the event rules (of a given tracing domain and channel)
5756 which match an instrumentation point. The other conditions are not
5757 supported as of LTTng{nbsp}{revision}.
5759 The LTTng tracer does not record an emitted event which passes
5760 a _disabled_ event rule.
5762 .Disable an event rule matching a Python logger (default channel).
5766 $ lttng disable-event --python my-logger
5770 .Disable an event rule matching all `java.util.logging` loggers (default channel).
5774 $ lttng disable-event --jul '*'
5778 .Disable _all_ the event rules of the default channel.
5780 The opt:lttng-disable-event(1):--all-events option is not, like the
5781 opt:lttng-enable-event(1):--all option of man:lttng-enable-event(1), the
5782 equivalent of the event name `*` (wildcard): it disables _all_ the event
5783 rules of a given channel.
5787 $ lttng disable-event --jul --all-events
5791 NOTE: You cannot delete an event rule once you create it.
5795 === Get the status of a tracing session
5797 To get the status of the current tracing session, that is, its
5798 parameters, its channels, event rules, and their attributes:
5800 * Use the man:lttng-status(1) command:
5810 To get the status of any tracing session:
5812 * Use the man:lttng-list(1) command with the tracing session's name:
5817 $ lttng list my-session
5821 Replace `my-session` with the desired tracing session's name.
5824 [[basic-tracing-session-control]]
5825 === Start and stop a tracing session
5827 Once you <<creating-destroying-tracing-sessions,create a tracing
5829 <<enabling-disabling-events,create one or more event rules>>,
5830 you can start and stop the tracers for this tracing session.
5832 To start tracing in the current tracing session:
5834 * Use the man:lttng-start(1) command:
5843 LTTng is very flexible: you can launch user applications before
5844 or after the you start the tracers. The tracers only record the events
5845 if they pass enabled event rules and if they occur while the tracers are
5848 To stop tracing in the current tracing session:
5850 * Use the man:lttng-stop(1) command:
5859 If there were <<channel-overwrite-mode-vs-discard-mode,lost event
5860 records>> or lost sub-buffers since the last time you ran
5861 man:lttng-start(1), warnings are printed when you run the
5862 man:lttng-stop(1) command.
5865 [[enabling-disabling-channels]]
5866 === Create a channel
5868 Once you create a tracing session, you can create a <<channel,channel>>
5869 with the man:lttng-enable-channel(1) command.
5871 Note that LTTng automatically creates a default channel when, for a
5872 given <<domain,tracing domain>>, no channels exist and you
5873 <<enabling-disabling-events,create>> the first event rule. This default
5874 channel is named `channel0` and its attributes are set to reasonable
5875 values. Therefore, you only need to create a channel when you need
5876 non-default attributes.
5878 You specify each non-default channel attribute with a command-line
5879 option when you use the man:lttng-enable-channel(1) command. The
5880 available command-line options are:
5882 [role="growable",cols="asciidoc,asciidoc"]
5883 .Command-line options for the man:lttng-enable-channel(1) command.
5885 |Option |Description
5891 <<channel-overwrite-mode-vs-discard-mode,event loss mode>> instead of
5892 the default _discard_ mode.
5894 |`--buffers-pid` (user space tracing domain only)
5897 Use the per-process <<channel-buffering-schemes,buffering scheme>>
5898 instead of the default per-user buffering scheme.
5900 |+--subbuf-size=__SIZE__+
5903 Allocate sub-buffers of +__SIZE__+ bytes (power of two), for each CPU,
5904 either for each Unix user (default), or for each instrumented process.
5906 See <<channel-subbuf-size-vs-subbuf-count,Sub-buffer count and size>>.
5908 |+--num-subbuf=__COUNT__+
5911 Allocate +__COUNT__+ sub-buffers (power of two), for each CPU, either
5912 for each Unix user (default), or for each instrumented process.
5914 See <<channel-subbuf-size-vs-subbuf-count,Sub-buffer count and size>>.
5916 |+--tracefile-size=__SIZE__+
5919 Set the maximum size of each trace file that this channel writes within
5920 a stream to +__SIZE__+ bytes instead of no maximum.
5922 See <<tracefile-rotation,Trace file count and size>>.
5924 |+--tracefile-count=__COUNT__+
5927 Limit the number of trace files that this channel creates to
5928 +__COUNT__+ channels instead of no limit.
5930 See <<tracefile-rotation,Trace file count and size>>.
5932 |+--switch-timer=__PERIODUS__+
5935 Set the <<channel-switch-timer,switch timer period>>
5936 to +__PERIODUS__+{nbsp}µs.
5938 |+--read-timer=__PERIODUS__+
5941 Set the <<channel-read-timer,read timer period>>
5942 to +__PERIODUS__+{nbsp}µs.
5944 |[[opt-blocking-timeout]]+--blocking-timeout=__TIMEOUTUS__+
5947 Set the timeout of user space applications which load LTTng-UST
5948 in blocking mode to +__TIMEOUTUS__+:
5951 Never block (non-blocking mode).
5954 Block forever until space is available in a sub-buffer to record
5957 __n__, a positive value::
5958 Wait for at most __n__ µs when trying to write into a sub-buffer.
5960 Note that, for this option to have any effect on an instrumented
5961 user space application, you need to run the application with a set
5962 env:LTTNG_UST_ALLOW_BLOCKING environment variable.
5964 |+--output=__TYPE__+ (Linux kernel tracing domain only)
5967 Set the channel's output type to +__TYPE__+, either `mmap` or `splice`.
5971 You can only create a channel in the Linux kernel and user space
5972 <<domain,tracing domains>>: other tracing domains have their own channel
5973 created on the fly when <<enabling-disabling-events,creating event
5978 Because of a current LTTng limitation, you must create all channels
5979 _before_ you <<basic-tracing-session-control,start tracing>> in a given
5980 tracing session, that is, before the first time you run
5983 Since LTTng automatically creates a default channel when you use the
5984 man:lttng-enable-event(1) command with a specific tracing domain, you
5985 cannot, for example, create a Linux kernel event rule, start tracing,
5986 and then create a user space event rule, because no user space channel
5987 exists yet and it's too late to create one.
5989 For this reason, make sure to configure your channels properly
5990 before starting the tracers for the first time!
5993 The following examples show how you can combine the previous
5994 command-line options to create simple to more complex channels.
5996 .Create a Linux kernel channel with default attributes.
6000 $ lttng enable-channel --kernel my-channel
6004 .Create a user space channel with 4 sub-buffers or 1{nbsp}MiB each, per CPU, per instrumented process.
6008 $ lttng enable-channel --userspace --num-subbuf=4 --subbuf-size=1M \
6009 --buffers-pid my-channel
6013 .[[blocking-timeout-example]]Create a default user space channel with an infinite blocking timeout.
6015 <<creating-destroying-tracing-sessions,Create a tracing-session>>,
6016 create the channel, <<enabling-disabling-events,create an event rule>>,
6017 and <<basic-tracing-session-control,start tracing>>:
6022 $ lttng enable-channel --userspace --blocking-timeout=-1 blocking-channel
6023 $ lttng enable-event --userspace --channel=blocking-channel --all
6027 Run an application instrumented with LTTng-UST and allow it to block:
6031 $ LTTNG_UST_ALLOW_BLOCKING=1 my-app
6035 .Create a Linux kernel channel which rotates 8 trace files of 4{nbsp}MiB each for each stream
6039 $ lttng enable-channel --kernel --tracefile-count=8 \
6040 --tracefile-size=4194304 my-channel
6044 .Create a user space channel in overwrite (or _flight recorder_) mode.
6048 $ lttng enable-channel --userspace --overwrite my-channel
6052 You can <<enabling-disabling-events,create>> the same event rule in
6053 two different channels:
6057 $ lttng enable-event --userspace --channel=my-channel app:tp
6058 $ lttng enable-event --userspace --channel=other-channel app:tp
6061 If both channels are enabled, when a tracepoint named `app:tp` is
6062 reached, LTTng records two events, one for each channel.
6066 === Disable a channel
6068 To disable a specific channel that you <<enabling-disabling-channels,created>>
6069 previously, use the man:lttng-disable-channel(1) command.
6071 .Disable a specific Linux kernel channel.
6075 $ lttng disable-channel --kernel my-channel
6079 The state of a channel precedes the individual states of event rules
6080 attached to it: event rules which belong to a disabled channel, even if
6081 they are enabled, are also considered disabled.
6085 === Add context fields to a channel
6087 Event record fields in trace files provide important information about
6088 events that occured previously, but sometimes some external context may
6089 help you solve a problem faster. Examples of context fields are:
6091 * The **process ID**, **thread ID**, **process name**, and
6092 **process priority** of the thread in which the event occurs.
6093 * The **hostname** of the system on which the event occurs.
6094 * The current values of many possible **performance counters** using
6096 ** CPU cycles, stalled cycles, idle cycles, and the other cycle types.
6098 ** Branch instructions, misses, and loads.
6100 * Any context defined at the application level (supported for the
6101 JUL and log4j <<domain,tracing domains>>).
6103 To get the full list of available context fields, see
6104 `lttng add-context --list`. Some context fields are reserved for a
6105 specific <<domain,tracing domain>> (Linux kernel or user space).
6107 You add context fields to <<channel,channels>>. All the events
6108 that a channel with added context fields records contain those fields.
6110 To add context fields to one or all the channels of a given tracing
6113 * Use the man:lttng-add-context(1) command.
6115 .Add context fields to all the channels of the current tracing session.
6117 The following command line adds the virtual process identifier and
6118 the per-thread CPU cycles count fields to all the user space channels
6119 of the current tracing session.
6123 $ lttng add-context --userspace --type=vpid --type=perf:thread:cpu-cycles
6127 .Add performance counter context fields by raw ID
6129 See man:lttng-add-context(1) for the exact format of the context field
6130 type, which is partly compatible with the format used in
6135 $ lttng add-context --userspace --type=perf:thread:raw:r0110:test
6136 $ lttng add-context --kernel --type=perf:cpu:raw:r0013c:x86unhalted
6140 .Add a context field to a specific channel.
6142 The following command line adds the thread identifier context field
6143 to the Linux kernel channel named `my-channel` in the current
6148 $ lttng add-context --kernel --channel=my-channel --type=tid
6152 .Add an application-specific context field to a specific channel.
6154 The following command line adds the `cur_msg_id` context field of the
6155 `retriever` context retriever for all the instrumented
6156 <<java-application,Java applications>> recording <<event,event records>>
6157 in the channel named `my-channel`:
6161 $ lttng add-context --kernel --channel=my-channel \
6162 --type='$app:retriever:cur_msg_id'
6165 IMPORTANT: Make sure to always quote the `$` character when you
6166 use man:lttng-add-context(1) from a shell.
6169 NOTE: You cannot remove context fields from a channel once you add it.
6174 === Track process IDs
6176 It's often useful to allow only specific process IDs (PIDs) to emit
6177 events. For example, you may wish to record all the system calls made by
6178 a given process (Ă la http://linux.die.net/man/1/strace[strace]).
6180 The man:lttng-track(1) and man:lttng-untrack(1) commands serve this
6181 purpose. Both commands operate on a whitelist of process IDs. You _add_
6182 entries to this whitelist with the man:lttng-track(1) command and remove
6183 entries with the man:lttng-untrack(1) command. Any process which has one
6184 of the PIDs in the whitelist is allowed to emit LTTng events which pass
6185 an enabled <<event,event rule>>.
6187 NOTE: The PID tracker tracks the _numeric process IDs_. Should a
6188 process with a given tracked ID exit and another process be given this
6189 ID, then the latter would also be allowed to emit events.
6191 .Track and untrack process IDs.
6193 For the sake of the following example, assume the target system has 16
6197 <<creating-destroying-tracing-sessions,create a tracing session>>,
6198 the whitelist contains all the possible PIDs:
6201 .All PIDs are tracked.
6202 image::track-all.png[]
6204 When the whitelist is full and you use the man:lttng-track(1) command to
6205 specify some PIDs to track, LTTng first clears the whitelist, then it
6206 tracks the specific PIDs. After:
6210 $ lttng track --pid=3,4,7,10,13
6216 .PIDs 3, 4, 7, 10, and 13 are tracked.
6217 image::track-3-4-7-10-13.png[]
6219 You can add more PIDs to the whitelist afterwards:
6223 $ lttng track --pid=1,15,16
6229 .PIDs 1, 15, and 16 are added to the whitelist.
6230 image::track-1-3-4-7-10-13-15-16.png[]
6232 The man:lttng-untrack(1) command removes entries from the PID tracker's
6233 whitelist. Given the previous example, the following command:
6237 $ lttng untrack --pid=3,7,10,13
6240 leads to this whitelist:
6243 .PIDs 3, 7, 10, and 13 are removed from the whitelist.
6244 image::track-1-4-15-16.png[]
6246 LTTng can track all possible PIDs again using the opt:track(1):--all
6251 $ lttng track --pid --all
6254 The result is, again:
6257 .All PIDs are tracked.
6258 image::track-all.png[]
6261 .Track only specific PIDs
6263 A very typical use case with PID tracking is to start with an empty
6264 whitelist, then <<basic-tracing-session-control,start the tracers>>, and
6265 then add PIDs manually while tracers are active. You can accomplish this
6266 by using the opt:lttng-untrack(1):--all option of the
6267 man:lttng-untrack(1) command to clear the whitelist after you
6268 <<creating-destroying-tracing-sessions,create a tracing session>>:
6272 $ lttng untrack --pid --all
6278 .No PIDs are tracked.
6279 image::untrack-all.png[]
6281 If you trace with this whitelist configuration, the tracer records no
6282 events for this <<domain,tracing domain>> because no processes are
6283 tracked. You can use the man:lttng-track(1) command as usual to track
6284 specific PIDs, for example:
6288 $ lttng track --pid=6,11
6294 .PIDs 6 and 11 are tracked.
6295 image::track-6-11.png[]
6300 [[saving-loading-tracing-session]]
6301 === Save and load tracing session configurations
6303 Configuring a <<tracing-session,tracing session>> can be long. Some of
6304 the tasks involved are:
6306 * <<enabling-disabling-channels,Create channels>> with
6307 specific attributes.
6308 * <<adding-context,Add context fields>> to specific channels.
6309 * <<enabling-disabling-events,Create event rules>> with specific log
6310 level and filter conditions.
6312 If you use LTTng to solve real world problems, chances are you have to
6313 record events using the same tracing session setup over and over,
6314 modifying a few variables each time in your instrumented program
6315 or environment. To avoid constant tracing session reconfiguration,
6316 the man:lttng(1) command-line tool can save and load tracing session
6317 configurations to/from XML files.
6319 To save a given tracing session configuration:
6321 * Use the man:lttng-save(1) command:
6326 $ lttng save my-session
6330 Replace `my-session` with the name of the tracing session to save.
6332 LTTng saves tracing session configurations to
6333 dir:{$LTTNG_HOME/.lttng/sessions} by default. Note that the
6334 env:LTTNG_HOME environment variable defaults to `$HOME` if not set. Use
6335 the opt:lttng-save(1):--output-path option to change this destination
6338 LTTng saves all configuration parameters, for example:
6340 * The tracing session name.
6341 * The trace data output path.
6342 * The channels with their state and all their attributes.
6343 * The context fields you added to channels.
6344 * The event rules with their state, log level and filter conditions.
6346 To load a tracing session:
6348 * Use the man:lttng-load(1) command:
6353 $ lttng load my-session
6357 Replace `my-session` with the name of the tracing session to load.
6359 When LTTng loads a configuration, it restores your saved tracing session
6360 as if you just configured it manually.
6362 See man:lttng(1) for the complete list of command-line options. You
6363 can also save and load all many sessions at a time, and decide in which
6364 directory to output the XML files.
6367 [[sending-trace-data-over-the-network]]
6368 === Send trace data over the network
6370 LTTng can send the recorded trace data to a remote system over the
6371 network instead of writing it to the local file system.
6373 To send the trace data over the network:
6375 . On the _remote_ system (which can also be the target system),
6376 start an LTTng <<lttng-relayd,relay daemon>> (man:lttng-relayd(8)):
6385 . On the _target_ system, create a tracing session configured to
6386 send trace data over the network:
6391 $ lttng create my-session --set-url=net://remote-system
6395 Replace `remote-system` by the host name or IP address of the
6396 remote system. See man:lttng-create(1) for the exact URL format.
6398 . On the target system, use the man:lttng(1) command-line tool as usual.
6399 When tracing is active, the target's consumer daemon sends sub-buffers
6400 to the relay daemon running on the remote system instead of flushing
6401 them to the local file system. The relay daemon writes the received
6402 packets to the local file system.
6404 The relay daemon writes trace files to
6405 +$LTTNG_HOME/lttng-traces/__hostname__/__session__+ by default, where
6406 +__hostname__+ is the host name of the target system and +__session__+
6407 is the tracing session name. Note that the env:LTTNG_HOME environment
6408 variable defaults to `$HOME` if not set. Use the
6409 opt:lttng-relayd(8):--output option of man:lttng-relayd(8) to write
6410 trace files to another base directory.
6415 === View events as LTTng emits them (noch:{LTTng} live)
6417 LTTng live is a network protocol implemented by the <<lttng-relayd,relay
6418 daemon>> (man:lttng-relayd(8)) to allow compatible trace viewers to
6419 display events as LTTng emits them on the target system while tracing is
6422 The relay daemon creates a _tee_: it forwards the trace data to both
6423 the local file system and to connected live viewers:
6426 .The relay daemon creates a _tee_, forwarding the trace data to both trace files and a connected live viewer.
6431 . On the _target system_, create a <<tracing-session,tracing session>>
6437 $ lttng create my-session --live
6441 This spawns a local relay daemon.
6443 . Start the live viewer and configure it to connect to the relay
6444 daemon. For example, with http://diamon.org/babeltrace[Babeltrace]:
6449 $ babeltrace --input-format=lttng-live \
6450 net://localhost/host/hostname/my-session
6457 * `hostname` with the host name of the target system.
6458 * `my-session` with the name of the tracing session to view.
6461 . Configure the tracing session as usual with the man:lttng(1)
6462 command-line tool, and <<basic-tracing-session-control,start tracing>>.
6464 You can list the available live tracing sessions with Babeltrace:
6468 $ babeltrace --input-format=lttng-live net://localhost
6471 You can start the relay daemon on another system. In this case, you need
6472 to specify the relay daemon's URL when you create the tracing session
6473 with the opt:lttng-create(1):--set-url option. You also need to replace
6474 `localhost` in the procedure above with the host name of the system on
6475 which the relay daemon is running.
6477 See man:lttng-create(1) and man:lttng-relayd(8) for the complete list of
6478 command-line options.
6482 [[taking-a-snapshot]]
6483 === Take a snapshot of the current sub-buffers of a tracing session
6485 The normal behavior of LTTng is to append full sub-buffers to growing
6486 trace data files. This is ideal to keep a full history of the events
6487 that occurred on the target system, but it can
6488 represent too much data in some situations. For example, you may wish
6489 to trace your application continuously until some critical situation
6490 happens, in which case you only need the latest few recorded
6491 events to perform the desired analysis, not multi-gigabyte trace files.
6493 With the man:lttng-snapshot(1) command, you can take a snapshot of the
6494 current sub-buffers of a given <<tracing-session,tracing session>>.
6495 LTTng can write the snapshot to the local file system or send it over
6500 . Create a tracing session in _snapshot mode_:
6505 $ lttng create my-session --snapshot
6509 The <<channel-overwrite-mode-vs-discard-mode,event loss mode>> of
6510 <<channel,channels>> created in this mode is automatically set to
6511 _overwrite_ (flight recorder mode).
6513 . Configure the tracing session as usual with the man:lttng(1)
6514 command-line tool, and <<basic-tracing-session-control,start tracing>>.
6516 . **Optional**: When you need to take a snapshot,
6517 <<basic-tracing-session-control,stop tracing>>.
6519 You can take a snapshot when the tracers are active, but if you stop
6520 them first, you are sure that the data in the sub-buffers does not
6521 change before you actually take the snapshot.
6528 $ lttng snapshot record --name=my-first-snapshot
6532 LTTng writes the current sub-buffers of all the current tracing
6533 session's channels to trace files on the local file system. Those trace
6534 files have `my-first-snapshot` in their name.
6536 There is no difference between the format of a normal trace file and the
6537 format of a snapshot: viewers of LTTng traces also support LTTng
6540 By default, LTTng writes snapshot files to the path shown by
6541 `lttng snapshot list-output`. You can change this path or decide to send
6542 snapshots over the network using either:
6544 . An output path or URL that you specify when you create the
6546 . An snapshot output path or URL that you add using
6547 `lttng snapshot add-output`
6548 . An output path or URL that you provide directly to the
6549 `lttng snapshot record` command.
6551 Method 3 overrides method 2, which overrides method 1. When you
6552 specify a URL, a relay daemon must listen on a remote system (see
6553 <<sending-trace-data-over-the-network,Send trace data over the network>>).
6558 === Use the machine interface
6560 With any command of the man:lttng(1) command-line tool, you can set the
6561 opt:lttng(1):--mi option to `xml` (before the command name) to get an
6562 XML machine interface output, for example:
6566 $ lttng --mi=xml enable-event --kernel --syscall open
6569 A schema definition (XSD) is
6570 https://github.com/lttng/lttng-tools/blob/stable-2.10/src/common/mi-lttng-3.0.xsd[available]
6571 to ease the integration with external tools as much as possible.
6575 [[metadata-regenerate]]
6576 === Regenerate the metadata of an LTTng trace
6578 An LTTng trace, which is a http://diamon.org/ctf[CTF] trace, has both
6579 data stream files and a metadata file. This metadata file contains,
6580 amongst other things, information about the offset of the clock sources
6581 used to timestamp <<event,event records>> when tracing.
6583 If, once a <<tracing-session,tracing session>> is
6584 <<basic-tracing-session-control,started>>, a major
6585 https://en.wikipedia.org/wiki/Network_Time_Protocol[NTP] correction
6586 happens, the trace's clock offset also needs to be updated. You
6587 can use the `metadata` item of the man:lttng-regenerate(1) command
6590 The main use case of this command is to allow a system to boot with
6591 an incorrect wall time and trace it with LTTng before its wall time
6592 is corrected. Once the system is known to be in a state where its
6593 wall time is correct, it can run `lttng regenerate metadata`.
6595 To regenerate the metadata of an LTTng trace:
6597 * Use the `metadata` item of the man:lttng-regenerate(1) command:
6602 $ lttng regenerate metadata
6608 `lttng regenerate metadata` has the following limitations:
6610 * Tracing session <<creating-destroying-tracing-sessions,created>>
6612 * User space <<channel,channels>>, if any, are using
6613 <<channel-buffering-schemes,per-user buffering>>.
6618 [[regenerate-statedump]]
6619 === Regenerate the state dump of a tracing session
6621 The LTTng kernel and user space tracers generate state dump
6622 <<event,event records>> when the application starts or when you
6623 <<basic-tracing-session-control,start a tracing session>>. An analysis
6624 can use the state dump event records to set an initial state before it
6625 builds the rest of the state from the following event records.
6626 http://tracecompass.org/[Trace Compass] is a notable example of an
6627 application which uses the state dump of an LTTng trace.
6629 When you <<taking-a-snapshot,take a snapshot>>, it's possible that the
6630 state dump event records are not included in the snapshot because they
6631 were recorded to a sub-buffer that has been consumed or overwritten
6634 You can use the `lttng regenerate statedump` command to emit the state
6635 dump event records again.
6637 To regenerate the state dump of the current tracing session, provided
6638 create it in snapshot mode, before you take a snapshot:
6640 . Use the `statedump` item of the man:lttng-regenerate(1) command:
6645 $ lttng regenerate statedump
6649 . <<basic-tracing-session-control,Stop the tracing session>>:
6658 . <<taking-a-snapshot,Take a snapshot>>:
6663 $ lttng snapshot record --name=my-snapshot
6667 Depending on the event throughput, you should run steps 1 and 2
6668 as closely as possible.
6670 NOTE: To record the state dump events, you need to
6671 <<enabling-disabling-events,create event rules>> which enable them.
6672 LTTng-UST state dump tracepoints start with `lttng_ust_statedump:`.
6673 LTTng-modules state dump tracepoints start with `lttng_statedump_`.
6677 [[persistent-memory-file-systems]]
6678 === Record trace data on persistent memory file systems
6680 https://en.wikipedia.org/wiki/Non-volatile_random-access_memory[Non-volatile random-access memory]
6681 (NVRAM) is random-access memory that retains its information when power
6682 is turned off (non-volatile). Systems with such memory can store data
6683 structures in RAM and retrieve them after a reboot, without flushing
6684 to typical _storage_.
6686 Linux supports NVRAM file systems thanks to either
6687 http://pramfs.sourceforge.net/[PRAMFS] or
6688 https://www.kernel.org/doc/Documentation/filesystems/dax.txt[DAX]{nbsp}+{nbsp}http://lkml.iu.edu/hypermail/linux/kernel/1504.1/03463.html[pmem]
6689 (requires Linux 4.1+).
6691 This section does not describe how to operate such file systems;
6692 we assume that you have a working persistent memory file system.
6694 When you create a <<tracing-session,tracing session>>, you can specify
6695 the path of the shared memory holding the sub-buffers. If you specify a
6696 location on an NVRAM file system, then you can retrieve the latest
6697 recorded trace data when the system reboots after a crash.
6699 To record trace data on a persistent memory file system and retrieve the
6700 trace data after a system crash:
6702 . Create a tracing session with a sub-buffer shared memory path located
6703 on an NVRAM file system:
6708 $ lttng create my-session --shm-path=/path/to/shm
6712 . Configure the tracing session as usual with the man:lttng(1)
6713 command-line tool, and <<basic-tracing-session-control,start tracing>>.
6715 . After a system crash, use the man:lttng-crash(1) command-line tool to
6716 view the trace data recorded on the NVRAM file system:
6721 $ lttng-crash /path/to/shm
6725 The binary layout of the ring buffer files is not exactly the same as
6726 the trace files layout. This is why you need to use man:lttng-crash(1)
6727 instead of your preferred trace viewer directly.
6729 To convert the ring buffer files to LTTng trace files:
6731 * Use the opt:lttng-crash(1):--extract option of man:lttng-crash(1):
6736 $ lttng-crash --extract=/path/to/trace /path/to/shm
6742 [[notif-trigger-api]]
6743 === Get notified when a channel's buffer usage is too high or too low
6745 With LTTng's $$C/C++$$ notification and trigger API, your user
6746 application can get notified when the buffer usage of one or more
6747 <<channel,channels>> becomes too low or too high. You can use this API
6748 and enable or disable <<event,event rules>> during tracing to avoid
6749 <<channel-overwrite-mode-vs-discard-mode,discarded event records>>.
6751 .Have a user application get notified when an LTTng channel's buffer usage is too high.
6753 In this example, we create and build an application which gets notified
6754 when the buffer usage of a specific LTTng channel is higher than
6755 75{nbsp}%. We only print that it is the case in the example, but we
6756 could as well use the API of <<liblttng-ctl-lttng,`liblttng-ctl`>> to
6757 disable event rules when this happens.
6759 . Create the application's C source file:
6767 #include <lttng/domain.h>
6768 #include <lttng/action/action.h>
6769 #include <lttng/action/notify.h>
6770 #include <lttng/condition/condition.h>
6771 #include <lttng/condition/buffer-usage.h>
6772 #include <lttng/condition/evaluation.h>
6773 #include <lttng/notification/channel.h>
6774 #include <lttng/notification/notification.h>
6775 #include <lttng/trigger/trigger.h>
6776 #include <lttng/endpoint.h>
6778 int main(int argc, char *argv[])
6780 int exit_status = 0;
6781 struct lttng_notification_channel *notification_channel;
6782 struct lttng_condition *condition;
6783 struct lttng_action *action;
6784 struct lttng_trigger *trigger;
6785 const char *tracing_session_name;
6786 const char *channel_name;
6789 tracing_session_name = argv[1];
6790 channel_name = argv[2];
6793 * Create a notification channel. A notification channel
6794 * connects the user application to the LTTng session daemon.
6795 * This notification channel can be used to listen for various
6796 * types of notifications.
6798 notification_channel = lttng_notification_channel_create(
6799 lttng_session_daemon_notification_endpoint);
6802 * Create a "high buffer usage" condition. In this case, the
6803 * condition is reached when the buffer usage is greater than or
6804 * equal to 75 %. We create the condition for a specific session
6805 * name, channel name, and for the user space tracing domain.
6807 * The "low buffer usage" condition type also exists.
6809 condition = lttng_condition_buffer_usage_high_create();
6810 lttng_condition_buffer_usage_set_threshold_ratio(condition, .75);
6811 lttng_condition_buffer_usage_set_session_name(
6812 condition, tracing_session_name);
6813 lttng_condition_buffer_usage_set_channel_name(condition,
6815 lttng_condition_buffer_usage_set_domain_type(condition,
6819 * Create an action (get a notification) to take when the
6820 * condition created above is reached.
6822 action = lttng_action_notify_create();
6825 * Create a trigger. A trigger associates a condition to an
6826 * action: the action is executed when the condition is reached.
6828 trigger = lttng_trigger_create(condition, action);
6830 /* Register the trigger to LTTng. */
6831 lttng_register_trigger(trigger);
6834 * Now that we have registered a trigger, a notification will be
6835 * emitted everytime its condition is met. To receive this
6836 * notification, we must subscribe to notifications that match
6837 * the same condition.
6839 lttng_notification_channel_subscribe(notification_channel, condition);
6842 * Notification loop. This can be in a dedicated thread to avoid
6843 * blocking the main thread.
6846 struct lttng_notification *notification;
6847 enum lttng_notification_channel_status status;
6848 const struct lttng_evaluation *notification_evaluation;
6849 const struct lttng_condition *notification_condition;
6850 double buffer_usage;
6852 /* Receive the next notification. */
6853 status = lttng_notification_channel_get_next_notification(
6854 notification_channel,
6858 case LTTNG_NOTIFICATION_CHANNEL_STATUS_OK:
6860 case LTTNG_NOTIFICATION_CHANNEL_STATUS_NOTIFICATIONS_DROPPED:
6862 * The session daemon can drop notifications if
6863 * a monitoring application is not consuming the
6864 * notifications fast enough.
6867 case LTTNG_NOTIFICATION_CHANNEL_STATUS_CLOSED:
6869 * The notification channel has been closed by the
6870 * session daemon. This is typically caused by a session
6871 * daemon shutting down (cleanly or because of a crash).
6875 /* Unhandled conditions or errors. */
6881 * A notification provides, amongst other things:
6883 * * The condition that caused this notification to be
6885 * * The condition evaluation, which provides more
6886 * specific information on the evaluation of the
6889 * The condition evaluation provides the buffer usage
6890 * value at the moment the condition was met.
6892 notification_condition = lttng_notification_get_condition(
6894 notification_evaluation = lttng_notification_get_evaluation(
6897 /* We're subscribed to only one condition. */
6898 assert(lttng_condition_get_type(notification_condition) ==
6899 LTTNG_CONDITION_TYPE_BUFFER_USAGE_HIGH);
6902 * Get the exact sampled buffer usage from the
6903 * condition evaluation.
6905 lttng_evaluation_buffer_usage_get_usage_ratio(
6906 notification_evaluation, &buffer_usage);
6909 * At this point, instead of printing a message, we
6910 * could do something to reduce the channel's buffer
6911 * usage, like disable specific events.
6913 printf("Buffer usage is %f %% in tracing session \"%s\", "
6914 "user space channel \"%s\".\n",
6915 buffer_usage * 100, tracing_session_name,
6917 lttng_notification_destroy(notification);
6921 lttng_action_destroy(action);
6922 lttng_condition_destroy(condition);
6923 lttng_trigger_destroy(trigger);
6924 lttng_notification_channel_destroy(notification_channel);
6930 . Build the `notif-app` application, linking it to `liblttng-ctl`:
6935 $ gcc -o notif-app notif-app.c -llttng-ctl
6939 . <<creating-destroying-tracing-sessions,Create a tracing session>>,
6940 <<enabling-disabling-events,create an event rule>> matching all the
6941 user space tracepoints, and
6942 <<basic-tracing-session-control,start tracing>>:
6947 $ lttng create my-session
6948 $ lttng enable-event --userspace --all
6953 If you create the channel manually with the man:lttng-enable-channel(1)
6954 command, you can control how frequently are the current values of the
6955 channel's properties sampled to evaluate user conditions with the
6956 opt:lttng-enable-channel(1):--monitor-timer option.
6958 . Run the `notif-app` application. This program accepts the
6959 <<tracing-session,tracing session>> name and the user space channel
6960 name as its two first arguments. The channel which LTTng automatically
6961 creates with the man:lttng-enable-event(1) command above is named
6967 $ ./notif-app my-session channel0
6971 . In another terminal, run an application with a very high event
6972 throughput so that the 75{nbsp}% buffer usage condition is reached.
6974 In the first terminal, the application should print lines like this:
6977 Buffer usage is 81.45197 % in tracing session "my-session", user space
6981 If you don't see anything, try modifying the condition in
6982 path:{notif-app.c} to a lower value (0.1, for example), rebuilding it
6983 (step 2) and running it again (step 4).
6990 [[lttng-modules-ref]]
6991 === noch:{LTTng-modules}
6995 [[lttng-tracepoint-enum]]
6996 ==== `LTTNG_TRACEPOINT_ENUM()` usage
6998 Use the `LTTNG_TRACEPOINT_ENUM()` macro to define an enumeration:
7002 LTTNG_TRACEPOINT_ENUM(name, TP_ENUM_VALUES(entries))
7007 * `name` with the name of the enumeration (C identifier, unique
7008 amongst all the defined enumerations).
7009 * `entries` with a list of enumeration entries.
7011 The available enumeration entry macros are:
7013 +ctf_enum_value(__name__, __value__)+::
7014 Entry named +__name__+ mapped to the integral value +__value__+.
7016 +ctf_enum_range(__name__, __begin__, __end__)+::
7017 Entry named +__name__+ mapped to the range of integral values between
7018 +__begin__+ (included) and +__end__+ (included).
7020 +ctf_enum_auto(__name__)+::
7021 Entry named +__name__+ mapped to the integral value following the
7022 last mapping's value.
7024 The last value of a `ctf_enum_value()` entry is its +__value__+
7027 The last value of a `ctf_enum_range()` entry is its +__end__+ parameter.
7029 If `ctf_enum_auto()` is the first entry in the list, its integral
7032 Use the `ctf_enum()` <<lttng-modules-tp-fields,field definition macro>>
7033 to use a defined enumeration as a tracepoint field.
7035 .Define an enumeration with `LTTNG_TRACEPOINT_ENUM()`.
7039 LTTNG_TRACEPOINT_ENUM(
7042 ctf_enum_auto("AUTO: EXPECT 0")
7043 ctf_enum_value("VALUE: 23", 23)
7044 ctf_enum_value("VALUE: 27", 27)
7045 ctf_enum_auto("AUTO: EXPECT 28")
7046 ctf_enum_range("RANGE: 101 TO 303", 101, 303)
7047 ctf_enum_auto("AUTO: EXPECT 304")
7055 [[lttng-modules-tp-fields]]
7056 ==== Tracepoint fields macros (for `TP_FIELDS()`)
7058 [[tp-fast-assign]][[tp-struct-entry]]The available macros to define
7059 tracepoint fields, which must be listed within `TP_FIELDS()` in
7060 `LTTNG_TRACEPOINT_EVENT()`, are:
7062 [role="func-desc growable",cols="asciidoc,asciidoc"]
7063 .Available macros to define LTTng-modules tracepoint fields
7065 |Macro |Description and parameters
7068 +ctf_integer(__t__, __n__, __e__)+
7070 +ctf_integer_nowrite(__t__, __n__, __e__)+
7072 +ctf_user_integer(__t__, __n__, __e__)+
7074 +ctf_user_integer_nowrite(__t__, __n__, __e__)+
7076 Standard integer, displayed in base 10.
7079 Integer C type (`int`, `long`, `size_t`, ...).
7085 Argument expression.
7088 +ctf_integer_hex(__t__, __n__, __e__)+
7090 +ctf_user_integer_hex(__t__, __n__, __e__)+
7092 Standard integer, displayed in base 16.
7101 Argument expression.
7103 |+ctf_integer_oct(__t__, __n__, __e__)+
7105 Standard integer, displayed in base 8.
7114 Argument expression.
7117 +ctf_integer_network(__t__, __n__, __e__)+
7119 +ctf_user_integer_network(__t__, __n__, __e__)+
7121 Integer in network byte order (big-endian), displayed in base 10.
7130 Argument expression.
7133 +ctf_integer_network_hex(__t__, __n__, __e__)+
7135 +ctf_user_integer_network_hex(__t__, __n__, __e__)+
7137 Integer in network byte order, displayed in base 16.
7146 Argument expression.
7149 +ctf_enum(__N__, __t__, __n__, __e__)+
7151 +ctf_enum_nowrite(__N__, __t__, __n__, __e__)+
7153 +ctf_user_enum(__N__, __t__, __n__, __e__)+
7155 +ctf_user_enum_nowrite(__N__, __t__, __n__, __e__)+
7160 Name of a <<lttng-tracepoint-enum,previously defined enumeration>>.
7163 Integer C type (`int`, `long`, `size_t`, ...).
7169 Argument expression.
7172 +ctf_string(__n__, __e__)+
7174 +ctf_string_nowrite(__n__, __e__)+
7176 +ctf_user_string(__n__, __e__)+
7178 +ctf_user_string_nowrite(__n__, __e__)+
7180 Null-terminated string; undefined behavior if +__e__+ is `NULL`.
7186 Argument expression.
7189 +ctf_array(__t__, __n__, __e__, __s__)+
7191 +ctf_array_nowrite(__t__, __n__, __e__, __s__)+
7193 +ctf_user_array(__t__, __n__, __e__, __s__)+
7195 +ctf_user_array_nowrite(__t__, __n__, __e__, __s__)+
7197 Statically-sized array of integers.
7200 Array element C type.
7206 Argument expression.
7212 +ctf_array_bitfield(__t__, __n__, __e__, __s__)+
7214 +ctf_array_bitfield_nowrite(__t__, __n__, __e__, __s__)+
7216 +ctf_user_array_bitfield(__t__, __n__, __e__, __s__)+
7218 +ctf_user_array_bitfield_nowrite(__t__, __n__, __e__, __s__)+
7220 Statically-sized array of bits.
7222 The type of +__e__+ must be an integer type. +__s__+ is the number
7223 of elements of such type in +__e__+, not the number of bits.
7226 Array element C type.
7232 Argument expression.
7238 +ctf_array_text(__t__, __n__, __e__, __s__)+
7240 +ctf_array_text_nowrite(__t__, __n__, __e__, __s__)+
7242 +ctf_user_array_text(__t__, __n__, __e__, __s__)+
7244 +ctf_user_array_text_nowrite(__t__, __n__, __e__, __s__)+
7246 Statically-sized array, printed as text.
7248 The string does not need to be null-terminated.
7251 Array element C type (always `char`).
7257 Argument expression.
7263 +ctf_sequence(__t__, __n__, __e__, __T__, __E__)+
7265 +ctf_sequence_nowrite(__t__, __n__, __e__, __T__, __E__)+
7267 +ctf_user_sequence(__t__, __n__, __e__, __T__, __E__)+
7269 +ctf_user_sequence_nowrite(__t__, __n__, __e__, __T__, __E__)+
7271 Dynamically-sized array of integers.
7273 The type of +__E__+ must be unsigned.
7276 Array element C type.
7282 Argument expression.
7285 Length expression C type.
7291 +ctf_sequence_hex(__t__, __n__, __e__, __T__, __E__)+
7293 +ctf_user_sequence_hex(__t__, __n__, __e__, __T__, __E__)+
7295 Dynamically-sized array of integers, displayed in base 16.
7297 The type of +__E__+ must be unsigned.
7300 Array element C type.
7306 Argument expression.
7309 Length expression C type.
7314 |+ctf_sequence_network(__t__, __n__, __e__, __T__, __E__)+
7316 Dynamically-sized array of integers in network byte order (big-endian),
7317 displayed in base 10.
7319 The type of +__E__+ must be unsigned.
7322 Array element C type.
7328 Argument expression.
7331 Length expression C type.
7337 +ctf_sequence_bitfield(__t__, __n__, __e__, __T__, __E__)+
7339 +ctf_sequence_bitfield_nowrite(__t__, __n__, __e__, __T__, __E__)+
7341 +ctf_user_sequence_bitfield(__t__, __n__, __e__, __T__, __E__)+
7343 +ctf_user_sequence_bitfield_nowrite(__t__, __n__, __e__, __T__, __E__)+
7345 Dynamically-sized array of bits.
7347 The type of +__e__+ must be an integer type. +__s__+ is the number
7348 of elements of such type in +__e__+, not the number of bits.
7350 The type of +__E__+ must be unsigned.
7353 Array element C type.
7359 Argument expression.
7362 Length expression C type.
7368 +ctf_sequence_text(__t__, __n__, __e__, __T__, __E__)+
7370 +ctf_sequence_text_nowrite(__t__, __n__, __e__, __T__, __E__)+
7372 +ctf_user_sequence_text(__t__, __n__, __e__, __T__, __E__)+
7374 +ctf_user_sequence_text_nowrite(__t__, __n__, __e__, __T__, __E__)+
7376 Dynamically-sized array, displayed as text.
7378 The string does not need to be null-terminated.
7380 The type of +__E__+ must be unsigned.
7382 The behaviour is undefined if +__e__+ is `NULL`.
7385 Sequence element C type (always `char`).
7391 Argument expression.
7394 Length expression C type.
7400 Use the `_user` versions when the argument expression, `e`, is
7401 a user space address. In the cases of `ctf_user_integer*()` and
7402 `ctf_user_float*()`, `&e` must be a user space address, thus `e` must
7405 The `_nowrite` versions omit themselves from the session trace, but are
7406 otherwise identical. This means the `_nowrite` fields won't be written
7407 in the recorded trace. Their primary purpose is to make some
7408 of the event context available to the
7409 <<enabling-disabling-events,event filters>> without having to
7410 commit the data to sub-buffers.
7416 Terms related to LTTng and to tracing in general:
7419 The http://diamon.org/babeltrace[Babeltrace] project, which includes
7420 the cmd:babeltrace command, some libraries, and Python bindings.
7422 <<channel-buffering-schemes,buffering scheme>>::
7423 A layout of sub-buffers applied to a given channel.
7425 <<channel,channel>>::
7426 An entity which is responsible for a set of ring buffers.
7428 <<event,Event rules>> are always attached to a specific channel.
7431 A reference of time for a tracer.
7433 <<lttng-consumerd,consumer daemon>>::
7434 A process which is responsible for consuming the full sub-buffers
7435 and write them to a file system or send them over the network.
7437 <<channel-overwrite-mode-vs-discard-mode,discard mode>>:: The event loss
7438 mode in which the tracer _discards_ new event records when there's no
7439 sub-buffer space left to store them.
7442 The consequence of the execution of an instrumentation
7443 point, like a tracepoint that you manually place in some source code,
7444 or a Linux kernel KProbe.
7446 An event is said to _occur_ at a specific time. Different actions can
7447 be taken upon the occurrence of an event, like record the event's payload
7450 <<channel-overwrite-mode-vs-discard-mode,event loss mode>>::
7451 The mechanism by which event records of a given channel are lost
7452 (not recorded) when there is no sub-buffer space left to store them.
7454 [[def-event-name]]event name::
7455 The name of an event, which is also the name of the event record.
7456 This is also called the _instrumentation point name_.
7459 A record, in a trace, of the payload of an event which occured.
7461 <<event,event rule>>::
7462 Set of conditions which must be satisfied for one or more occuring
7463 events to be recorded.
7465 `java.util.logging`::
7467 https://docs.oracle.com/javase/7/docs/api/java/util/logging/package-summary.html[core logging facilities].
7469 <<instrumenting,instrumentation>>::
7470 The use of LTTng probes to make a piece of software traceable.
7472 instrumentation point::
7473 A point in the execution path of a piece of software that, when
7474 reached by this execution, can emit an event.
7476 instrumentation point name::
7477 See _<<def-event-name,event name>>_.
7480 A http://logging.apache.org/log4j/1.2/[logging library] for Java
7481 developed by the Apache Software Foundation.
7484 Level of severity of a log statement or user space
7485 instrumentation point.
7488 The _Linux Trace Toolkit: next generation_ project.
7490 <<lttng-cli,cmd:lttng>>::
7491 A command-line tool provided by the LTTng-tools project which you
7492 can use to send and receive control messages to and from a
7496 The https://github.com/lttng/lttng-analyses[LTTng analyses] project,
7497 which is a set of analyzing programs that are used to obtain a
7498 higher level view of an LTTng trace.
7500 cmd:lttng-consumerd::
7501 The name of the consumer daemon program.
7504 A utility provided by the LTTng-tools project which can convert
7505 ring buffer files (usually
7506 <<persistent-memory-file-systems,saved on a persistent memory file system>>)
7509 LTTng Documentation::
7512 <<lttng-live,LTTng live>>::
7513 A communication protocol between the relay daemon and live viewers
7514 which makes it possible to see events "live", as they are received by
7517 <<lttng-modules,LTTng-modules>>::
7518 The https://github.com/lttng/lttng-modules[LTTng-modules] project,
7519 which contains the Linux kernel modules to make the Linux kernel
7520 instrumentation points available for LTTng tracing.
7523 The name of the relay daemon program.
7525 cmd:lttng-sessiond::
7526 The name of the session daemon program.
7529 The https://github.com/lttng/lttng-tools[LTTng-tools] project, which
7530 contains the various programs and libraries used to
7531 <<controlling-tracing,control tracing>>.
7533 <<lttng-ust,LTTng-UST>>::
7534 The https://github.com/lttng/lttng-ust[LTTng-UST] project, which
7535 contains libraries to instrument user applications.
7537 <<lttng-ust-agents,LTTng-UST Java agent>>::
7538 A Java package provided by the LTTng-UST project to allow the
7539 LTTng instrumentation of `java.util.logging` and Apache log4j 1.2
7542 <<lttng-ust-agents,LTTng-UST Python agent>>::
7543 A Python package provided by the LTTng-UST project to allow the
7544 LTTng instrumentation of Python logging statements.
7546 <<channel-overwrite-mode-vs-discard-mode,overwrite mode>>::
7547 The event loss mode in which new event records overwrite older
7548 event records when there's no sub-buffer space left to store them.
7550 <<channel-buffering-schemes,per-process buffering>>::
7551 A buffering scheme in which each instrumented process has its own
7552 sub-buffers for a given user space channel.
7554 <<channel-buffering-schemes,per-user buffering>>::
7555 A buffering scheme in which all the processes of a Unix user share the
7556 same sub-buffer for a given user space channel.
7558 <<lttng-relayd,relay daemon>>::
7559 A process which is responsible for receiving the trace data sent by
7560 a distant consumer daemon.
7563 A set of sub-buffers.
7565 <<lttng-sessiond,session daemon>>::
7566 A process which receives control commands from you and orchestrates
7567 the tracers and various LTTng daemons.
7569 <<taking-a-snapshot,snapshot>>::
7570 A copy of the current data of all the sub-buffers of a given tracing
7571 session, saved as trace files.
7574 One part of an LTTng ring buffer which contains event records.
7577 The time information attached to an event when it is emitted.
7580 A set of files which are the concatenations of one or more
7581 flushed sub-buffers.
7584 The action of recording the events emitted by an application
7585 or by a system, or to initiate such recording by controlling
7589 The http://tracecompass.org[Trace Compass] project and application.
7592 An instrumentation point using the tracepoint mechanism of the Linux
7593 kernel or of LTTng-UST.
7595 tracepoint definition::
7596 The definition of a single tracepoint.
7599 The name of a tracepoint.
7601 tracepoint provider::
7602 A set of functions providing tracepoints to an instrumented user
7605 Not to be confused with a _tracepoint provider package_: many tracepoint
7606 providers can exist within a tracepoint provider package.
7608 tracepoint provider package::
7609 One or more tracepoint providers compiled as an object file or as
7613 A software which records emitted events.
7615 <<domain,tracing domain>>::
7616 A namespace for event sources.
7618 <<tracing-group,tracing group>>::
7619 The Unix group in which a Unix user can be to be allowed to trace the
7622 <<tracing-session,tracing session>>::
7623 A stateful dialogue between you and a <<lttng-sessiond,session
7627 An application running in user space, as opposed to a Linux kernel
7628 module, for example.