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 links and tables.
64 * **<<glossary,Glossary>>** is a specialized dictionary of terms related
65 to LTTng or to the field of software tracing.
68 include::../common/convention.txt[]
71 include::../common/acknowledgements.txt[]
75 == What's new in LTTng {revision}?
77 * **Tracing control**:
78 ** Dynamic filter support for <<event,event rules>> in the Linux kernel
79 <<domain,tracing domain>>. For example:
84 lttng enable-event --kernel irq_handler_entry --filter='irq == 28'
88 ** Wildcard support in the instrumentation point name of an event rule
89 in the Linux kernel tracing domain. For example:
94 lttng enable-event --kernel 'sched_*'
98 ** New `lttng track` and `lttng untrack` commands to make
99 <<pid-tracking,PID tracking>> super-fast for both the Linux kernel
100 and the user space tracing domains.
102 When LTTng _tracks_ one or more PIDs, only the processes having those PIDs
103 can emit events for a given tracing session.
105 ** New `--shm-path` option of the `lttng create` command to specify the
106 path where LTTng creates the shared memory holding the ring buffers.
108 This feature is useful when used with persistent memory file systems to
109 extract the latest recorded trace data in the event of a crash requiring
112 The new man:lttng-crash(1) command-line utility can extract trace data
113 from such a file (see <<persistent-memory-file-systems,Record trace data
114 on persistent memory file systems>>).
116 * **User space tracing**:
117 ** New <<python-application,LTTng-UST Python agent>> which makes it easy
118 to trace existing Python applications that are using the standard
119 https://docs.python.org/3/howto/logging.html[`logging` package].
121 This agent is compatible with both the Python 2 and Python 3 languages.
123 ** New <<tracelog,`tracelog()`>> facility to ease the migration from
126 `tracelog()` is similar to <<tracef,`tracef()`>>,
127 but it accepts an additional log level parameter.
129 ** Plugin support in LTTng-UST to provide a custom clock source and to
130 retrieve the current CPU number.
132 This feature exists for very advanced use cases.
135 https://github.com/lttng/lttng-ust/tree/stable-{revision}/doc/examples/clock-override[clock-override]
137 https://github.com/lttng/lttng-ust/tree/stable-{revision}/doc/examples/getcpu-override[getcpu-override]
138 examples for more details.
140 Moreover, LTTng{nbsp}{revision} boasts great stability, benifiting from
141 piles of bug fixes and more-than-welcome internal refactorings.
143 To learn more about the new features of LTTng{nbsp}{revision}, see
144 https://lttng.org/blog/2015/10/14/lttng-2.7-released/[the release announcement].
150 What is LTTng? As its name suggests, the _Linux Trace Toolkit: next
151 generation_ is a modern toolkit for tracing Linux systems and
152 applications. So your first question might be:
159 As the history of software engineering progressed and led to what
160 we now take for granted--complex, numerous and
161 interdependent software applications running in parallel on
162 sophisticated operating systems like Linux--the authors of such
163 components, software developers, began feeling a natural
164 urge to have tools that would ensure the robustness and good performance
165 of their masterpieces.
167 One major achievement in this field is, inarguably, the
168 https://www.gnu.org/software/gdb/[GNU debugger (GDB)],
169 an essential tool for developers to find and fix bugs. But even the best
170 debugger won't help make your software run faster, and nowadays, faster
171 software means either more work done by the same hardware, or cheaper
172 hardware for the same work.
174 A _profiler_ is often the tool of choice to identify performance
175 bottlenecks. Profiling is suitable to identify _where_ performance is
176 lost in a given software. The profiler outputs a profile, a statistical
177 summary of observed events, which you may use to discover which
178 functions took the most time to execute. However, a profiler won't
179 report _why_ some identified functions are the bottleneck. Bottlenecks
180 might only occur when specific conditions are met, conditions that are
181 sometimes impossible to capture by a statistical profiler, or impossible
182 to reproduce with an application altered by the overhead of an
183 event-based profiler. For a thorough investigation of software
184 performance issues, a history of execution is essential, with the
185 recorded values of variables and context fields you choose, and
186 with as little influence as possible on the instrumented software. This
187 is where tracing comes in handy.
189 _Tracing_ is a technique used to understand what goes on in a running
190 software system. The software used for tracing is called a _tracer_,
191 which is conceptually similar to a tape recorder. When recording,
192 specific instrumentation points placed in the software source code
193 generate events that are saved on a giant tape: a _trace_ file. You
194 can trace user applications and the operating system at the same time,
195 opening the possibility of resolving a wide range of problems that would
196 otherwise be extremely challenging.
198 Tracing is often compared to _logging_. However, tracers and loggers are
199 two different tools, serving two different purposes. Tracers are
200 designed to record much lower-level events that occur much more
201 frequently than log messages, often in the range of thousands per
202 second, with very little execution overhead. Logging is more appropriate
203 for a very high-level analysis of less frequent events: user accesses,
204 exceptional conditions (errors and warnings, for example), database
205 transactions, instant messaging communications, and such. Simply put,
206 logging is one of the many use cases that can be satisfied with tracing.
208 The list of recorded events inside a trace file can be read manually
209 like a log file for the maximum level of detail, but it is generally
210 much more interesting to perform application-specific analyses to
211 produce reduced statistics and graphs that are useful to resolve a
212 given problem. Trace viewers and analyzers are specialized tools
215 In the end, this is what LTTng is: a powerful, open source set of
216 tools to trace the Linux kernel and user applications at the same time.
217 LTTng is composed of several components actively maintained and
218 developed by its link:/community/#where[community].
221 [[lttng-alternatives]]
222 === Alternatives to noch:{LTTng}
224 Excluding proprietary solutions, a few competing software tracers
227 * https://github.com/dtrace4linux/linux[dtrace4linux] is a port of
228 Sun Microsystems's DTrace to Linux. The cmd:dtrace tool interprets
229 user scripts and is responsible for loading code into the
230 Linux kernel for further execution and collecting the outputted data.
231 * https://en.wikipedia.org/wiki/Berkeley_Packet_Filter[eBPF] is a
232 subsystem in the Linux kernel in which a virtual machine can execute
233 programs passed from the user space to the kernel. You can attach
234 such programs to tracepoints and KProbes thanks to a system call, and
235 they can output data to the user space when executed thanks to
236 different mechanisms (pipe, VM register values, and eBPF maps, to name
238 * https://www.kernel.org/doc/Documentation/trace/ftrace.txt[ftrace]
239 is the de facto function tracer of the Linux kernel. Its user
240 interface is a set of special files in sysfs.
241 * https://perf.wiki.kernel.org/[perf] is
242 a performance analyzing tool for Linux which supports hardware
243 performance counters, tracepoints, as well as other counters and
244 types of probes. perf's controlling utility is the cmd:perf command
246 * http://linux.die.net/man/1/strace[strace]
247 is a command-line utility which records system calls made by a
248 user process, as well as signal deliveries and changes of process
249 state. strace makes use of https://en.wikipedia.org/wiki/Ptrace[ptrace]
250 to fulfill its function.
251 * http://www.sysdig.org/[sysdig], like SystemTap, uses scripts to
252 analyze Linux kernel events. You write scripts, or _chisels_ in
253 sysdig's jargon, in Lua and sysdig executes them while the system is
254 being traced or afterwards. sysdig's interface is the cmd:sysdig
255 command-line tool as well as the curses-based cmd:csysdig tool.
256 * https://sourceware.org/systemtap/[SystemTap] is a Linux kernel and
257 user space tracer which uses custom user scripts to produce plain text
258 traces. SystemTap converts the scripts to the C language, and then
259 compiles them as Linux kernel modules which are loaded to produce
260 trace data. SystemTap's primary user interface is the cmd:stap
263 The main distinctive features of LTTng is that it produces correlated
264 kernel and user space traces, as well as doing so with the lowest
265 overhead amongst other solutions. It produces trace files in the
266 http://diamon.org/ctf[CTF] format, a file format optimized
267 for the production and analyses of multi-gigabyte data.
269 LTTng is the result of more than 10 years of active open source
270 development by a community of passionate developers.
271 LTTng{nbsp}{revision} is currently available on major desktop and server
274 The main interface for tracing control is a single command-line tool
275 named cmd:lttng. The latter can create several tracing sessions, enable
276 and disable events on the fly, filter events efficiently with custom
277 user expressions, start and stop tracing, and much more. LTTng can
278 record the traces on the file system or send them over the network, and
279 keep them totally or partially. You can view the traces once tracing
280 becomes inactive or in real-time.
282 <<installing-lttng,Install LTTng now>> and
283 <<getting-started,start tracing>>!
289 **LTTng** is a set of software <<plumbing,components>> which interact to
290 <<instrumenting,instrument>> the Linux kernel and user applications, and
291 to <<controlling-tracing,control tracing>> (start and stop
292 tracing, enable and disable event rules, and the rest). Those
293 components are bundled into the following packages:
295 * **LTTng-tools**: Libraries and command-line interface to
297 * **LTTng-modules**: Linux kernel modules to instrument and
299 * **LTTng-UST**: Libraries and Java/Python packages to instrument and
300 trace user applications.
302 Most distributions mark the LTTng-modules and LTTng-UST packages as
303 optional when installing LTTng-tools (which is always required). In the
304 following sections, we always provide the steps to install all three,
307 * You only need to install LTTng-modules if you intend to trace the
309 * You only need to install LTTng-UST if you intend to trace user
313 .Availability of LTTng{nbsp}{revision} for major Linux distributions.
315 |Distribution |Available in releases |Alternatives
318 |<<ubuntu,Ubuntu{nbsp}16.04 _Xenial Xerus_>>
319 |LTTng{nbsp}2.8 for Ubuntu{nbsp}16.10 _Yakkety Yak_.
321 LTTng{nbsp}{revision} for Ubuntu{nbsp}12.04 _Precise Pangolin_,
322 Ubuntu{nbsp}14.04 _Trusty Tahr_, and Ubuntu{nbsp}16.04 _Xenial Xerus_:
323 <<ubuntu-ppa,use the LTTng Stable{nbsp}{revision} PPA>>.
325 <<building-from-source,Build LTTng{nbsp}{revision} from source>> for
326 other Ubuntu releases.
330 |LTTng{nbsp}{revision} for Fedora{nbsp}25 and Fedora{nbsp}26 (not
333 <<building-from-source,Build LTTng{nbsp}{revision} from source>> for
334 other Fedora releases.
338 |LTTng{nbsp}2.8 for Debian "stretch" (testing).
340 <<building-from-source,Build LTTng{nbsp}{revision} from source>> for
341 other Debian releases.
344 |<<opensuse,openSUSE Leap{nbsp}42.1>>
345 |<<building-from-source,Build LTTng{nbsp}{revision} from source>> for
346 other openSUSE releases.
350 |<<building-from-source,Build LTTng{nbsp}{revision} from source>>.
354 |LTTng{nbsp}2.8 for Alpine Linux "edge".
356 LTTng{nbsp}2.8 for Alpine Linux{nbsp}3.5 (not released yet).
358 <<building-from-source,Build LTTng{nbsp}{revision} from source>> for
359 other Alpine Linux releases.
362 |See http://packages.efficios.com/[EfficiOS Enterprise Packages].
366 |<<"buildroot","Buildroot{nbsp}2016.02, Buildroot{nbsp}2016.05,
367 and Buildroot{nbsp}2016.08">>
368 |LTTng{nbsp}2.8 for Buildroot{nbsp}2016.11 (not released yet).
370 <<building-from-source,Build LTTng{nbsp}{revision} from source>> for
371 other Buildroot releases.
373 |OpenEmbedded and Yocto
374 |<<oe-yocto,`openembedded-core` layer from 1{nbsp}December 2016 until
375 3{nbsp}September 2016>>
376 |LTTng{nbsp}2.8 for OpenEmbedded since 3{nbsp}September 2016.
378 <<building-from-source,Build LTTng{nbsp}{revision} from source>> for
379 other OpenEmbedded releases.
384 === [[ubuntu-official-repositories]]Ubuntu
386 LTTng{nbsp}{revision} is available on Ubuntu 16.04 _Xenial Xerus_. For
387 previous releases of Ubuntu, <<ubuntu-ppa,use the LTTng
388 Stable{nbsp}{revision} PPA>>.
390 To install LTTng{nbsp}{revision} on Ubuntu{nbsp}16.04 _Xenial Xerus_:
392 . Install the main LTTng{nbsp}{revision} packages:
397 sudo apt-get install lttng-tools
398 sudo apt-get install lttng-modules-dkms
399 sudo apt-get install liblttng-ust-dev
403 . **If you need to instrument and trace
404 <<java-application,Java applications>>**, install the LTTng-UST
410 sudo apt-get install liblttng-ust-agent-java
414 . **If you need to instrument and trace
415 <<python-application,Python applications>>**, install the
416 LTTng-UST Python agent:
421 sudo apt-get install python3-lttngust
427 ==== noch:{LTTng} Stable {revision} PPA
430 https://launchpad.net/~lttng/+archive/ubuntu/stable-{revision}[LTTng Stable{nbsp}{revision} PPA]
431 offers the latest stable LTTng{nbsp}{revision} packages for:
433 * Ubuntu{nbsp}12.04 _Precise Pangolin_
434 * Ubuntu{nbsp}14.04 _Trusty Tahr_
435 * Ubuntu{nbsp}16.04 _Xenial Xerus_
437 To install LTTng{nbsp}{revision} from the LTTng Stable{nbsp}{revision}
440 . Add the LTTng Stable{nbsp}{revision} PPA repository and update the
446 sudo apt-add-repository ppa:lttng/stable-2.7
451 . Install the main LTTng{nbsp}{revision} packages:
456 sudo apt-get install lttng-tools
457 sudo apt-get install lttng-modules-dkms
458 sudo apt-get install liblttng-ust-dev
462 . **If you need to instrument and trace
463 <<java-application,Java applications>>**, install the LTTng-UST
469 sudo apt-get install liblttng-ust-agent-java
473 . **If you need to instrument and trace
474 <<python-application,Python applications>>**, install the
475 LTTng-UST Python agent:
480 sudo apt-get install python3-lttngust
486 === noch:{openSUSE}/RPM
488 To install LTTng{nbsp}{revision} on openSUSE Leap{nbsp}42.1:
490 * Install the main LTTng{nbsp}{revision} packages:
495 sudo zypper install lttng-tools
496 sudo zypper install lttng-modules
497 sudo zypper install lttng-ust-devel
502 .Java and Python application instrumentation and tracing
504 If you need to instrument and trace <<java-application,Java
505 applications>> on openSUSE, you need to build and install
506 LTTng-UST{nbsp}{revision} <<building-from-source,from source>> and pass
507 the `--enable-java-agent-jul`, `--enable-java-agent-log4j`, or
508 `--enable-java-agent-all` options to the `configure` script, depending
509 on which Java logging framework you use.
511 If you need to instrument and trace <<python-application,Python
512 applications>> on openSUSE, you need to build and install
513 LTTng-UST{nbsp}{revision} from source and pass the
514 `--enable-python-agent` option to the `configure` script.
521 To install LTTng{nbsp}{revision} on Buildroot{nbsp}2016.02,
522 Buildroot{nbsp}2016.05, or Buildroot{nbsp}2016.08:
524 . Launch the Buildroot configuration tool:
533 . In **Kernel**, check **Linux kernel**.
534 . In **Toolchain**, check **Enable WCHAR support**.
535 . In **Target packages**{nbsp}→ **Debugging, profiling and benchmark**,
536 check **lttng-modules** and **lttng-tools**.
537 . In **Target packages**{nbsp}→ **Libraries**{nbsp}→
538 **Other**, check **lttng-libust**.
542 === OpenEmbedded and Yocto
544 LTTng{nbsp}{revision} recipes are available in the
545 http://layers.openembedded.org/layerindex/branch/master/layer/openembedded-core/[`openembedded-core`]
546 layer of OpenEmbedded since 1{nbsp}December 2016 until
547 3{nbsp}September 2016 under the following names:
553 With BitBake, the simplest way to include LTTng recipes in your target
554 image is to add them to `IMAGE_INSTALL_append` in path:{conf/local.conf}:
557 IMAGE_INSTALL_append = " lttng-tools lttng-modules lttng-ust"
562 . Select a machine and an image recipe.
563 . Click **Edit image recipe**.
564 . Under the **All recipes** tab, search for **lttng**.
565 . Check the desired LTTng recipes.
568 .Java and Python application instrumentation and tracing
570 If you need to instrument and trace <<java-application,Java
571 applications>> on openSUSE, you need to build and install
572 LTTng-UST{nbsp}{revision} <<building-from-source,from source>> and pass
573 the `--enable-java-agent-jul`, `--enable-java-agent-log4j`, or
574 `--enable-java-agent-all` options to the `configure` script, depending
575 on which Java logging framework you use.
577 If you need to instrument and trace <<python-application,Python
578 applications>> on openSUSE, you need to build and install
579 LTTng-UST{nbsp}{revision} from source and pass the
580 `--enable-python-agent` option to the `configure` script.
584 [[enterprise-distributions]]
585 === RHEL, SUSE, and other enterprise distributions
587 To install LTTng on enterprise Linux distributions, such as Red Hat
588 Enterprise Linux (RHEL) and SUSE Linux Enterprise Server (SUSE), please
589 see http://packages.efficios.com/[EfficiOS Enterprise Packages].
592 [[building-from-source]]
593 === Build from source
595 To build and install LTTng{nbsp}{revision} from source:
597 . Using your distribution's package manager, or from source, install
598 the following dependencies of LTTng-tools and LTTng-UST:
601 * https://sourceforge.net/projects/libuuid/[libuuid]
602 * http://directory.fsf.org/wiki/Popt[popt]
603 * http://liburcu.org/[Userspace RCU]
604 * http://www.xmlsoft.org/[libxml2]
607 . Download, build, and install the latest LTTng-modules{nbsp}{revision}:
613 wget http://lttng.org/files/lttng-modules/lttng-modules-latest-2.7.tar.bz2 &&
614 tar -xf lttng-modules-latest-2.7.tar.bz2 &&
615 cd lttng-modules-2.7.* &&
617 sudo make modules_install &&
622 . Download, build, and install the latest LTTng-UST{nbsp}{revision}:
628 wget http://lttng.org/files/lttng-ust/lttng-ust-latest-2.7.tar.bz2 &&
629 tar -xf lttng-ust-latest-2.7.tar.bz2 &&
630 cd lttng-ust-2.7.* &&
640 .Java and Python application tracing
642 If you need to instrument and trace <<java-application,Java
643 applications>>, pass the `--enable-java-agent-jul`,
644 `--enable-java-agent-log4j`, or `--enable-java-agent-all` options to the
645 `configure` script, depending on which Java logging framework you use.
647 If you need to instrument and trace <<python-application,Python
648 applications>>, pass the `--enable-python-agent` option to the
649 `configure` script. You can set the `PYTHON` environment variable to the
650 path to the Python interpreter for which to install the LTTng-UST Python
658 By default, LTTng-UST libraries are installed to
659 dir:{/usr/local/lib}, which is the de facto directory in which to
660 keep self-compiled and third-party libraries.
662 When <<building-tracepoint-providers-and-user-application,linking an
663 instrumented user application with `liblttng-ust`>>:
665 * Append `/usr/local/lib` to the env:LD_LIBRARY_PATH environment
667 * Pass the `-L/usr/local/lib` and `-Wl,-rpath,/usr/local/lib` options to
668 man:gcc(1), man:g++(1), or man:clang(1).
672 . Download, build, and install the latest LTTng-tools{nbsp}{revision}:
678 wget http://lttng.org/files/lttng-tools/lttng-tools-latest-2.7.tar.bz2 &&
679 tar -xf lttng-tools-latest-2.7.tar.bz2 &&
680 cd lttng-tools-2.7.* &&
688 TIP: The https://github.com/eepp/vlttng[vlttng tool] can do all the
689 previous steps automatically for a given version of LTTng and confine
690 the installed files in a specific directory. This can be useful to test
691 LTTng without installing it on your system.
697 This is a short guide to get started quickly with LTTng kernel and user
700 Before you follow this guide, make sure to <<installing-lttng,install>>
703 This tutorial walks you through the steps to:
705 . <<tracing-the-linux-kernel,Trace the Linux kernel>>.
706 . <<tracing-your-own-user-application,Trace a user application>> written
708 . <<viewing-and-analyzing-your-traces,View and analyze the
712 [[tracing-the-linux-kernel]]
713 === Trace the Linux kernel
715 The following command lines start with cmd:sudo because you need root
716 privileges to trace the Linux kernel. You can avoid using cmd:sudo if
717 your Unix user is a member of the <<lttng-sessiond,tracing group>>.
719 . Create a <<tracing-session,tracing session>>:
724 sudo lttng create my-kernel-session
728 . List the available kernel tracepoints and system calls:
737 . Create an <<event,event rule>> which matches the desired event names,
738 for example `sched_switch` and `sched_process_fork`:
743 sudo lttng enable-event --kernel sched_switch,sched_process_fork
747 You can also create an event rule which _matches_ all the Linux kernel
748 tracepoints (this will generate a lot of data when tracing):
753 sudo lttng enable-event --kernel --all
766 . Do some operation on your system for a few seconds. For example,
767 load a website, or list the files of a directory.
768 . Stop tracing and destroy the tracing session:
778 The `destroy` command does not destroy the trace data; it only destroys
779 the state of the tracing session.
781 By default, LTTng saves the traces in
782 +$LTTNG_HOME/lttng-traces/__name__-__date__-__time__+,
783 where +__name__+ is the tracing session name. Note that the
784 env:LTTNG_HOME environment variable defaults to `$HOME` if not set.
786 See <<viewing-and-analyzing-your-traces,View and analyze the
787 recorded events>> to view the recorded events.
790 [[tracing-your-own-user-application]]
791 === Trace a user application
793 This section steps you through a simple example to trace a
794 _Hello world_ program written in C.
796 To create the traceable user application:
798 . Create the tracepoint provider header file, which defines the
799 tracepoints and the events they can generate:
805 #undef TRACEPOINT_PROVIDER
806 #define TRACEPOINT_PROVIDER hello_world
808 #undef TRACEPOINT_INCLUDE
809 #define TRACEPOINT_INCLUDE "./hello-tp.h"
811 #if !defined(_HELLO_TP_H) || defined(TRACEPOINT_HEADER_MULTI_READ)
814 #include <lttng/tracepoint.h>
824 ctf_string(my_string_field, my_string_arg)
825 ctf_integer(int, my_integer_field, my_integer_arg)
829 #endif /* _HELLO_TP_H */
831 #include <lttng/tracepoint-event.h>
835 . Create the tracepoint provider package source file:
841 #define TRACEPOINT_CREATE_PROBES
842 #define TRACEPOINT_DEFINE
844 #include "hello-tp.h"
848 . Build the tracepoint provider package:
853 gcc -c -I. hello-tp.c
857 . Create the _Hello World_ application source file:
864 #include "hello-tp.h"
866 int main(int argc, char *argv[])
870 puts("Hello, World!\nPress Enter to continue...");
873 * The following getchar() call is only placed here for the purpose
874 * of this demonstration, to pause the application in order for
875 * you to have time to list its tracepoints. It is not
881 * A tracepoint() call.
883 * Arguments, as defined in hello-tp.h:
885 * 1. Tracepoint provider name (required)
886 * 2. Tracepoint name (required)
887 * 3. my_integer_arg (first user-defined argument)
888 * 4. my_string_arg (second user-defined argument)
890 * Notice the tracepoint provider and tracepoint names are
891 * NOT strings: they are in fact parts of variables that the
892 * macros in hello-tp.h create.
894 tracepoint(hello_world, my_first_tracepoint, 23, "hi there!");
896 for (x = 0; x < argc; ++x) {
897 tracepoint(hello_world, my_first_tracepoint, x, argv[x]);
900 puts("Quitting now!");
901 tracepoint(hello_world, my_first_tracepoint, x * x, "x^2");
908 . Build the application:
917 . Link the application with the tracepoint provider package,
918 `liblttng-ust`, and `libdl`:
923 gcc -o hello hello.o hello-tp.o -llttng-ust -ldl
927 Here's the whole build process:
930 .User space tracing tutorial's build steps.
931 image::ust-flow.png[]
933 To trace the user application:
935 . Run the application with a few arguments:
940 ./hello world and beyond
949 Press Enter to continue...
953 . Start an LTTng <<lttng-sessiond,session daemon>>:
958 lttng-sessiond --daemonize
962 Note that a session daemon might already be running, for example as
963 a service that the distribution's service manager started.
965 . List the available user space tracepoints:
970 lttng list --userspace
974 You see the `hello_world:my_first_tracepoint` tracepoint listed
975 under the `./hello` process.
977 . Create a <<tracing-session,tracing session>>:
982 lttng create my-user-space-session
986 . Create an <<event,event rule>> which matches the
987 `hello_world:my_first_tracepoint` event name:
992 lttng enable-event --userspace hello_world:my_first_tracepoint
1005 . Go back to the running `hello` application and press Enter. The
1006 program executes all `tracepoint()` instrumentation points and exits.
1007 . Stop tracing and destroy the tracing session:
1017 The `destroy` command does not destroy the trace data; it only destroys
1018 the state of the tracing session.
1020 By default, LTTng saves the traces in
1021 +$LTTNG_HOME/lttng-traces/__name__-__date__-__time__+,
1022 where +__name__+ is the tracing session name. Note that the
1023 env:LTTNG_HOME environment variable defaults to `$HOME` if not set.
1025 See <<viewing-and-analyzing-your-traces,View and analyze the
1026 recorded events>> to view the recorded events.
1029 [[viewing-and-analyzing-your-traces]]
1030 === View and analyze the recorded events
1032 Once you have completed the <<tracing-the-linux-kernel,Trace the Linux
1033 kernel>> and <<tracing-your-own-user-application,Trace a user
1034 application>> tutorials, you can inspect the recorded events.
1036 Many tools are available to read LTTng traces:
1038 * **cmd:babeltrace** is a command-line utility which converts trace
1039 formats; it supports the format that LTTng produces, CTF, as well as a
1040 basic text output which can be ++grep++ed. The cmd:babeltrace command
1041 is part of the http://diamon.org/babeltrace[Babeltrace] project.
1042 * Babeltrace also includes
1043 **https://www.python.org/[Python] bindings** so
1044 that you can easily open and read an LTTng trace with your own script,
1045 benefiting from the power of Python.
1046 * http://tracecompass.org/[**Trace Compass**]
1047 is a graphical user interface for viewing and analyzing any type of
1048 logs or traces, including LTTng's.
1049 * https://github.com/lttng/lttng-analyses[**LTTng analyses**] is a
1050 project which includes many high-level analyses of LTTng kernel
1051 traces, like scheduling statistics, interrupt frequency distribution,
1052 top CPU usage, and more.
1054 NOTE: This section assumes that the traces recorded during the previous
1055 tutorials were saved to their default location, in the
1056 dir:{$LTTNG_HOME/lttng-traces} directory. Note that the env:LTTNG_HOME
1057 environment variable defaults to `$HOME` if not set.
1060 [[viewing-and-analyzing-your-traces-bt]]
1061 ==== Use the cmd:babeltrace command-line tool
1063 The simplest way to list all the recorded events of a trace is to pass
1064 its path to cmd:babeltrace with no options:
1068 babeltrace ~/lttng-traces/my-user-space-session*
1071 cmd:babeltrace finds all traces recursively within the given path and
1072 prints all their events, merging them in chronological order.
1074 You can pipe the output of cmd:babeltrace into a tool like man:grep(1) for
1079 babeltrace ~/lttng-traces/my-kernel-session* | grep sys_
1082 You can pipe the output of cmd:babeltrace into a tool like man:wc(1) to
1083 count the recorded events:
1087 babeltrace ~/lttng-traces/my-kernel-session* | grep sys_read | wc --lines
1091 [[viewing-and-analyzing-your-traces-bt-python]]
1092 ==== Use the Babeltrace Python bindings
1094 The <<viewing-and-analyzing-your-traces-bt,text output of cmd:babeltrace>>
1095 is useful to isolate events by simple matching using man:grep(1) and
1096 similar utilities. However, more elaborate filters, such as keeping only
1097 event records with a field value falling within a specific range, are
1098 not trivial to write using a shell. Moreover, reductions and even the
1099 most basic computations involving multiple event records are virtually
1100 impossible to implement.
1102 Fortunately, Babeltrace ships with Python 3 bindings which makes it easy
1103 to read the event records of an LTTng trace sequentially and compute the
1104 desired information.
1106 The following script accepts an LTTng Linux kernel trace path as its
1107 first argument and prints the short names of the top 5 running processes
1108 on CPU 0 during the whole trace:
1113 from collections import Counter
1119 if len(sys.argv) != 2:
1120 msg = 'Usage: python3 {} TRACEPATH'.format(sys.argv[0])
1121 print(msg, file=sys.stderr)
1124 # A trace collection contains one or more traces
1125 col = babeltrace.TraceCollection()
1127 # Add the trace provided by the user (LTTng traces always have
1129 if col.add_trace(sys.argv[1], 'ctf') is None:
1130 raise RuntimeError('Cannot add trace')
1132 # This counter dict contains execution times:
1134 # task command name -> total execution time (ns)
1135 exec_times = Counter()
1137 # This contains the last `sched_switch` timestamp
1141 for event in col.events:
1142 # Keep only `sched_switch` events
1143 if event.name != 'sched_switch':
1146 # Keep only events which happened on CPU 0
1147 if event['cpu_id'] != 0:
1151 cur_ts = event.timestamp
1157 # Previous task command (short) name
1158 prev_comm = event['prev_comm']
1160 # Initialize entry in our dict if not yet done
1161 if prev_comm not in exec_times:
1162 exec_times[prev_comm] = 0
1164 # Compute previous command execution time
1165 diff = cur_ts - last_ts
1167 # Update execution time of this command
1168 exec_times[prev_comm] += diff
1170 # Update last timestamp
1174 for name, ns in exec_times.most_common(5):
1176 print('{:20}{} s'.format(name, s))
1181 if __name__ == '__main__':
1182 sys.exit(0 if top5proc() else 1)
1189 python3 top5proc.py ~/lttng-traces/my-kernel-session*/kernel
1195 swapper/0 48.607245889 s
1196 chromium 7.192738188 s
1197 pavucontrol 0.709894415 s
1198 Compositor 0.660867933 s
1199 Xorg.bin 0.616753786 s
1202 Note that `swapper/0` is the "idle" process of CPU 0 on Linux; since we
1203 weren't using the CPU that much when tracing, its first position in the
1208 == [[understanding-lttng]]Core concepts
1210 From a user's perspective, the LTTng system is built on a few concepts,
1211 or objects, on which the <<lttng-cli,cmd:lttng command-line tool>>
1212 operates by sending commands to the <<lttng-sessiond,session daemon>>.
1213 Understanding how those objects relate to eachother is key in mastering
1216 The core concepts are:
1218 * <<tracing-session,Tracing session>>
1219 * <<domain,Tracing domain>>
1220 * <<channel,Channel and ring buffer>>
1221 * <<"event","Instrumentation point, event rule, event, and event record">>
1227 A _tracing session_ is a stateful dialogue between you and
1228 a <<lttng-sessiond,session daemon>>. You can
1229 <<creating-destroying-tracing-sessions,create a new tracing
1230 session>> with the `lttng create` command.
1232 Anything that you do when you control LTTng tracers happens within a
1233 tracing session. In particular, a tracing session:
1236 * Has its own set of trace files.
1237 * Has its own state of activity (started or stopped).
1238 * Has its own <<tracing-session-mode,mode>> (local, network streaming,
1240 * Has its own <<channel,channels>> which have their own
1241 <<event,event rules>>.
1244 .A _tracing session_ contains <<channel,channels>> that are members of <<domain,tracing domains>> and contain <<event,event rules>>.
1245 image::concepts.png[]
1247 Those attributes and objects are completely isolated between different
1250 A tracing session is analogous to a cash machine session:
1251 the operations you do on the banking system through the cash machine do
1252 not alter the data of other users of the same system. In the case of
1253 the cash machine, a session lasts as long as your bank card is inside.
1254 In the case of LTTng, a tracing session lasts from the `lttng create`
1255 command to the `lttng destroy` command.
1258 .Each Unix user has its own set of tracing sessions.
1259 image::many-sessions.png[]
1262 [[tracing-session-mode]]
1263 ==== Tracing session mode
1265 LTTng can send the generated trace data to different locations. The
1266 _tracing session mode_ dictates where to send it. The following modes
1267 are available in LTTng{nbsp}{revision}:
1270 LTTng writes the traces to the file system of the machine being traced
1273 Network streaming mode::
1274 LTTng sends the traces over the network to a
1275 <<lttng-relayd,relay daemon>> running on a remote system.
1278 LTTng does not write the traces by default. Instead, you can request
1279 LTTng to <<taking-a-snapshot,take a snapshot>>, that is, a copy of the
1280 current tracing buffers, and to write it to the target's file system
1281 or to send it over the network to a <<lttng-relayd,relay daemon>>
1282 running on a remote system.
1285 This mode is similar to the network streaming mode, but a live
1286 trace viewer can connect to the distant relay daemon to
1287 <<lttng-live,view event records as LTTng generates them>> by
1294 A _tracing domain_ is a namespace for event sources. A tracing domain
1295 has its own properties and features.
1297 There are currently five available tracing domains:
1301 * `java.util.logging` (JUL)
1305 You must specify a tracing domain when using some commands to avoid
1306 ambiguity. For example, since all the domains support named tracepoints
1307 as event sources (instrumentation points that you manually insert in the
1308 source code), you need to specify a tracing domain when
1309 <<enabling-disabling-events,creating an event rule>> because all the
1310 tracing domains could have tracepoints with the same names.
1312 Some features are reserved to specific tracing domains. Dynamic function
1313 entry and return instrumentation points, for example, are currently only
1314 supported in the Linux kernel tracing domain, but support for other
1315 tracing domains could be added in the future.
1317 You can create <<channel,channels>> in the Linux kernel and user space
1318 tracing domains. The other tracing domains have a single default
1323 === Channel and ring buffer
1325 A _channel_ is an object which is responsible for a set of ring buffers.
1326 Each ring buffer is divided into multiple sub-buffers. When an LTTng
1327 tracer emits an event, it can record it to one or more
1328 sub-buffers. The attributes of a channel determine what to do when
1329 there's no space left for a new event record because all sub-buffers
1330 are full, where to send a full sub-buffer, and other behaviours.
1332 A channel is always associated to a <<domain,tracing domain>>. The
1333 `java.util.logging` (JUL), log4j, and Python tracing domains each have
1334 a default channel which you cannot configure.
1336 A channel also owns <<event,event rules>>. When an LTTng tracer emits
1337 an event, it records it to the sub-buffers of all
1338 the enabled channels with a satisfied event rule, as long as those
1339 channels are part of active <<tracing-session,tracing sessions>>.
1342 [[channel-buffering-schemes]]
1343 ==== Per-user vs. per-process buffering schemes
1345 A channel has at least one ring buffer _per CPU_. LTTng always
1346 records an event to the ring buffer associated to the CPU on which it
1349 Two _buffering schemes_ are available when you
1350 <<enabling-disabling-channels,create a channel>> in the
1351 user space <<domain,tracing domain>>:
1353 Per-user buffering::
1354 Allocate one set of ring buffers--one per CPU--shared by all the
1355 instrumented processes of each Unix user.
1359 .Per-user buffering scheme.
1360 image::per-user-buffering.png[]
1363 Per-process buffering::
1364 Allocate one set of ring buffers--one per CPU--for each
1365 instrumented process.
1369 .Per-process buffering scheme.
1370 image::per-process-buffering.png[]
1373 The per-process buffering scheme tends to consume more memory than the
1374 per-user option because systems generally have more instrumented
1375 processes than Unix users running instrumented processes. However, the
1376 per-process buffering scheme ensures that one process having a high
1377 event throughput won't fill all the shared sub-buffers of the same
1380 The Linux kernel tracing domain has only one available buffering scheme
1381 which is to allocate a single set of ring buffers for the whole system.
1382 This scheme is similar to the per-user option, but with a single, global
1383 user "running" the kernel.
1386 [[channel-overwrite-mode-vs-discard-mode]]
1387 ==== Overwrite vs. discard event loss modes
1389 When an event occurs, LTTng records it to a specific sub-buffer (yellow
1390 arc in the following animation) of a specific channel's ring buffer.
1391 When there's no space left in a sub-buffer, the tracer marks it as
1392 consumable (red) and another, empty sub-buffer starts receiving the
1393 following event records. A <<lttng-consumerd,consumer daemon>>
1394 eventually consumes the marked sub-buffer (returns to white).
1397 [role="docsvg-channel-subbuf-anim"]
1402 In an ideal world, sub-buffers are consumed faster than they are filled,
1403 as is the case in the previous animation. In the real world,
1404 however, all sub-buffers can be full at some point, leaving no space to
1405 record the following events.
1407 By design, LTTng is a _non-blocking_ tracer: when no empty sub-buffer is
1408 available, it is acceptable to lose event records when the alternative
1409 would be to cause substantial delays in the instrumented application's
1410 execution. LTTng privileges performance over integrity; it aims at
1411 perturbing the traced system as little as possible in order to make
1412 tracing of subtle race conditions and rare interrupt cascades possible.
1414 When it comes to losing event records because no empty sub-buffer is
1415 available, the channel's _event loss mode_ determines what to do. The
1416 available event loss modes are:
1419 Drop the newest event records until a the tracer
1420 releases a sub-buffer.
1423 Clear the sub-buffer containing the oldest event records and start
1424 writing the newest event records there.
1426 This mode is sometimes called _flight recorder mode_ because it's
1428 https://en.wikipedia.org/wiki/Flight_recorder[flight recorder]:
1429 always keep a fixed amount of the latest data.
1431 Which mechanism you should choose depends on your context: prioritize
1432 the newest or the oldest event records in the ring buffer?
1434 Beware that, in overwrite mode, the tracer abandons a whole sub-buffer
1435 as soon as a there's no space left for a new event record, whereas in
1436 discard mode, the tracer only discards the event record that doesn't
1439 In discard mode, LTTng increments a count of lost event records when
1440 an event record is lost and saves this count to the trace. In
1441 overwrite mode, LTTng keeps no information when it overwrites a
1442 sub-buffer before consuming it.
1444 There are a few ways to decrease your probability of losing event
1446 <<channel-subbuf-size-vs-subbuf-count,Sub-buffer count and size>> shows
1447 how you can fine-une the sub-buffer count and size of a channel to
1448 virtually stop losing event records, though at the cost of greater
1452 [[channel-subbuf-size-vs-subbuf-count]]
1453 ==== Sub-buffer count and size
1455 When you <<enabling-disabling-channels,create a channel>>, you can
1456 set its number of sub-buffers and their size.
1458 Note that there is noticeable CPU overhead introduced when
1459 switching sub-buffers (marking a full one as consumable and switching
1460 to an empty one for the following events to be recorded). Knowing this,
1461 the following list presents a few practical situations along with how
1462 to configure the sub-buffer count and size for them:
1464 * **High event throughput**: In general, prefer bigger sub-buffers to
1465 lower the risk of losing event records.
1467 Having bigger sub-buffers also ensures a lower sub-buffer switching
1470 The number of sub-buffers is only meaningful if you create the channel
1471 in overwrite mode: in this case, if a sub-buffer overwrite happens, the
1472 other sub-buffers are left unaltered.
1474 * **Low event throughput**: In general, prefer smaller sub-buffers
1475 since the risk of losing event records is low.
1477 Because events occur less frequently, the sub-buffer switching frequency
1478 should remain low and thus the tracer's overhead should not be a
1481 * **Low memory system**: If your target system has a low memory
1482 limit, prefer fewer first, then smaller sub-buffers.
1484 Even if the system is limited in memory, you want to keep the
1485 sub-buffers as big as possible to avoid a high sub-buffer switching
1488 Note that LTTng uses http://diamon.org/ctf/[CTF] as its trace format,
1489 which means event data is very compact. For example, the average
1490 LTTng kernel event record weights about 32{nbsp}bytes. Thus, a
1491 sub-buffer size of 1{nbsp}MiB is considered big.
1493 The previous situations highlight the major trade-off between a few big
1494 sub-buffers and more, smaller sub-buffers: sub-buffer switching
1495 frequency vs. how much data is lost in overwrite mode. Assuming a
1496 constant event throughput and using the overwrite mode, the two
1497 following configurations have the same ring buffer total size:
1500 [role="docsvg-channel-subbuf-size-vs-count-anim"]
1505 * **2 sub-buffers of 4{nbsp}MiB each**: Expect a very low sub-buffer
1506 switching frequency, but if a sub-buffer overwrite happens, half of
1507 the event records so far (4{nbsp}MiB) are definitely lost.
1508 * **8 sub-buffers of 1{nbsp}MiB each**: Expect 4{nbsp}times the tracer's
1509 overhead as the previous configuration, but if a sub-buffer
1510 overwrite happens, only the eighth of event records so far are
1513 In discard mode, the sub-buffers count parameter is pointless: use two
1514 sub-buffers and set their size according to the requirements of your
1518 [[channel-switch-timer]]
1519 ==== Switch timer period
1521 The _switch timer period_ is an important configurable attribute of
1522 a channel to ensure periodic sub-buffer flushing.
1524 When the _switch timer_ expires, a sub-buffer switch happens. You can
1525 set the switch timer period attribute when you
1526 <<enabling-disabling-channels,create a channel>> to ensure that event
1527 data is consumed and committed to trace files or to a distant relay
1528 daemon periodically in case of a low event throughput.
1531 [role="docsvg-channel-switch-timer"]
1536 This attribute is also convenient when you use big sub-buffers to cope
1537 with a sporadic high event throughput, even if the throughput is
1541 [[channel-read-timer]]
1542 ==== Read timer period
1544 By default, the LTTng tracers use a notification mechanism to signal a
1545 full sub-buffer so that a consumer daemon can consume it. When such
1546 notifications must be avoided, for example in real-time applications,
1547 you can use the channel's _read timer_ instead. When the read timer
1548 fires, the <<lttng-consumerd,consumer daemon>> checks for full,
1549 consumable sub-buffers.
1552 [[tracefile-rotation]]
1553 ==== Trace file count and size
1555 By default, trace files can grow as large as needed. You can set the
1556 maximum size of each trace file that a channel writes when you
1557 <<enabling-disabling-channels,create a channel>>. When the size of
1558 a trace file reaches the channel's fixed maximum size, LTTng creates
1559 another file to contain the next event records. LTTng appends a file
1560 count to each trace file name in this case.
1562 If you set the trace file size attribute when you create a channel, the
1563 maximum number of trace files that LTTng creates is _unlimited_ by
1564 default. To limit them, you can also set a maximum number of trace
1565 files. When the number of trace files reaches the channel's fixed
1566 maximum count, the oldest trace file is overwritten. This mechanism is
1567 called _trace file rotation_.
1571 === Instrumentation point, event rule, event, and event record
1573 An _event rule_ is a set of conditions which must be **all** satisfied
1574 for LTTng to record an occuring event.
1576 You set the conditions when you <<enabling-disabling-events,create
1579 You always attach an event rule to <<channel,channel>> when you create
1582 When an event passes the conditions of an event rule, LTTng records it
1583 in one of the attached channel's sub-buffers.
1585 The available conditions, as of LTTng{nbsp}{revision}, are:
1587 * The event rule _is enabled_.
1588 * The instrumentation point's type _is{nbsp}T_.
1589 * The instrumentation point's name (sometimes called _event name_)
1590 _matches{nbsp}N_, but _is not{nbsp}E_.
1591 * The instrumentation point's log level _is as severe as{nbsp}L_, or
1592 _is exactly{nbsp}L_.
1593 * The fields of the event's payload _satisfy_ a filter
1594 expression{nbsp}__F__.
1596 As you can see, all the conditions but the dynamic filter are related to
1597 the event rule's status or to the instrumentation point, not to the
1598 occurring events. This is why, without a filter, checking if an event
1599 passes an event rule is not a dynamic task: when you create or modify an
1600 event rule, all the tracers of its tracing domain enable or disable the
1601 instrumentation points themselves once. This is possible because the
1602 attributes of an instrumentation point (type, name, and log level) are
1603 defined statically. In other words, without a dynamic filter, the tracer
1604 _does not evaluate_ the arguments of an instrumentation point unless it
1605 matches an enabled event rule.
1607 Note that, for LTTng to record an event, the <<channel,channel>> to
1608 which a matching event rule is attached must also be enabled, and the
1609 tracing session owning this channel must be active.
1612 .Logical path from an instrumentation point to an event record.
1613 image::event-rule.png[]
1615 .Event, event record, or event rule?
1617 With so many similar terms, it's easy to get confused.
1619 An **event** is the consequence of the execution of an _instrumentation
1620 point_, like a tracepoint that you manually place in some source code,
1621 or a Linux kernel KProbe. An event is said to _occur_ at a specific
1622 time. Different actions can be taken upon the occurance of an event,
1623 like record the event's payload to a buffer.
1625 An **event record** is the representation of an event in a sub-buffer. A
1626 tracer is responsible for capturing the payload of an event, current
1627 context variables, the event's ID, and the event's timestamp. LTTng
1628 can append this sub-buffer to a trace file.
1630 An **event rule** is a set of conditions which must all be satisfied for
1631 LTTng to record an occuring event. Events still occur without
1632 satisfying event rules, but LTTng does not record them.
1637 == Components of noch:{LTTng}
1639 The second _T_ in _LTTng_ stands for _toolkit_: it would be wrong
1640 to call LTTng a simple _tool_ since it is composed of multiple
1641 interacting components. This section describes those components,
1642 explains their respective roles, and shows how they connect together to
1643 form the LTTng ecosystem.
1645 The following diagram shows how the most important components of LTTng
1646 interact with user applications, the Linux kernel, and you:
1649 .Control and trace data paths between LTTng components.
1650 image::plumbing.png[]
1652 The LTTng project incorporates:
1654 * **LTTng-tools**: Libraries and command-line interface to
1655 control tracing sessions.
1656 ** <<lttng-sessiond,Session daemon>> (man:lttng-sessiond(8)).
1657 ** <<lttng-consumerd,Consumer daemon>> (man:lttng-consumerd(8)).
1658 ** <<lttng-relayd,Relay daemon>> (man:lttng-relayd(8)).
1659 ** <<liblttng-ctl-lttng,Tracing control library>> (`liblttng-ctl`).
1660 ** <<lttng-cli,Tracing control command-line tool>> (man:lttng(1)).
1661 * **LTTng-UST**: Libraries and Java/Python packages to trace user
1663 ** <<lttng-ust,User space tracing library>> (`liblttng-ust`) and its
1664 headers to instrument and trace any native user application.
1665 ** <<prebuilt-ust-helpers,Preloadable user space tracing helpers>>:
1666 *** `liblttng-ust-libc-wrapper`
1667 *** `liblttng-ust-pthread-wrapper`
1668 *** `liblttng-ust-cyg-profile`
1669 *** `liblttng-ust-cyg-profile-fast`
1670 *** `liblttng-ust-dl`
1671 ** User space tracepoint provider source files generator command-line
1672 tool (man:lttng-gen-tp(1)).
1673 ** <<lttng-ust-agents,LTTng-UST Java agent>> to instrument and trace
1674 Java applications using `java.util.logging` or
1675 Apache log4j 1.2 logging.
1676 ** <<lttng-ust-agents,LTTng-UST Python agent>> to instrument
1677 Python applications using the standard `logging` package.
1678 * **LTTng-modules**: <<lttng-modules,Linux kernel modules>> to trace
1680 ** LTTng kernel tracer module.
1681 ** Tracing ring buffer kernel modules.
1682 ** Probe kernel modules.
1683 ** LTTng logger kernel module.
1687 === Tracing control command-line interface
1690 .The tracing control command-line interface.
1691 image::plumbing-lttng-cli.png[]
1693 The _man:lttng(1) command-line tool_ is the standard user interface to
1694 control LTTng <<tracing-session,tracing sessions>>. The cmd:lttng tool
1695 is part of LTTng-tools.
1697 The cmd:lttng tool is linked with
1698 <<liblttng-ctl-lttng,`liblttng-ctl`>> to communicate with
1699 one or more <<lttng-sessiond,session daemons>> behind the scenes.
1701 The cmd:lttng tool has a Git-like interface:
1705 lttng <general options> <command> <command options>
1708 The <<controlling-tracing,Tracing control>> section explores the
1709 available features of LTTng using the cmd:lttng tool.
1712 [[liblttng-ctl-lttng]]
1713 === Tracing control library
1716 .The tracing control library.
1717 image::plumbing-liblttng-ctl.png[]
1719 The _LTTng control library_, `liblttng-ctl`, is used to communicate
1720 with a <<lttng-sessiond,session daemon>> using a C API that hides the
1721 underlying protocol's details. `liblttng-ctl` is part of LTTng-tools.
1723 The <<lttng-cli,cmd:lttng command-line tool>>
1724 is linked with `liblttng-ctl`.
1726 You can use `liblttng-ctl` in C or $$C++$$ source code by including its
1731 #include <lttng/lttng.h>
1734 Some objects are referenced by name (C string), such as tracing
1735 sessions, but most of them require to create a handle first using
1736 `lttng_create_handle()`.
1738 The best available developer documentation for `liblttng-ctl` is, as of
1739 LTTng{nbsp}{revision}, its installed header files. Every function and
1740 structure is thoroughly documented.
1744 === User space tracing library
1747 .The user space tracing library.
1748 image::plumbing-liblttng-ust.png[]
1750 The _user space tracing library_, `liblttng-ust` (see man:lttng-ust(3)),
1751 is the LTTng user space tracer. It receives commands from a
1752 <<lttng-sessiond,session daemon>>, for example to
1753 enable and disable specific instrumentation points, and writes event
1754 records to ring buffers shared with a
1755 <<lttng-consumerd,consumer daemon>>.
1756 `liblttng-ust` is part of LTTng-UST.
1758 Public C header files are installed beside `liblttng-ust` to
1759 instrument any <<c-application,C or $$C++$$ application>>.
1761 <<lttng-ust-agents,LTTng-UST agents>>, which are regular Java and Python
1762 packages, use their own library providing tracepoints which is
1763 linked with `liblttng-ust`.
1765 An application or library does not have to initialize `liblttng-ust`
1766 manually: its constructor does the necessary tasks to properly register
1767 to a session daemon. The initialization phase also enables the
1768 instrumentation points matching the <<event,event rules>> that you
1772 [[lttng-ust-agents]]
1773 === User space tracing agents
1776 .The user space tracing agents.
1777 image::plumbing-lttng-ust-agents.png[]
1779 The _LTTng-UST Java and Python agents_ are regular Java and Python
1780 packages which add LTTng tracing capabilities to the
1781 native logging frameworks. The LTTng-UST agents are part of LTTng-UST.
1783 In the case of Java, the
1784 https://docs.oracle.com/javase/7/docs/api/java/util/logging/package-summary.html[`java.util.logging`
1785 core logging facilities] and
1786 https://logging.apache.org/log4j/1.2/[Apache log4j 1.2] are supported.
1787 Note that Apache Log4{nbsp}2 is not supported.
1789 In the case of Python, the standard
1790 https://docs.python.org/3/library/logging.html[`logging`] package
1791 is supported. Both Python 2 and Python 3 modules can import the
1792 LTTng-UST Python agent package.
1794 The applications using the LTTng-UST agents are in the
1795 `java.util.logging` (JUL),
1796 log4j, and Python <<domain,tracing domains>>.
1798 Both agents use the same mechanism to trace the log statements. When an
1799 agent is initialized, it creates a log handler that attaches to the root
1800 logger. The agent also registers to a <<lttng-sessiond,session daemon>>.
1801 When the application executes a log statement, it is passed to the
1802 agent's log handler by the root logger. The agent's log handler calls a
1803 native function in a tracepoint provider package shared library linked
1804 with <<lttng-ust,`liblttng-ust`>>, passing the formatted log message and
1805 other fields, like its logger name and its log level. This native
1806 function contains a user space instrumentation point, hence tracing the
1809 The log level condition of an
1810 <<event,event rule>> is considered when tracing
1811 a Java or a Python application, and it's compatible with the standard
1812 JUL, log4j, and Python log levels.
1816 === LTTng kernel modules
1819 .The LTTng kernel modules.
1820 image::plumbing-lttng-modules.png[]
1822 The _LTTng kernel modules_ are a set of Linux kernel modules
1823 which implement the kernel tracer of the LTTng project. The LTTng
1824 kernel modules are part of LTTng-modules.
1826 The LTTng kernel modules include:
1828 * A set of _probe_ modules.
1830 Each module attaches to a specific subsystem
1831 of the Linux kernel using its tracepoint instrument points. There are
1832 also modules to attach to the entry and return points of the Linux
1833 system call functions.
1835 * _Ring buffer_ modules.
1837 A ring buffer implementation is provided as kernel modules. The LTTng
1838 kernel tracer writes to the ring buffer; a
1839 <<lttng-consumerd,consumer daemon>> reads from the ring buffer.
1841 * The _LTTng kernel tracer_ module.
1842 * The _LTTng logger_ module.
1844 The LTTng logger module implements the special path:{/proc/lttng-logger}
1845 file so that any executable can generate LTTng events by opening and
1846 writing to this file.
1848 See <<proc-lttng-logger-abi,LTTng logger>>.
1850 Generally, you do not have to load the LTTng kernel modules manually
1851 (using man:modprobe(8), for example): a root <<lttng-sessiond,session
1852 daemon>> loads the necessary modules when starting. If you have extra
1853 probe modules, you can specify to load them to the session daemon on
1856 The LTTng kernel modules are installed in
1857 +/usr/lib/modules/__release__/extra+ by default, where +__release__+ is
1858 the kernel release (see `uname --kernel-release`).
1865 .The session daemon.
1866 image::plumbing-sessiond.png[]
1868 The _session daemon_, man:lttng-sessiond(8), is a daemon responsible for
1869 managing tracing sessions and for controlling the various components of
1870 LTTng. The session daemon is part of LTTng-tools.
1872 The session daemon sends control requests to and receives control
1875 * The <<lttng-ust,user space tracing library>>.
1877 Any instance of the user space tracing library first registers to
1878 a session daemon. Then, the session daemon can send requests to
1879 this instance, such as:
1882 ** Get the list of tracepoints.
1883 ** Share an <<event,event rule>> so that the user space tracing library
1884 can enable or disable tracepoints. Amongst the possible conditions
1885 of an event rule is a filter expression which `liblttng-ust` evalutes
1886 when an event occurs.
1887 ** Share <<channel,channel>> attributes and ring buffer locations.
1890 The session daemon and the user space tracing library use a Unix
1891 domain socket for their communication.
1893 * The <<lttng-ust-agents,user space tracing agents>>.
1895 Any instance of a user space tracing agent first registers to
1896 a session daemon. Then, the session daemon can send requests to
1897 this instance, such as:
1900 ** Get the list of loggers.
1901 ** Enable or disable a specific logger.
1904 The session daemon and the user space tracing agent use a TCP connection
1905 for their communication.
1907 * The <<lttng-modules,LTTng kernel tracer>>.
1908 * The <<lttng-consumerd,consumer daemon>>.
1910 The session daemon sends requests to the consumer daemon to instruct
1911 it where to send the trace data streams, amongst other information.
1913 * The <<lttng-relayd,relay daemon>>.
1915 The session daemon receives commands from the
1916 <<liblttng-ctl-lttng,tracing control library>>.
1918 The root session daemon loads the appropriate
1919 <<lttng-modules,LTTng kernel modules>> on startup. It also spawns
1920 a <<lttng-consumerd,consumer daemon>> as soon as you create
1921 an <<event,event rule>>.
1923 The session daemon does not send and receive trace data: this is the
1924 role of the <<lttng-consumerd,consumer daemon>> and
1925 <<lttng-relayd,relay daemon>>. It does, however, generate the
1926 http://diamon.org/ctf/[CTF] metadata stream.
1928 Each Unix user can have its own session daemon instance. The
1929 tracing sessions managed by different session daemons are completely
1932 The root user's session daemon is the only one which is
1933 allowed to control the LTTng kernel tracer, and its spawned consumer
1934 daemon is the only one which is allowed to consume trace data from the
1935 LTTng kernel tracer. Note, however, that any Unix user which is a member
1936 of the <<tracing-group,tracing group>> is allowed
1937 to create <<channel,channels>> in the
1938 Linux kernel <<domain,tracing domain>>, and thus to trace the Linux
1941 The <<lttng-cli,cmd:lttng command-line tool>> automatically starts a
1942 session daemon when using its `create` command if none is currently
1943 running. You can also start the session daemon manually.
1950 .The consumer daemon.
1951 image::plumbing-consumerd.png[]
1953 The _consumer daemon_, man:lttng-consumerd(8), is a daemon which shares
1954 ring buffers with user applications or with the LTTng kernel modules to
1955 collect trace data and send it to some location (on disk or to a
1956 <<lttng-relayd,relay daemon>> over the network). The consumer daemon
1957 is part of LTTng-tools.
1959 You do not start a consumer daemon manually: a consumer daemon is always
1960 spawned by a <<lttng-sessiond,session daemon>> as soon as you create an
1961 <<event,event rule>>, that is, before you start tracing. When you kill
1962 its owner session daemon, the consumer daemon also exits because it is
1963 the session daemon's child process. Command-line options of
1964 man:lttng-sessiond(8) target the consumer daemon process.
1966 There are up to two running consumer daemons per Unix user, whereas only
1967 one session daemon can run per user. This is because each process can be
1968 either 32-bit or 64-bit: if the target system runs a mixture of 32-bit
1969 and 64-bit processes, it is more efficient to have separate
1970 corresponding 32-bit and 64-bit consumer daemons. The root user is an
1971 exception: it can have up to _three_ running consumer daemons: 32-bit
1972 and 64-bit instances for its user applications, and one more
1973 reserved for collecting kernel trace data.
1981 image::plumbing-relayd.png[]
1983 The _relay daemon_, man:lttng-relayd(8), is a daemon acting as a bridge
1984 between remote session and consumer daemons, local trace files, and a
1985 remote live trace viewer. The relay daemon is part of LTTng-tools.
1987 The main purpose of the relay daemon is to implement a receiver of
1988 <<sending-trace-data-over-the-network,trace data over the network>>.
1989 This is useful when the target system does not have much file system
1990 space to record trace files locally.
1992 The relay daemon is also a server to which a
1993 <<lttng-live,live trace viewer>> can
1994 connect. The live trace viewer sends requests to the relay daemon to
1995 receive trace data as the target system emits events. The
1996 communication protocol is named _LTTng live_; it is used over TCP
1999 Note that you can start the relay daemon on the target system directly.
2000 This is the setup of choice when the use case is to view events as
2001 the target system emits them without the need of a remote system.
2005 == [[using-lttng]]Instrumentation
2007 There are many examples of tracing and monitoring in our everyday life:
2009 * You have access to real-time and historical weather reports and
2010 forecasts thanks to weather stations installed around the country.
2011 * You know your heart is safe thanks to an electrocardiogram.
2012 * You make sure not to drive your car too fast and to have enough fuel
2013 to reach your destination thanks to gauges visible on your dashboard.
2015 All the previous examples have something in common: they rely on
2016 **instruments**. Without the electrodes attached to the surface of your
2017 body's skin, cardiac monitoring is futile.
2019 LTTng, as a tracer, is no different from those real life examples. If
2020 you're about to trace a software system or, in other words, record its
2021 history of execution, you better have **instrumentation points** in the
2022 subject you're tracing, that is, the actual software.
2024 Various ways were developed to instrument a piece of software for LTTng
2025 tracing. The most straightforward one is to manually place
2026 instrumentation points, called _tracepoints_, in the software's source
2027 code. It is also possible to add instrumentation points dynamically in
2028 the Linux kernel <<domain,tracing domain>>.
2030 If you're only interested in tracing the Linux kernel, your
2031 instrumentation needs are probably already covered by LTTng's built-in
2032 <<lttng-modules,Linux kernel tracepoints>>. You may also wish to trace a
2033 user application which is already instrumented for LTTng tracing.
2034 In such cases, you can skip this whole section and read the topics of
2035 the <<controlling-tracing,Tracing control>> section.
2037 Many methods are available to instrument a piece of software for LTTng
2040 * <<c-application,User space instrumentation for C and $$C++$$
2042 * <<prebuilt-ust-helpers,Prebuilt user space tracing helpers>>.
2043 * <<java-application,User space Java agent>>.
2044 * <<python-application,User space Python agent>>.
2045 * <<proc-lttng-logger-abi,LTTng logger>>.
2046 * <<instrumenting-linux-kernel,LTTng kernel tracepoints>>.
2050 === [[cxx-application]]User space instrumentation for C and $$C++$$ applications
2052 The procedure to instrument a C or $$C++$$ user application with
2053 the <<lttng-ust,LTTng user space tracing library>>, `liblttng-ust`, is:
2055 . <<tracepoint-provider,Create the source files of a tracepoint provider
2057 . <<probing-the-application-source-code,Add tracepoints to
2058 the application's source code>>.
2059 . <<building-tracepoint-providers-and-user-application,Build and link
2060 a tracepoint provider package and the user application>>.
2062 If you need quick, man:printf(3)-like instrumentation, you can skip
2063 those steps and use <<tracef,`tracef()`>> or <<tracelog,`tracelog()`>>
2066 IMPORTANT: You need to <<installing-lttng,install>> LTTng-UST to
2067 instrument a user application with `liblttng-ust`.
2070 [[tracepoint-provider]]
2071 ==== Create the source files of a tracepoint provider package
2073 A _tracepoint provider_ is a set of compiled functions which provide
2074 **tracepoints** to an application, the type of instrumentation point
2075 supported by LTTng-UST. Those functions can emit events with
2076 user-defined fields and serialize those events as event records to one
2077 or more LTTng-UST <<channel,channel>> sub-buffers. The `tracepoint()`
2078 macro, which you <<probing-the-application-source-code,insert in a user
2079 application's source code>>, calls those functions.
2081 A _tracepoint provider package_ is an object file (`.o`) or a shared
2082 library (`.so`) which contains one or more tracepoint providers.
2083 Its source files are:
2085 * One or more <<tpp-header,tracepoint provider header>> (`.h`).
2086 * A <<tpp-source,tracepoint provider package source>> (`.c`).
2088 A tracepoint provider package is dynamically linked with `liblttng-ust`,
2089 the LTTng user space tracer, at run time.
2092 .User application linked with `liblttng-ust` and containing a tracepoint provider.
2093 image::ust-app.png[]
2095 NOTE: If you need quick, man:printf(3)-like instrumentation, you can
2096 skip creating and using a tracepoint provider and use
2097 <<tracef,`tracef()`>> or <<tracelog,`tracelog()`>> instead.
2101 ===== Create a tracepoint provider header file template
2103 A _tracepoint provider header file_ contains the tracepoint
2104 definitions of a tracepoint provider.
2106 To create a tracepoint provider header file:
2108 . Start from this template:
2112 .Tracepoint provider header file template (`.h` file extension).
2114 #undef TRACEPOINT_PROVIDER
2115 #define TRACEPOINT_PROVIDER provider_name
2117 #undef TRACEPOINT_INCLUDE
2118 #define TRACEPOINT_INCLUDE "./tp.h"
2120 #if !defined(_TP_H) || defined(TRACEPOINT_HEADER_MULTI_READ)
2123 #include <lttng/tracepoint.h>
2126 * Use TRACEPOINT_EVENT(), TRACEPOINT_EVENT_CLASS(),
2127 * TRACEPOINT_EVENT_INSTANCE(), and TRACEPOINT_LOGLEVEL() here.
2132 #include <lttng/tracepoint-event.h>
2138 * `provider_name` with the name of your tracepoint provider.
2139 * `"tp.h"` with the name of your tracepoint provider header file.
2141 . Below the `#include <lttng/tracepoint.h>` line, put your
2142 <<defining-tracepoints,tracepoint definitions>>.
2144 Your tracepoint provider name must be unique amongst all the possible
2145 tracepoint provider names used on the same target system. We
2146 suggest to include the name of your project or company in the name,
2147 for example, `org_lttng_my_project_tpp`.
2149 TIP: [[lttng-gen-tp]]You can use the man:lttng-gen-tp(1) tool to create
2150 this boilerplate for you. When using cmd:lttng-gen-tp, all you need to
2151 write are the <<defining-tracepoints,tracepoint definitions>>.
2154 [[defining-tracepoints]]
2155 ===== Create a tracepoint definition
2157 A _tracepoint definition_ defines, for a given tracepoint:
2159 * Its **input arguments**. They are the macro parameters that the
2160 `tracepoint()` macro accepts for this particular tracepoint
2161 in the user application's source code.
2162 * Its **output event fields**. They are the sources of event fields
2163 that form the payload of any event that the execution of the
2164 `tracepoint()` macro emits for this particular tracepoint.
2166 You can create a tracepoint definition by using the
2167 `TRACEPOINT_EVENT()` macro below the `#include <lttng/tracepoint.h>`
2169 <<tpp-header,tracepoint provider header file template>>.
2171 The syntax of the `TRACEPOINT_EVENT()` macro is:
2174 .`TRACEPOINT_EVENT()` macro syntax.
2177 /* Tracepoint provider name */
2180 /* Tracepoint name */
2183 /* Input arguments */
2188 /* Output event fields */
2197 * `provider_name` with your tracepoint provider name.
2198 * `tracepoint_name` with your tracepoint name.
2199 * `arguments` with the <<tpp-def-input-args,input arguments>>.
2200 * `fields` with the <<tpp-def-output-fields,output event field>>
2203 This tracepoint emits events named `provider_name:tracepoint_name`.
2206 .Event name's length limitation
2208 The concatenation of the tracepoint provider name and the
2209 tracepoint name must not exceed **254 characters**. If it does, the
2210 instrumented application compiles and runs, but LTTng throws multiple
2211 warnings and you could experience serious issues.
2214 [[tpp-def-input-args]]The syntax of the `TP_ARGS()` macro is:
2217 .`TP_ARGS()` macro syntax.
2226 * `type` with the C type of the argument.
2227 * `arg_name` with the argument name.
2229 You can repeat `type` and `arg_name` up to 10 times to have
2230 more than one argument.
2232 .`TP_ARGS()` usage with three arguments.
2244 The `TP_ARGS()` and `TP_ARGS(void)` forms are valid to create a
2245 tracepoint definition with no input arguments.
2247 [[tpp-def-output-fields]]The `TP_FIELDS()` macro contains a list of
2248 `ctf_*()` macros. Each `ctf_*()` macro defines one event field.
2249 See <<liblttng-ust-tp-fields,Tracepoint fields macros>> for a
2250 complete description of the available `ctf_*()` macros.
2251 A `ctf_*()` macro specifies the type, size, and byte order of
2254 Each `ctf_*()` macro takes an _argument expression_ parameter. This is a
2255 C expression that the tracer evalutes at the `tracepoint()` macro site
2256 in the application's source code. This expression provides a field's
2257 source of data. The argument expression can include input argument names
2258 listed in the `TP_ARGS()` macro.
2260 Each `ctf_*()` macro also takes a _field name_ parameter. Field names
2261 must be unique within a given tracepoint definition.
2263 Here's a complete tracepoint definition example:
2265 .Tracepoint definition.
2267 The following tracepoint definition defines a tracepoint which takes
2268 three input arguments and has four output event fields.
2272 #include "my-custom-structure.h"
2278 const struct my_custom_structure*, my_custom_structure,
2283 ctf_string(query_field, query)
2284 ctf_float(double, ratio_field, ratio)
2285 ctf_integer(int, recv_size, my_custom_structure->recv_size)
2286 ctf_integer(int, send_size, my_custom_structure->send_size)
2291 You can refer to this tracepoint definition with the `tracepoint()`
2292 macro in your application's source code like this:
2296 tracepoint(my_provider, my_tracepoint,
2297 my_structure, some_ratio, the_query);
2301 NOTE: The LTTng tracer only evaluates tracepoint arguments at run time
2302 if they satisfy an enabled <<event,event rule>>.
2305 [[using-tracepoint-classes]]
2306 ===== Use a tracepoint class
2308 A _tracepoint class_ is a class of tracepoints which share the same
2309 output event field definitions. A _tracepoint instance_ is one
2310 instance of such a defined tracepoint class, with its own tracepoint
2313 The <<defining-tracepoints,`TRACEPOINT_EVENT()` macro>> is actually a
2314 shorthand which defines both a tracepoint class and a tracepoint
2315 instance at the same time.
2317 When you build a tracepoint provider package, the C or $$C++$$ compiler
2318 creates one serialization function for each **tracepoint class**. A
2319 serialization function is responsible for serializing the event fields
2320 of a tracepoint to a sub-buffer when tracing.
2322 For various performance reasons, when your situation requires multiple
2323 tracepoint definitions with different names, but with the same event
2324 fields, we recommend that you manually create a tracepoint class
2325 and instantiate as many tracepoint instances as needed. One positive
2326 effect of such a design, amongst other advantages, is that all
2327 tracepoint instances of the same tracepoint class reuse the same
2328 serialization function, thus reducing
2329 https://en.wikipedia.org/wiki/Cache_pollution[cache pollution].
2331 .Use a tracepoint class and tracepoint instances.
2333 Consider the following three tracepoint definitions:
2345 ctf_integer(int, userid, userid)
2346 ctf_integer(size_t, len, len)
2358 ctf_integer(int, userid, userid)
2359 ctf_integer(size_t, len, len)
2371 ctf_integer(int, userid, userid)
2372 ctf_integer(size_t, len, len)
2377 In this case, we create three tracepoint classes, with one implicit
2378 tracepoint instance for each of them: `get_account`, `get_settings`, and
2379 `get_transaction`. However, they all share the same event field names
2380 and types. Hence three identical, yet independent serialization
2381 functions are created when you build the tracepoint provider package.
2383 A better design choice is to define a single tracepoint class and three
2384 tracepoint instances:
2388 /* The tracepoint class */
2389 TRACEPOINT_EVENT_CLASS(
2390 /* Tracepoint provider name */
2393 /* Tracepoint class name */
2396 /* Input arguments */
2402 /* Output event fields */
2404 ctf_integer(int, userid, userid)
2405 ctf_integer(size_t, len, len)
2409 /* The tracepoint instances */
2410 TRACEPOINT_EVENT_INSTANCE(
2411 /* Tracepoint provider name */
2414 /* Tracepoint class name */
2417 /* Tracepoint name */
2420 /* Input arguments */
2426 TRACEPOINT_EVENT_INSTANCE(
2435 TRACEPOINT_EVENT_INSTANCE(
2448 [[assigning-log-levels]]
2449 ===== Assign a log level to a tracepoint definition
2451 You can assign an optional _log level_ to a
2452 <<defining-tracepoints,tracepoint definition>>.
2454 Assigning different levels of severity to tracepoint definitions can
2455 be useful: when you <<enabling-disabling-events,create an event rule>>,
2456 you can target tracepoints having a log level as severe as a specific
2459 The concept of LTTng-UST log levels is similar to the levels found
2460 in typical logging frameworks:
2462 * In a logging framework, the log level is given by the function
2463 or method name you use at the log statement site: `debug()`,
2464 `info()`, `warn()`, `error()`, and so on.
2465 * In LTTng-UST, you statically assign the log level to a tracepoint
2466 definition; any `tracepoint()` macro invocation which refers to
2467 this definition has this log level.
2469 You can assign a log level to a tracepoint definition with the
2470 `TRACEPOINT_LOGLEVEL()` macro. You must use this macro _after_ the
2471 <<defining-tracepoints,`TRACEPOINT_EVENT()`>> or
2472 <<using-tracepoint-classes,`TRACEPOINT_INSTANCE()`>> macro for a given
2475 The syntax of the `TRACEPOINT_LOGLEVEL()` macro is:
2478 .`TRACEPOINT_LOGLEVEL()` macro syntax.
2480 TRACEPOINT_LOGLEVEL(provider_name, tracepoint_name, log_level)
2485 * `provider_name` with the tracepoint provider name.
2486 * `tracepoint_name` with the tracepoint name.
2487 * `log_level` with the log level to assign to the tracepoint
2488 definition named `tracepoint_name` in the `provider_name`
2489 tracepoint provider.
2491 See <<liblttng-ust-tracepoint-loglevel,Tracepoint log levels>> for
2492 a list of available log level names.
2494 .Assign the `TRACE_DEBUG_UNIT` log level to a tracepoint definition.
2498 /* Tracepoint definition */
2507 ctf_integer(int, userid, userid)
2508 ctf_integer(size_t, len, len)
2512 /* Log level assignment */
2513 TRACEPOINT_LOGLEVEL(my_app, get_transaction, TRACE_DEBUG_UNIT)
2519 ===== Create a tracepoint provider package source file
2521 A _tracepoint provider package source file_ is a C source file which
2522 includes a <<tpp-header,tracepoint provider header file>> to expand its
2523 macros into event serialization and other functions.
2525 You can always use the following tracepoint provider package source
2529 .Tracepoint provider package source file template.
2531 #define TRACEPOINT_CREATE_PROBES
2536 Replace `tp.h` with the name of your <<tpp-header,tracepoint provider
2537 header file>> name. You may also include more than one tracepoint
2538 provider header file here to create a tracepoint provider package
2539 holding more than one tracepoint providers.
2542 [[probing-the-application-source-code]]
2543 ==== Add tracepoints to an application's source code
2545 Once you <<tpp-header,create a tracepoint provider header file>>, you
2546 can use the `tracepoint()` macro in your application's
2547 source code to insert the tracepoints that this header
2548 <<defining-tracepoints,defined>> defines.
2550 The `tracepoint()` macro takes at least two parameters: the tracepoint
2551 provider name and the tracepoint name. The corresponding tracepoint
2552 definition defines the other parameters.
2554 .`tracepoint()` usage.
2556 The following <<defining-tracepoints,tracepoint definition>> defines a
2557 tracepoint which takes two input arguments and has two output event
2561 .Tracepoint provider header file.
2563 #include "my-custom-structure.h"
2570 const char*, cmd_name
2573 ctf_string(cmd_name, cmd_name)
2574 ctf_integer(int, number_of_args, argc)
2579 You can refer to this tracepoint definition with the `tracepoint()`
2580 macro in your application's source code like this:
2583 .Application's source file.
2587 int main(int argc, char* argv[])
2589 tracepoint(my_provider, my_tracepoint, argc, argv[0]);
2595 Note how the application's source code includes
2596 the tracepoint provider header file containing the tracepoint
2597 definitions to use, path:{tp.h}.
2600 .`tracepoint()` usage with a complex tracepoint definition.
2602 Consider this complex tracepoint definition, where multiple event
2603 fields refer to the same input arguments in their argument expression
2607 .Tracepoint provider header file.
2609 /* For `struct stat` */
2610 #include <sys/types.h>
2611 #include <sys/stat.h>
2623 ctf_integer(int, my_constant_field, 23 + 17)
2624 ctf_integer(int, my_int_arg_field, my_int_arg)
2625 ctf_integer(int, my_int_arg_field2, my_int_arg * my_int_arg)
2626 ctf_integer(int, sum4_field, my_str_arg[0] + my_str_arg[1] +
2627 my_str_arg[2] + my_str_arg[3])
2628 ctf_string(my_str_arg_field, my_str_arg)
2629 ctf_integer_hex(off_t, size_field, st->st_size)
2630 ctf_float(double, size_dbl_field, (double) st->st_size)
2631 ctf_sequence_text(char, half_my_str_arg_field, my_str_arg,
2632 size_t, strlen(my_str_arg) / 2)
2637 You can refer to this tracepoint definition with the `tracepoint()`
2638 macro in your application's source code like this:
2641 .Application's source file.
2643 #define TRACEPOINT_DEFINE
2650 stat("/etc/fstab", &s);
2651 tracepoint(my_provider, my_tracepoint, 23, "Hello, World!", &s);
2657 If you look at the event record that LTTng writes when tracing this
2658 program, assuming the file size of path:{/etc/fstab} is 301{nbsp}bytes,
2659 it should look like this:
2661 .Event record fields
2663 |Field's name |Field's value
2664 |`my_constant_field` |40
2665 |`my_int_arg_field` |23
2666 |`my_int_arg_field2` |529
2668 |`my_str_arg_field` |`Hello, World!`
2669 |`size_field` |0x12d
2670 |`size_dbl_field` |301.0
2671 |`half_my_str_arg_field` |`Hello,`
2675 Sometimes, the arguments you pass to `tracepoint()` are expensive to
2676 compute--they use the call stack, for example. To avoid this
2677 computation when the tracepoint is disabled, you can use the
2678 `tracepoint_enabled()` and `do_tracepoint()` macros.
2680 The syntax of the `tracepoint_enabled()` and `do_tracepoint()` macros
2684 .`tracepoint_enabled()` and `do_tracepoint()` macros syntax.
2686 tracepoint_enabled(provider_name, tracepoint_name)
2687 do_tracepoint(provider_name, tracepoint_name, ...)
2692 * `provider_name` with the tracepoint provider name.
2693 * `tracepoint_name` with the tracepoint name.
2695 `tracepoint_enabled()` returns a non-zero value if the tracepoint named
2696 `tracepoint_name` from the provider named `provider_name` is enabled
2699 `do_tracepoint()` is like `tracepoint()`, except that it doesn't check
2700 if the tracepoint is enabled. Using `tracepoint()` with
2701 `tracepoint_enabled()` is dangerous since `tracepoint()` also contains
2702 the `tracepoint_enabled()` check, thus a race condition is
2703 possible in this situation:
2706 .Possible race condition when using `tracepoint_enabled()` with `tracepoint()`.
2708 if (tracepoint_enabled(my_provider, my_tracepoint)) {
2709 stuff = prepare_stuff();
2712 tracepoint(my_provider, my_tracepoint, stuff);
2715 If the tracepoint is enabled after the condition, then `stuff` is not
2716 prepared: the emitted event will either contain wrong data, or the whole
2717 application could crash (segmentation fault, for example).
2719 NOTE: Neither `tracepoint_enabled()` nor `do_tracepoint()` have an
2720 `STAP_PROBEV()` call. If you need it, you must emit
2724 [[building-tracepoint-providers-and-user-application]]
2725 ==== Build and link a tracepoint provider package and an application
2727 Once you have one or more <<tpp-header,tracepoint provider header
2728 files>> and a <<tpp-source,tracepoint provider package source file>>,
2729 you can create the tracepoint provider package by compiling its source
2730 file. From here, multiple build and run scenarios are possible. The
2731 following table shows common application and library configurations
2732 along with the required command lines to achieve them.
2734 In the following diagrams, we use the following file names:
2737 Executable application.
2740 Application's object file.
2743 Tracepoint provider package object file.
2746 Tracepoint provider package archive file.
2749 Tracepoint provider package shared object file.
2752 User library object file.
2755 User library shared object file.
2757 The red star indicates that this object file is instrumented
2758 (contains code which uses the `tracepoint()` macro). The spring
2759 symbol between the application and a library means the application is
2760 linked with the library at build time.
2762 We assume that path:{.} is part of the env:LD_LIBRARY_PATH environment
2763 variable in the following instructions.
2765 [role="growable ust-scenarios",cols="asciidoc,asciidoc"]
2766 .Common tracepoint provider package scenarios.
2768 |Scenario |Instructions
2771 The instrumented application is statically linked with
2772 the tracepoint provider package object.
2774 image::ust-sit+app-linked-with-tp-o+app-instrumented.png[]
2777 include::../common/ust-sit-step-tp-o.txt[]
2779 To build the instrumented application:
2781 . In path:{app.c}, before including path:{tpp.h}, add the following line:
2786 #define TRACEPOINT_DEFINE
2790 . Compile the application source file:
2799 . Build the application:
2804 gcc -o app app.o tpp.o -llttng-ust -ldl
2808 To run the instrumented application:
2810 * Start the application:
2820 The instrumented application is statically linked with the
2821 tracepoint provider package archive file.
2823 image::ust-sit+app-linked-with-tp-a+app-instrumented.png[]
2826 To create the tracepoint provider package archive file:
2828 . Compile the <<tpp-source,tracepoint provider package source file>>:
2837 . Create the tracepoint provider package archive file:
2846 To build the instrumented application:
2848 . In path:{app.c}, before including path:{tpp.h}, add the following line:
2853 #define TRACEPOINT_DEFINE
2857 . Compile the application source file:
2866 . Build the application:
2871 gcc -o app app.o tpp.a -llttng-ust -ldl
2875 To run the instrumented application:
2877 * Start the application:
2887 The instrumented application is linked with the tracepoint provider
2888 package shared object.
2890 image::ust-sit+app-linked-with-tp-so+app-instrumented.png[]
2893 include::../common/ust-sit-step-tp-so.txt[]
2895 To build the instrumented application:
2897 . In path:{app.c}, before including path:{tpp.h}, add the following line:
2902 #define TRACEPOINT_DEFINE
2906 . Compile the application source file:
2915 . Build the application:
2920 gcc -o app app.o -ldl -L. -ltpp
2924 To run the instrumented application:
2926 * Start the application:
2936 The tracepoint provider package shared object is preloaded before the
2937 instrumented application starts.
2939 image::ust-sit+tp-so-preloaded+app-instrumented.png[]
2942 include::../common/ust-sit-step-tp-so.txt[]
2944 To build the instrumented application:
2946 . In path:{app.c}, before including path:{tpp.h}, add the
2952 #define TRACEPOINT_DEFINE
2953 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
2957 . Compile the application source file:
2966 . Build the application:
2971 gcc -o app app.o -ldl
2975 To run the instrumented application with tracing support:
2977 * Preload the tracepoint provider package shared object and
2978 start the application:
2983 LD_PRELOAD=./libtpp.so ./app
2987 To run the instrumented application without tracing support:
2989 * Start the application:
2999 The instrumented application dynamically loads the tracepoint provider
3000 package shared object.
3002 See the <<dlclose-warning,warning about `dlclose()`>>.
3004 image::ust-sit+app-dlopens-tp-so+app-instrumented.png[]
3007 include::../common/ust-sit-step-tp-so.txt[]
3009 To build the instrumented application:
3011 . In path:{app.c}, before including path:{tpp.h}, add the
3017 #define TRACEPOINT_DEFINE
3018 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3022 . Compile the application source file:
3031 . Build the application:
3036 gcc -o app app.o -ldl
3040 To run the instrumented application:
3042 * Start the application:
3052 The application is linked with the instrumented user library.
3054 The instrumented user library is statically linked with the tracepoint
3055 provider package object file.
3057 image::ust-sit+app-linked-with-lib+lib-linked-with-tp-o+lib-instrumented.png[]
3060 include::../common/ust-sit-step-tp-o-fpic.txt[]
3062 To build the instrumented user library:
3064 . In path:{emon.c}, before including path:{tpp.h}, add the
3070 #define TRACEPOINT_DEFINE
3074 . Compile the user library source file:
3079 gcc -I. -fpic -c emon.c
3083 . Build the user library shared object:
3088 gcc -shared -o libemon.so emon.o tpp.o -llttng-ust -ldl
3092 To build the application:
3094 . Compile the application source file:
3103 . Build the application:
3108 gcc -o app app.o -L. -lemon
3112 To run the application:
3114 * Start the application:
3124 The application is linked with the instrumented user library.
3126 The instrumented user library is linked with the tracepoint provider
3127 package shared object.
3129 image::ust-sit+app-linked-with-lib+lib-linked-with-tp-so+lib-instrumented.png[]
3132 include::../common/ust-sit-step-tp-so.txt[]
3134 To build the instrumented user library:
3136 . In path:{emon.c}, before including path:{tpp.h}, add the
3142 #define TRACEPOINT_DEFINE
3146 . Compile the user library source file:
3151 gcc -I. -fpic -c emon.c
3155 . Build the user library shared object:
3160 gcc -shared -o libemon.so emon.o -ldl -L. -ltpp
3164 To build the application:
3166 . Compile the application source file:
3175 . Build the application:
3180 gcc -o app app.o -L. -lemon
3184 To run the application:
3186 * Start the application:
3196 The tracepoint provider package shared object is preloaded before the
3199 The application is linked with the instrumented user library.
3201 image::ust-sit+tp-so-preloaded+app-linked-with-lib+lib-instrumented.png[]
3204 include::../common/ust-sit-step-tp-so.txt[]
3206 To build the instrumented user library:
3208 . In path:{emon.c}, before including path:{tpp.h}, add the
3214 #define TRACEPOINT_DEFINE
3215 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3219 . Compile the user library source file:
3224 gcc -I. -fpic -c emon.c
3228 . Build the user library shared object:
3233 gcc -shared -o libemon.so emon.o -ldl
3237 To build the application:
3239 . Compile the application source file:
3248 . Build the application:
3253 gcc -o app app.o -L. -lemon
3257 To run the application with tracing support:
3259 * Preload the tracepoint provider package shared object and
3260 start the application:
3265 LD_PRELOAD=./libtpp.so ./app
3269 To run the application without tracing support:
3271 * Start the application:
3281 The application is linked with the instrumented user library.
3283 The instrumented user library dynamically loads the tracepoint provider
3284 package shared object.
3286 See the <<dlclose-warning,warning about `dlclose()`>>.
3288 image::ust-sit+app-linked-with-lib+lib-dlopens-tp-so+lib-instrumented.png[]
3291 include::../common/ust-sit-step-tp-so.txt[]
3293 To build the instrumented user library:
3295 . In path:{emon.c}, before including path:{tpp.h}, add the
3301 #define TRACEPOINT_DEFINE
3302 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3306 . Compile the user library source file:
3311 gcc -I. -fpic -c emon.c
3315 . Build the user library shared object:
3320 gcc -shared -o libemon.so emon.o -ldl
3324 To build the application:
3326 . Compile the application source file:
3335 . Build the application:
3340 gcc -o app app.o -L. -lemon
3344 To run the application:
3346 * Start the application:
3356 The application dynamically loads the instrumented user library.
3358 The instrumented user library is linked with the tracepoint provider
3359 package shared object.
3361 See the <<dlclose-warning,warning about `dlclose()`>>.
3363 image::ust-sit+app-dlopens-lib+lib-linked-with-tp-so+lib-instrumented.png[]
3366 include::../common/ust-sit-step-tp-so.txt[]
3368 To build the instrumented user library:
3370 . In path:{emon.c}, before including path:{tpp.h}, add the
3376 #define TRACEPOINT_DEFINE
3380 . Compile the user library source file:
3385 gcc -I. -fpic -c emon.c
3389 . Build the user library shared object:
3394 gcc -shared -o libemon.so emon.o -ldl -L. -ltpp
3398 To build the application:
3400 . Compile the application source file:
3409 . Build the application:
3414 gcc -o app app.o -ldl -L. -lemon
3418 To run the application:
3420 * Start the application:
3430 The application dynamically loads the instrumented user library.
3432 The instrumented user library dynamically loads the tracepoint provider
3433 package shared object.
3435 See the <<dlclose-warning,warning about `dlclose()`>>.
3437 image::ust-sit+app-dlopens-lib+lib-dlopens-tp-so+lib-instrumented.png[]
3440 include::../common/ust-sit-step-tp-so.txt[]
3442 To build the instrumented user library:
3444 . In path:{emon.c}, before including path:{tpp.h}, add the
3450 #define TRACEPOINT_DEFINE
3451 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3455 . Compile the user library source file:
3460 gcc -I. -fpic -c emon.c
3464 . Build the user library shared object:
3469 gcc -shared -o libemon.so emon.o -ldl
3473 To build the application:
3475 . Compile the application source file:
3484 . Build the application:
3489 gcc -o app app.o -ldl -L. -lemon
3493 To run the application:
3495 * Start the application:
3505 The tracepoint provider package shared object is preloaded before the
3508 The application dynamically loads the instrumented user library.
3510 image::ust-sit+tp-so-preloaded+app-dlopens-lib+lib-instrumented.png[]
3513 include::../common/ust-sit-step-tp-so.txt[]
3515 To build the instrumented user library:
3517 . In path:{emon.c}, before including path:{tpp.h}, add the
3523 #define TRACEPOINT_DEFINE
3524 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3528 . Compile the user library source file:
3533 gcc -I. -fpic -c emon.c
3537 . Build the user library shared object:
3542 gcc -shared -o libemon.so emon.o -ldl
3546 To build the application:
3548 . Compile the application source file:
3557 . Build the application:
3562 gcc -o app app.o -L. -lemon
3566 To run the application with tracing support:
3568 * Preload the tracepoint provider package shared object and
3569 start the application:
3574 LD_PRELOAD=./libtpp.so ./app
3578 To run the application without tracing support:
3580 * Start the application:
3590 The application is statically linked with the tracepoint provider
3591 package object file.
3593 The application is linked with the instrumented user library.
3595 image::ust-sit+app-linked-with-tp-o+app-linked-with-lib+lib-instrumented.png[]
3598 include::../common/ust-sit-step-tp-o.txt[]
3600 To build the instrumented user library:
3602 . In path:{emon.c}, before including path:{tpp.h}, add the
3608 #define TRACEPOINT_DEFINE
3612 . Compile the user library source file:
3617 gcc -I. -fpic -c emon.c
3621 . Build the user library shared object:
3626 gcc -shared -o libemon.so emon.o
3630 To build the application:
3632 . Compile the application source file:
3641 . Build the application:
3646 gcc -o app app.o tpp.o -llttng-ust -ldl -L. -lemon
3650 To run the instrumented application:
3652 * Start the application:
3662 The application is statically linked with the tracepoint provider
3663 package object file.
3665 The application dynamically loads the instrumented user library.
3667 image::ust-sit+app-linked-with-tp-o+app-dlopens-lib+lib-instrumented.png[]
3670 include::../common/ust-sit-step-tp-o.txt[]
3672 To build the application:
3674 . In path:{app.c}, before including path:{tpp.h}, add the following line:
3679 #define TRACEPOINT_DEFINE
3683 . Compile the application source file:
3692 . Build the application:
3697 gcc -Wl,--export-dynamic -o app app.o tpp.o \
3702 The `--export-dynamic` option passed to the linker is necessary for the
3703 dynamically loaded library to ``see'' the tracepoint symbols defined in
3706 To build the instrumented user library:
3708 . Compile the user library source file:
3713 gcc -I. -fpic -c emon.c
3717 . Build the user library shared object:
3722 gcc -shared -o libemon.so emon.o
3726 To run the application:
3728 * Start the application:
3740 .Do not use man:dlclose(3) on a tracepoint provider package
3742 Never use man:dlclose(3) on any shared object which:
3744 * Is linked with, statically or dynamically, a tracepoint provider
3746 * Calls man:dlopen(3) itself to dynamically open a tracepoint provider
3747 package shared object.
3749 This is currently considered **unsafe** due to a lack of reference
3750 counting from LTTng-UST to the shared object.
3752 A known workaround (available since glibc 2.2) is to use the
3753 `RTLD_NODELETE` flag when calling man:dlopen(3) initially. This has the
3754 effect of not unloading the loaded shared object, even if man:dlclose(3)
3757 You can also preload the tracepoint provider package shared object with
3758 the env:LD_PRELOAD environment variable to overcome this limitation.
3762 [[using-lttng-ust-with-daemons]]
3763 ===== Use noch:{LTTng-UST} with daemons
3765 If your instrumented application calls man:fork(2), man:clone(2),
3766 or BSD's man:rfork(2), without a following man:exec(3)-family
3767 system call, you must preload the path:{liblttng-ust-fork.so} shared
3768 object when starting the application.
3772 LD_PRELOAD=liblttng-ust-fork.so ./my-app
3775 If your tracepoint provider package is
3776 a shared library which you also preload, you must put both
3777 shared objects in env:LD_PRELOAD:
3781 LD_PRELOAD=liblttng-ust-fork.so:/path/to/tp.so ./my-app
3785 [[lttng-ust-pkg-config]]
3786 ===== Use noch:{pkg-config}
3788 On some distributions, LTTng-UST ships with a
3789 https://www.freedesktop.org/wiki/Software/pkg-config/[pkg-config]
3790 metadata file. If this is your case, then you can use cmd:pkg-config to
3791 build an application on the command line:
3795 gcc -o my-app my-app.o tp.o $(pkg-config --cflags --libs lttng-ust)
3799 [[instrumenting-32-bit-app-on-64-bit-system]]
3800 ===== [[advanced-instrumenting-techniques]]Build a 32-bit instrumented application for a 64-bit target system
3802 In order to trace a 32-bit application running on a 64-bit system,
3803 LTTng must use a dedicated 32-bit
3804 <<lttng-consumerd,consumer daemon>>.
3806 The following steps show how to build and install a 32-bit consumer
3807 daemon, which is _not_ part of the default 64-bit LTTng build, how to
3808 build and install the 32-bit LTTng-UST libraries, and how to build and
3809 link an instrumented 32-bit application in that context.
3811 To build a 32-bit instrumented application for a 64-bit target system,
3812 assuming you have a fresh target system with no installed Userspace RCU
3815 . Download, build, and install a 32-bit version of Userspace RCU:
3821 wget http://lttng.org/files/urcu/userspace-rcu-latest-0.9.tar.bz2 &&
3822 tar -xf userspace-rcu-latest-0.9.tar.bz2 &&
3823 cd userspace-rcu-0.9.* &&
3824 ./configure --libdir=/usr/local/lib32 CFLAGS=-m32 &&
3826 sudo make install &&
3831 . Using your distribution's package manager, or from source, install
3832 the following 32-bit versions of the following dependencies of
3833 LTTng-tools and LTTng-UST:
3836 * https://sourceforge.net/projects/libuuid/[libuuid]
3837 * http://directory.fsf.org/wiki/Popt[popt]
3838 * http://www.xmlsoft.org/[libxml2]
3841 . Download, build, and install a 32-bit version of the latest
3842 LTTng-UST{nbsp}{revision}:
3848 wget http://lttng.org/files/lttng-ust/lttng-ust-latest-2.7.tar.bz2 &&
3849 tar -xf lttng-ust-latest-2.7.tar.bz2 &&
3850 cd lttng-ust-2.7.* &&
3851 ./configure --libdir=/usr/local/lib32 \
3852 CFLAGS=-m32 CXXFLAGS=-m32 \
3853 LDFLAGS='-L/usr/local/lib32 -L/usr/lib32' &&
3855 sudo make install &&
3862 Depending on your distribution,
3863 32-bit libraries could be installed at a different location than
3864 `/usr/lib32`. For example, Debian is known to install
3865 some 32-bit libraries in `/usr/lib/i386-linux-gnu`.
3867 In this case, make sure to set `LDFLAGS` to all the
3868 relevant 32-bit library paths, for example:
3872 LDFLAGS='-L/usr/lib/i386-linux-gnu -L/usr/lib32'
3876 . Download the latest LTTng-tools{nbsp}{revision}, build, and install
3877 the 32-bit consumer daemon:
3883 wget http://lttng.org/files/lttng-tools/lttng-tools-latest-2.7.tar.bz2 &&
3884 tar -xf lttng-tools-latest-2.7.tar.bz2 &&
3885 cd lttng-tools-2.7.* &&
3886 ./configure --libdir=/usr/local/lib32 CFLAGS=-m32 CXXFLAGS=-m32 \
3887 LDFLAGS='-L/usr/local/lib32 -L/usr/lib32' &&
3889 cd src/bin/lttng-consumerd &&
3890 sudo make install &&
3895 . From your distribution or from source,
3896 <<installing-lttng,install>> the 64-bit versions of
3897 LTTng-UST and Userspace RCU.
3898 . Download, build, and install the 64-bit version of the
3899 latest LTTng-tools{nbsp}{revision}:
3905 wget http://lttng.org/files/lttng-tools/lttng-tools-latest-2.7.tar.bz2 &&
3906 tar -xf lttng-tools-latest-2.7.tar.bz2 &&
3907 cd lttng-tools-2.7.* &&
3908 ./configure --with-consumerd32-libdir=/usr/local/lib32 \
3909 --with-consumerd32-bin=/usr/local/lib32/lttng/libexec/lttng-consumerd &&
3911 sudo make install &&
3916 . Pass the following options to man:gcc(1), man:g++(1), or man:clang(1)
3917 when linking your 32-bit application:
3920 -m32 -L/usr/lib32 -L/usr/local/lib32 \
3921 -Wl,-rpath,/usr/lib32,-rpath,/usr/local/lib32
3924 For example, let's rebuild the quick start example in
3925 <<tracing-your-own-user-application,Trace a user application>> as an
3926 instrumented 32-bit application:
3931 gcc -m32 -c -I. hello-tp.c
3933 gcc -m32 -o hello hello.o hello-tp.o \
3934 -L/usr/lib32 -L/usr/local/lib32 \
3935 -Wl,-rpath,/usr/lib32,-rpath,/usr/local/lib32 \
3940 No special action is required to execute the 32-bit application and
3941 to trace it: use the command-line man:lttng(1) tool as usual.
3948 `tracef()` is a small LTTng-UST API designed for quick,
3949 man:printf(3)-like instrumentation without the burden of
3950 <<tracepoint-provider,creating>> and
3951 <<building-tracepoint-providers-and-user-application,building>>
3952 a tracepoint provider package.
3954 To use `tracef()` in your application:
3956 . In the C or C++ source files where you need to use `tracef()`,
3957 include `<lttng/tracef.h>`:
3962 #include <lttng/tracef.h>
3966 . In the application's source code, use `tracef()` like you would use
3974 tracef("my message: %d (%s)", my_integer, my_string);
3980 . Link your application with `liblttng-ust`:
3985 gcc -o app app.c -llttng-ust
3989 To trace the events that `tracef()` calls emit:
3991 * <<enabling-disabling-events,Create an event rule>> which matches the
3992 `lttng_ust_tracef:*` event name:
3997 lttng enable-event --userspace 'lttng_ust_tracef:*'
4002 .Limitations of `tracef()`
4004 The `tracef()` utility function was developed to make user space tracing
4005 super simple, albeit with notable disadvantages compared to
4006 <<defining-tracepoints,user-defined tracepoints>>:
4008 * All the emitted events have the same tracepoint provider and
4009 tracepoint names, respectively `lttng_ust_tracef` and `event`.
4010 * There is no static type checking.
4011 * The only event record field you actually get, named `msg`, is a string
4012 potentially containing the values you passed to `tracef()`
4013 using your own format string. This also means that you cannot filter
4014 events with a custom expression at run time because there are no
4016 * Since `tracef()` uses the C standard library's man:vasprintf(3)
4017 function behind the scenes to format the strings at run time, its
4018 expected performance is lower than with user-defined tracepoints,
4019 which do not require a conversion to a string.
4021 Taking this into consideration, `tracef()` is useful for some quick
4022 prototyping and debugging, but you should not consider it for any
4023 permanent and serious applicative instrumentation.
4029 ==== Use `tracelog()`
4031 The `tracelog()` API is very similar to <<tracef,`tracef()`>>, with
4032 the difference that it accepts an additional log level parameter.
4034 The goal of `tracelog()` is to ease the migration from logging to
4037 To use `tracelog()` in your application:
4039 . In the C or C++ source files where you need to use `tracelog()`,
4040 include `<lttng/tracelog.h>`:
4045 #include <lttng/tracelog.h>
4049 . In the application's source code, use `tracelog()` like you would use
4050 man:printf(3), except for the first parameter which is the log
4058 tracelog(TRACE_WARNING, "my message: %d (%s)",
4059 my_integer, my_string);
4065 See <<liblttng-ust-tracepoint-loglevel,Tracepoint log levels>> for
4066 a list of available log level names.
4068 . Link your application with `liblttng-ust`:
4073 gcc -o app app.c -llttng-ust
4077 To trace the events that `tracelog()` calls emit with a log level
4078 _as severe as_ a specific log level:
4080 * <<enabling-disabling-events,Create an event rule>> which matches the
4081 `lttng_ust_tracelog:*` event name and a minimum level
4087 lttng enable-event --userspace 'lttng_ust_tracelog:*'
4088 --loglevel=TRACE_WARNING
4092 To trace the events that `tracelog()` calls emit with a
4093 _specific log level_:
4095 * Create an event rule which matches the `lttng_ust_tracelog:*`
4096 event name and a specific log level:
4101 lttng enable-event --userspace 'lttng_ust_tracelog:*'
4102 --loglevel-only=TRACE_INFO
4107 [[prebuilt-ust-helpers]]
4108 === Prebuilt user space tracing helpers
4110 The LTTng-UST package provides a few helpers in the form or preloadable
4111 shared objects which automatically instrument system functions and
4114 The helper shared objects are normally found in dir:{/usr/lib}. If you
4115 built LTTng-UST <<building-from-source,from source>>, they are probably
4116 located in dir:{/usr/local/lib}.
4118 The installed user space tracing helpers in LTTng-UST{nbsp}{revision}
4121 path:{liblttng-ust-libc-wrapper.so}::
4122 path:{liblttng-ust-pthread-wrapper.so}::
4123 <<liblttng-ust-libc-pthread-wrapper,C{nbsp}standard library
4124 memory and POSIX threads function tracing>>.
4126 path:{liblttng-ust-cyg-profile.so}::
4127 path:{liblttng-ust-cyg-profile-fast.so}::
4128 <<liblttng-ust-cyg-profile,Function entry and exit tracing>>.
4130 path:{liblttng-ust-dl.so}::
4131 <<liblttng-ust-dl,Dynamic linker tracing>>.
4133 To use a user space tracing helper with any user application:
4135 * Preload the helper shared object when you start the application:
4140 LD_PRELOAD=liblttng-ust-libc-wrapper.so my-app
4144 You can preload more than one helper:
4149 LD_PRELOAD=liblttng-ust-libc-wrapper.so:liblttng-ust-dl.so my-app
4155 [[liblttng-ust-libc-pthread-wrapper]]
4156 ==== Instrument C standard library memory and POSIX threads functions
4158 The path:{liblttng-ust-libc-wrapper.so} and
4159 path:{liblttng-ust-pthread-wrapper.so} helpers
4160 add instrumentation to some C standard library and POSIX
4164 .Functions instrumented by preloading path:{liblttng-ust-libc-wrapper.so}.
4166 |TP provider name |TP name |Instrumented function
4168 .6+|`lttng_ust_libc` |`malloc` |man:malloc(3)
4169 |`calloc` |man:calloc(3)
4170 |`realloc` |man:realloc(3)
4171 |`free` |man:free(3)
4172 |`memalign` |man:memalign(3)
4173 |`posix_memalign` |man:posix_memalign(3)
4177 .Functions instrumented by preloading path:{liblttng-ust-pthread-wrapper.so}.
4179 |TP provider name |TP name |Instrumented function
4181 .4+|`lttng_ust_pthread` |`pthread_mutex_lock_req` |man:pthread_mutex_lock(3p) (request time)
4182 |`pthread_mutex_lock_acq` |man:pthread_mutex_lock(3p) (acquire time)
4183 |`pthread_mutex_trylock` |man:pthread_mutex_trylock(3p)
4184 |`pthread_mutex_unlock` |man:pthread_mutex_unlock(3p)
4187 When you preload the shared object, it replaces the functions listed
4188 in the previous tables by wrappers which contain tracepoints and call
4189 the replaced functions.
4192 [[liblttng-ust-cyg-profile]]
4193 ==== Instrument function entry and exit
4195 The path:{liblttng-ust-cyg-profile*.so} helpers can add instrumentation
4196 to the entry and exit points of functions.
4198 man:gcc(1) and man:clang(1) have an option named
4199 https://gcc.gnu.org/onlinedocs/gcc/Code-Gen-Options.html[`-finstrument-functions`]
4200 which generates instrumentation calls for entry and exit to functions.
4201 The LTTng-UST function tracing helpers,
4202 path:{liblttng-ust-cyg-profile.so} and
4203 path:{liblttng-ust-cyg-profile-fast.so}, take advantage of this feature
4204 to add tracepoints to the two generated functions (which contain
4205 `cyg_profile` in their names, hence the helper's name).
4207 To use the LTTng-UST function tracing helper, the source files to
4208 instrument must be built using the `-finstrument-functions` compiler
4211 There are two versions of the LTTng-UST function tracing helper:
4213 * **path:{liblttng-ust-cyg-profile-fast.so}** is a lightweight variant
4214 that you should only use when it can be _guaranteed_ that the
4215 complete event stream is recorded without any lost event record.
4216 Any kind of duplicate information is left out.
4218 This version contains the following tracepoints:
4222 .Points instrumented by preloading path:{liblttng-ust-cyg-profile-fast.so}.
4224 |TP provider name |TP name |Instrumented points
4226 .2+|`lttng_ust_cyg_profile_fast`
4232 Address of called function.
4239 Assuming no event record is lost, having only the function addresses on
4240 entry is enough to create a call graph, since an event record always
4241 contains the ID of the CPU that generated it.
4243 You can use a tool like
4244 https://sourceware.org/binutils/docs/binutils/addr2line.html[cmd:addr2line]
4245 to convert function addresses back to source file names and
4248 * **path:{liblttng-ust-cyg-profile.so}** is a more robust variant
4249 which also works in use cases where event records might get discarded or
4250 not recorded from application startup.
4251 In these cases, the trace analyzer needs more information to be
4252 able to reconstruct the program flow.
4254 This version contains the following tracepoints:
4258 .Points instrumented by preloading path:{liblttng-ust-cyg-profile.so}.
4260 |TP provider name |TP name |Instrumented point
4262 .2+|`lttng_ust_cyg_profile`
4268 Address of called function.
4277 Address of called function.
4284 TIP: It's sometimes a good idea to limit the number of source files that
4285 you compile with the `-finstrument-functions` option to prevent LTTng
4286 from writing an excessive amount of trace data at run time. When using
4287 man:gcc(1), you can use the
4288 `-finstrument-functions-exclude-function-list` option to avoid
4289 instrument entries and exits of specific function names.
4291 All the tracepoints that this helper contains have the
4292 <<liblttng-ust-tracepoint-loglevel,log level>> `TRACE_DEBUG_FUNCTION`.
4297 ==== Instrument the dynamic linker
4299 The path:{liblttng-ust-dl.so} helper adds instrumentation to the
4300 man:dlopen(3) and man:dlclose(3) function calls.
4303 .Functions instrumented by preloading path:{liblttng-ust-dl.so}.
4305 |TP provider name |TP name |Instrumented function
4313 Memory base address (where the dynamic linker placed the shared
4317 File system path to the loaded shared object.
4320 File size of the the loaded shared object.
4323 Last modification time (seconds since Epoch time) of the loaded shared
4330 Memory base address (where the dynamic linker placed the shared
4336 [[java-application]]
4337 === User space Java agent
4339 You can instrument a Java application which uses one of the following
4342 * The https://docs.oracle.com/javase/7/docs/api/java/util/logging/package-summary.html[**`java.util.logging`**]
4343 (JUL) core logging facilities.
4344 * http://logging.apache.org/log4j/1.2/[**Apache log4j 1.2**], since
4345 LTTng 2.6. Note that Apache Log4j{nbsp}2 is not supported.
4347 Each log statement emits an LTTng event once the
4348 application initializes the <<lttng-ust-agents,LTTng-UST Java agent>>
4352 .LTTng-UST Java agent imported by a Java application.
4353 image::java-app.png[]
4355 NOTE: We use http://openjdk.java.net/[OpenJDK] 7 for development and
4356 https://ci.lttng.org/[continuous integration], thus this version is
4357 directly supported. However, the LTTng-UST Java agent is also
4358 tested with OpenJDK 6.
4360 To use the LTTng-UST Java agent:
4362 . In the Java application's source code, import the LTTng-UST Java
4368 import org.lttng.ust.agent.LTTngAgent;
4372 . As soon as possible after the entry point of the application,
4373 initialize the LTTng-UST Java agent:
4378 LTTngAgent lttngAgent = LTTngAgent.getLTTngAgent();
4382 Any log statement that the application executes before this
4383 initialization does not emit an LTTng event.
4385 . Use `java.util.logging` and/or log4j log statements and configuration
4386 as usual. Since the LTTng-UST Java agent adds a handler to the _root_
4387 loggers, you can trace any log statement from any logger.
4389 . Before exiting the application, dispose the LTTng-UST Java agent:
4394 lttngAgent.dispose();
4398 This is not strictly necessary, but it is recommended for a clean
4399 disposal of the agent's resources.
4401 Any log statement that the application executes after this disposal does
4402 not emit an LTTng event.
4404 . Include the LTTng-UST Java agent's JAR file, path:{liblttng-ust-agent.jar},
4406 https://docs.oracle.com/javase/tutorial/essential/environment/paths.html[class path]
4407 when building the Java application.
4409 path:{liblttng-ust-agent.jar} is typically located in
4410 dir:{/usr/share/java}.
4412 IMPORTANT: The LTTng-UST Java agent must be
4413 <<installing-lttng,installed>> for the logging framework your
4416 .[[jul]]Use the LTTng-UST Java agent with `java.util.logging`.
4421 import java.util.logging.Logger;
4422 import org.lttng.ust.agent.LTTngAgent;
4426 private static final int answer = 42;
4428 public static void main(String[] argv) throws Exception
4431 Logger logger = Logger.getLogger("jello");
4433 // Call this as soon as possible (before logging)
4434 LTTngAgent lttngAgent = LTTngAgent.getLTTngAgent();
4437 logger.info("some info");
4438 logger.warning("some warning");
4440 logger.finer("finer information; the answer is " + answer);
4442 logger.severe("error!");
4444 // Not mandatory, but cleaner
4445 lttngAgent.dispose();
4450 You can build this example like this:
4454 javac -cp /usr/share/java/liblttng-ust-agent.jar Test.java
4457 You can run the compiled class like this:
4461 java -cp /usr/share/java/liblttng-ust-agent.jar:. Test
4465 .[[log4j]]Use the LTTng-UST Java agent with Apache log4j 1.2.
4470 import org.apache.log4j.Logger;
4471 import org.apache.log4j.BasicConfigurator;
4472 import org.lttng.ust.agent.LTTngAgent;
4476 private static final int answer = 42;
4478 public static void main(String[] argv) throws Exception
4480 // Create and configure a logger
4481 Logger logger = Logger.getLogger(Test.class);
4482 BasicConfigurator.configure();
4484 // Call this as soon as possible (before logging)
4485 LTTngAgent lttngAgent = LTTngAgent.getLTTngAgent();
4488 logger.info("some info");
4489 logger.warn("some warning");
4491 logger.debug("debug information; the answer is " + answer);
4493 logger.error("error!");
4494 logger.fatal("fatal error!");
4496 // Not mandatory, but cleaner
4497 lttngAgent.dispose();
4502 You can build this example like this:
4506 javac -cp /usr/share/java/liblttng-ust-agent.jar:$LOG4JCP Test.java
4509 where `$LOG4JCP` is the path to log4j's JAR file.
4511 You can run the compiled class like this:
4515 java -cp /usr/share/java/liblttng-ust-agent.jar:$LOG4JCP:. Test
4519 When you <<enabling-disabling-events,create an event rule>>, use the
4520 `--jul` (`java.util.logging`) or `--log4j` (log4j) option to target
4522 <<domain,tracing domain>>. You can also use the `--loglevel` or
4523 `--loglevel-only` option to target a range of JUL/log4j log levels or a
4524 specific JUL/log4j log level.
4528 [[python-application]]
4529 === User space Python agent
4531 You can instrument a Python 2 or Python 3 application which uses the
4532 standard https://docs.python.org/3/library/logging.html[`logging`]
4535 Each log statement emits an LTTng event once the
4536 application module imports the
4537 <<lttng-ust-agents,LTTng-UST Python agent>> package.
4540 .A Python application importing the LTTng-UST Python agent.
4541 image::python-app.png[]
4543 To use the LTTng-UST Python agent:
4545 . In the Python application's source code, import the LTTng-UST Python
4555 The LTTng-UST Python agent automatically adds its logging handler to the
4556 root logger at import time.
4558 Any log statement that the application executes before this import does
4559 not emit an LTTng event.
4561 IMPORTANT: The LTTng-UST Python agent must be
4562 <<installing-lttng,installed>>.
4564 . Use log statements and logging configuration as usual.
4565 Since the LTTng-UST Python agent adds a handler to the _root_
4566 logger, you can trace any log statement from any logger.
4568 .Use the LTTng-UST Python agent.
4578 logging.basicConfig()
4579 logger = logging.getLogger('my-logger')
4582 logger.debug('debug message')
4583 logger.info('info message')
4584 logger.warn('warn message')
4585 logger.error('error message')
4586 logger.critical('critical message')
4590 if __name__ == '__main__':
4594 NOTE: `logging.basicConfig()`, which adds to the root logger a basic
4595 logging handler which prints to the standard error stream, is not
4596 strictly required for LTTng-UST tracing to work, but in versions of
4597 Python preceding 3.2, you could see a warning message which indicates
4598 that no handler exists for the logger `my-logger`.
4601 When you <<enabling-disabling-events,create an event rule>>, use the
4602 `--python` option to target the Python
4603 <<domain,tracing domain>>. You can also use
4604 the `--loglevel` or `--loglevel-only` option to target a range of
4605 Python log levels or a specific Python log level.
4607 When an application imports the LTTng-UST Python agent, the agent tries
4608 to register to a <<lttng-sessiond,session daemon>>. Note that you must
4609 start the session daemon _before_ you start the Python application.
4610 If a session daemon is found, the agent tries to register to it
4611 during 5{nbsp}seconds, after which the application continues without
4612 LTTng tracing support. You can override this timeout value with the
4613 env:LTTNG_UST_PYTHON_REGISTER_TIMEOUT environment variable
4616 If the session daemon stops while a Python application with an imported
4617 LTTng-UST Python agent runs, the agent retries to connect and to
4618 register to a session daemon every 3{nbsp}seconds. You can override this
4619 delay with the env:LTTNG_UST_PYTHON_REGISTER_RETRY_DELAY environment
4624 [[proc-lttng-logger-abi]]
4627 The `lttng-tracer` Linux kernel module, part of
4628 <<lttng-modules,LTTng-modules>>, creates the special LTTng logger file
4629 path:{/proc/lttng-logger} when it's loaded. Any application can write
4630 text data to this file to emit an LTTng event.
4633 .An application writes to the LTTng logger file to emit an LTTng event.
4634 image::lttng-logger.png[]
4636 The LTTng logger is the quickest method--not the most efficient,
4637 however--to add instrumentation to an application. It is designed
4638 mostly to instrument shell scripts:
4642 echo "Some message, some $variable" > /proc/lttng-logger
4645 Any event that the LTTng logger emits is named `lttng_logger` and
4646 belongs to the Linux kernel <<domain,tracing domain>>. However, unlike
4647 other instrumentation points in the kernel tracing domain, **any Unix
4648 user** can <<enabling-disabling-events,create an event rule>> which
4649 matches its event name, not only the root user or users in the tracing
4652 To use the LTTng logger:
4654 * From any application, write text data to the path:{/proc/lttng-logger}
4657 The `msg` field of `lttng_logger` event records contains the
4660 NOTE: The maximum message length of an LTTng logger event is
4661 1024{nbsp}bytes. Writing more than this makes the LTTng logger emit more
4662 than one event to contain the remaining data.
4664 You should not use the LTTng logger to trace a user application which
4665 can be instrumented in a more efficient way, namely:
4667 * <<c-application,C and $$C++$$ applications>>.
4668 * <<java-application,Java applications>>.
4669 * <<python-application,Python applications>>.
4672 [[instrumenting-linux-kernel]]
4673 === LTTng kernel tracepoints
4675 NOTE: This section shows how to _add_ instrumentation points to the
4676 Linux kernel. The kernel's subsystems are already thoroughly
4677 instrumented at strategic places for LTTng when you
4678 <<installing-lttng,install>> the <<lttng-modules,LTTng-modules>>
4682 There are two methods to instrument the Linux kernel:
4684 . <<linux-add-lttng-layer,Add an LTTng layer>> over an existing ftrace
4685 tracepoint which uses the `TRACE_EVENT()` API.
4687 Choose this if you want to instrumentation a Linux kernel tree with an
4688 instrumentation point compatible with ftrace, perf, and SystemTap.
4690 . Use an <<linux-lttng-tracepoint-event,LTTng-only approach>> to
4691 instrument an out-of-tree kernel module.
4693 Choose this if you don't need ftrace, perf, or SystemTap support.
4697 [[linux-add-lttng-layer]]
4698 ==== [[instrumenting-linux-kernel-itself]][[mainline-trace-event]][[lttng-adaptation-layer]]Add an LTTng layer to an existing ftrace tracepoint
4700 This section shows how to add an LTTng layer to existing ftrace
4701 instrumentation using the `TRACE_EVENT()` API.
4703 This section does not document the `TRACE_EVENT()` macro. You can
4704 read the following articles to learn more about this API:
4706 * http://lwn.net/Articles/379903/[Using the TRACE_EVENT() macro (Part 1)]
4707 * http://lwn.net/Articles/381064/[Using the TRACE_EVENT() macro (Part 2)]
4708 * http://lwn.net/Articles/383362/[Using the TRACE_EVENT() macro (Part 3)]
4710 The following procedure assumes that your ftrace tracepoints are
4711 correctly defined in their own header and that they are created in
4712 one source file using the `CREATE_TRACE_POINTS` definition.
4714 To add an LTTng layer over an existing ftrace tracepoint:
4716 . Make sure the following kernel configuration options are
4722 * `CONFIG_HIGH_RES_TIMERS`
4723 * `CONFIG_TRACEPOINTS`
4726 . Build the Linux source tree with your custom ftrace tracepoints.
4727 . Boot the resulting Linux image on your target system.
4729 Confirm that the tracepoints exist by looking for their names in the
4730 dir:{/sys/kernel/debug/tracing/events/subsys} directory, where `subsys`
4731 is your subsystem's name.
4733 . Get a copy of the latest LTTng-modules{nbsp}{revision}:
4739 wget http://lttng.org/files/lttng-modules/lttng-modules-latest-2.8.tar.bz2 &&
4740 tar -xf lttng-modules-latest-2.8.tar.bz2 &&
4741 cd lttng-modules-2.8.*
4745 . In dir:{instrumentation/events/lttng-module}, relative to the root
4746 of the LTTng-modules source tree, create a header file named
4747 +__subsys__.h+ for your custom subsystem +__subsys__+ and write your
4748 LTTng-modules tracepoint definitions using the LTTng-modules
4751 Start with this template:
4755 .path:{instrumentation/events/lttng-module/my_subsys.h}
4758 #define TRACE_SYSTEM my_subsys
4760 #if !defined(_LTTNG_MY_SUBSYS_H) || defined(TRACE_HEADER_MULTI_READ)
4761 #define _LTTNG_MY_SUBSYS_H
4763 #include "../../../probes/lttng-tracepoint-event.h"
4764 #include <linux/tracepoint.h>
4766 LTTNG_TRACEPOINT_EVENT(
4768 * Format is identical to TRACE_EVENT()'s version for the three
4769 * following macro parameters:
4772 TP_PROTO(int my_int, const char *my_string),
4773 TP_ARGS(my_int, my_string),
4775 /* LTTng-modules specific macros */
4777 ctf_integer(int, my_int_field, my_int)
4778 ctf_string(my_bar_field, my_bar)
4782 #endif /* !defined(_LTTNG_MY_SUBSYS_H) || defined(TRACE_HEADER_MULTI_READ) */
4784 #include "../../../probes/define_trace.h"
4788 The entries in the `TP_FIELDS()` section are the list of fields for the
4789 LTTng tracepoint. This is similar to the `TP_STRUCT__entry()` part of
4790 ftrace's `TRACE_EVENT()` macro.
4792 See <<lttng-modules-tp-fields,Tracepoint fields macros>> for a
4793 complete description of the available `ctf_*()` macros.
4795 . Create the LTTng-modules probe's kernel module C source file,
4796 +probes/lttng-probe-__subsys__.c+, where +__subsys__+ is your
4801 .path:{probes/lttng-probe-my-subsys.c}
4803 #include <linux/module.h>
4804 #include "../lttng-tracer.h"
4807 * Build-time verification of mismatch between mainline
4808 * TRACE_EVENT() arguments and the LTTng-modules adaptation
4809 * layer LTTNG_TRACEPOINT_EVENT() arguments.
4811 #include <trace/events/my_subsys.h>
4813 /* Create LTTng tracepoint probes */
4814 #define LTTNG_PACKAGE_BUILD
4815 #define CREATE_TRACE_POINTS
4816 #define TRACE_INCLUDE_PATH ../instrumentation/events/lttng-module
4818 #include "../instrumentation/events/lttng-module/my_subsys.h"
4820 MODULE_LICENSE("GPL and additional rights");
4821 MODULE_AUTHOR("Your name <your-email>");
4822 MODULE_DESCRIPTION("LTTng my_subsys probes");
4823 MODULE_VERSION(__stringify(LTTNG_MODULES_MAJOR_VERSION) "."
4824 __stringify(LTTNG_MODULES_MINOR_VERSION) "."
4825 __stringify(LTTNG_MODULES_PATCHLEVEL_VERSION)
4826 LTTNG_MODULES_EXTRAVERSION);
4830 . Edit path:{probes/Makefile} and add your new kernel module object
4831 next to the existing ones:
4835 .path:{probes/Makefile}
4839 obj-m += lttng-probe-module.o
4840 obj-m += lttng-probe-power.o
4842 obj-m += lttng-probe-my-subsys.o
4848 . Build and install the LTTng kernel modules:
4853 make KERNELDIR=/path/to/linux
4854 sudo make modules_install
4858 Replace `/path/to/linux` with the path to the Linux source tree where
4859 you defined and used tracepoints with ftrace's `TRACE_EVENT()` macro.
4861 Note that you can also use the
4862 <<lttng-tracepoint-event-code,`LTTNG_TRACEPOINT_EVENT_CODE()` macro>>
4863 instead of `LTTNG_TRACEPOINT_EVENT()` to use custom local variables and
4864 C code that need to be executed before the event fields are recorded.
4866 The best way to learn how to use the previous LTTng-modules macros is to
4867 inspect the existing LTTng-modules tracepoint definitions in the
4868 dir:{instrumentation/events/lttng-module} header files. Compare them
4869 with the Linux kernel mainline versions in the
4870 dir:{include/trace/events} directory of the Linux source tree.
4874 [[lttng-tracepoint-event-code]]
4875 ===== Use custom C code to access the data for tracepoint fields
4877 Although we recommended to always use the
4878 <<lttng-adaptation-layer,`LTTNG_TRACEPOINT_EVENT()`>> macro to describe
4879 the arguments and fields of an LTTng-modules tracepoint when possible,
4880 sometimes you need a more complex process to access the data that the
4881 tracer records as event record fields. In other words, you need local
4882 variables and multiple C{nbsp}statements instead of simple
4883 argument-based expressions that you pass to the
4884 <<lttng-modules-tp-fields,`ctf_*()` macros of `TP_FIELDS()`>>.
4886 You can use the `LTTNG_TRACEPOINT_EVENT_CODE()` macro instead of
4887 `LTTNG_TRACEPOINT_EVENT()` to declare custom local variables and define
4888 a block of C{nbsp}code to be executed before LTTng records the fields.
4889 The structure of this macro is:
4892 .`LTTNG_TRACEPOINT_EVENT_CODE()` macro syntax.
4894 LTTNG_TRACEPOINT_EVENT_CODE(
4896 * Format identical to the LTTNG_TRACEPOINT_EVENT()
4897 * version for the following three macro parameters:
4900 TP_PROTO(int my_int, const char *my_string),
4901 TP_ARGS(my_int, my_string),
4903 /* Declarations of custom local variables */
4906 unsigned long b = 0;
4907 const char *name = "(undefined)";
4908 struct my_struct *my_struct;
4912 * Custom code which uses both tracepoint arguments
4913 * (in TP_ARGS()) and local variables (in TP_locvar()).
4915 * Local variables are actually members of a structure pointed
4916 * to by the special variable tp_locvar.
4920 tp_locvar->a = my_int + 17;
4921 tp_locvar->my_struct = get_my_struct_at(tp_locvar->a);
4922 tp_locvar->b = my_struct_compute_b(tp_locvar->my_struct);
4923 tp_locvar->name = my_struct_get_name(tp_locvar->my_struct);
4924 put_my_struct(tp_locvar->my_struct);
4933 * Format identical to the LTTNG_TRACEPOINT_EVENT()
4934 * version for this, except that tp_locvar members can be
4935 * used in the argument expression parameters of
4936 * the ctf_*() macros.
4939 ctf_integer(unsigned long, my_struct_b, tp_locvar->b)
4940 ctf_integer(int, my_struct_a, tp_locvar->a)
4941 ctf_string(my_string_field, my_string)
4942 ctf_string(my_struct_name, tp_locvar->name)
4947 IMPORTANT: The C code defined in `TP_code()` must not have any side
4948 effects when executed. In particular, the code must not allocate
4949 memory or get resources without deallocating this memory or putting
4950 those resources afterwards.
4953 [[instrumenting-linux-kernel-tracing]]
4954 ==== Load and unload a custom probe kernel module
4956 You must load a <<lttng-adaptation-layer,created LTTng-modules probe
4957 kernel module>> in the kernel before it can emit LTTng events.
4959 To load the default probe kernel modules and a custom probe kernel
4962 * Use the `--extra-kmod-probes` option to give extra probe modules
4963 to load when starting a root <<lttng-sessiond,session daemon>>:
4966 .Load the `my_subsys`, `usb`, and the default probe modules.
4970 sudo lttng-sessiond --extra-kmod-probes=my_subsys,usb
4975 You only need to pass the subsystem name, not the whole kernel module
4978 To load _only_ a given custom probe kernel module:
4980 * Use the `--kmod-probes` option to give the probe modules
4981 to load when starting a root session daemon:
4984 .Load only the `my_subsys` and `usb` probe modules.
4988 sudo lttng-sessiond --kmod-probes=my_subsys,usb
4993 To confirm that a probe module is loaded:
5000 lsmod | grep lttng_probe_usb
5004 To unload the loaded probe modules:
5006 * Kill the session daemon with `SIGTERM`:
5011 sudo pkill lttng-sessiond
5015 You can also use man:modprobe(8)'s `--remove` option if the session
5016 daemon terminates abnormally.
5019 [[controlling-tracing]]
5022 Once an application or a Linux kernel is
5023 <<instrumenting,instrumented>> for LTTng tracing,
5026 This section is divided in topics on how to use the various
5027 <<plumbing,components of LTTng>>, in particular the <<lttng-cli,cmd:lttng
5028 command-line tool>>, to _control_ the LTTng daemons and tracers.
5030 Note that the <<online-lttng-manpages,Online LTTng man pages>> are
5031 more comprehensive than the guides of this section. Refer to them if
5032 your use case is not included in this section.
5036 === Start a session daemon
5038 In some situations, you need to run a <<lttng-sessiond,session daemon>>
5039 _before_ you can use the cmd:lttng command-line tool.
5041 You will see the following error when you run a command while no session
5045 Error: No session daemon is available
5048 The only command that automatically runs a session daemon is `create`,
5049 which you use to <<creating-destroying-tracing-sessions,create a tracing
5050 session>>. While this is most of the time the first operation that you
5051 do, sometimes it's not. Some examples are:
5053 * <<list-instrumentation-points,List the available instrumentation points>>.
5054 * <<saving-loading-tracing-session,Load a tracing session configuration>>.
5056 [[tracing-group]] Each Unix user must have its own running session
5057 daemon to trace user applications. The session daemon that the root user
5058 starts is the only one allowed to control the LTTng kernel tracer. Users
5059 that are part of the _tracing group_ can control the root session
5060 daemon. The default tracing group name is `tracing`; you can set it to
5061 something else with the `--group` option when you start the root session
5064 To start a user session daemon:
5066 * Run cmd:lttng-sessiond:
5071 lttng-sessiond --daemonize
5075 To start the root session daemon:
5077 * Run cmd:lttng-sessiond as the root user:
5082 sudo lttng-sessiond --daemonize
5086 In both cases, remove the `--daemonize` option to start the session
5087 daemon in foreground.
5089 To stop a session daemon, use cmd:kill on its process ID (standard
5092 Note that some Linux distributions could manage the LTTng session daemon
5093 as a service. In this case, you should use the service manager to
5094 start, restart, and stop session daemons.
5097 [[creating-destroying-tracing-sessions]]
5098 === Create and destroy a tracing session
5100 Almost all the LTTng control operations happen in the scope of
5101 a <<tracing-session,tracing session>>, which is the dialogue between the
5102 <<lttng-sessiond,session daemon>> and you.
5104 To create a tracing session with a generated name:
5106 * Use the `create` command:
5115 The created tracing session's name is `auto` followed by the
5118 To create a tracing session with a specific name:
5120 * Use the optional argument of the `create` command:
5125 lttng create my-session
5129 Replace `my-session` with the specific tracing session name.
5131 LTTng appends the creation date to the created tracing session's name.
5133 LTTng writes the traces of a tracing session in
5134 +$LTTNG_HOME/lttng-trace/__name__+ by default, where +__name__+ is the
5135 name of the tracing session. Note that the env:LTTNG_HOME environment
5136 variable defaults to `$HOME` if not set.
5138 To output LTTng traces to a non-default location:
5140 * Use the `--output` option of the `create` command:
5145 lttng create --output=/tmp/some-directory my-session
5149 You may create as many tracing sessions as you wish.
5151 To list all the existing tracing sessions for your Unix user:
5153 * Use the `list` command:
5162 When you create a tracing session, it is set as the _current tracing
5163 session_. The following man:lttng(1) commands operate on the current
5164 tracing session when you don't specify one:
5166 [role="list-3-cols"]
5182 To change the current tracing session:
5184 * Use the `set-session` command:
5189 lttng set-session new-session
5193 Replace `new-session` by the name of the new current tracing session.
5195 When you are done tracing in a given tracing session, you can destroy
5196 it. This operation frees the resources taken by the tracing session
5197 to destroy; it does not destroy the trace data that LTTng wrote for
5198 this tracing session.
5200 To destroy the current tracing session:
5202 * Use the `destroy` command:
5212 [[list-instrumentation-points]]
5213 === List the available instrumentation points
5215 The <<lttng-sessiond,session daemon>> can query the running instrumented
5216 user applications and the Linux kernel to get a list of available
5217 instrumentation points. For the Linux kernel <<domain,tracing domain>>,
5218 they are tracepoints and system calls. For the user space tracing
5219 domain, they are tracepoints. For the other tracing domains, they are
5222 To list the available instrumentation points:
5224 * Use the `list` command with the requested tracing domain's option
5228 * `--kernel`: Linux kernel tracepoints (your Unix user must be a root
5229 user, or it must be a member of the tracing group).
5230 * `--kernel --syscall`: Linux kernel system calls (your Unix user must
5231 be a root user, or it must be a member of the tracing group).
5232 * `--userspace`: user space tracepoints.
5233 * `--jul`: `java.util.logging` loggers.
5234 * `--log4j`: Apache log4j loggers.
5235 * `--python`: Python loggers.
5238 .List the available user space tracepoints.
5242 lttng list --userspace
5246 .List the available Linux kernel system call tracepoints.
5250 lttng list --kernel --syscall
5255 [[enabling-disabling-events]]
5256 === Create and enable an event rule
5258 Once you <<creating-destroying-tracing-sessions,create a tracing
5259 session>>, you can create <<event,event rules>> with the
5260 `enable-event` command.
5262 You specify each condition with a command-line option. The available
5263 condition options are shown in the following table.
5265 [role="growable",cols="asciidoc,asciidoc,default"]
5266 .Condition command-line options for the `enable-event` command.
5268 |Option |Description |Applicable tracing domains
5274 . +--probe=__ADDR__+
5275 . +--function=__ADDR__+
5278 Instead of using the default _tracepoint_ instrumentation type, use:
5280 . A Linux system call.
5281 . A Linux https://lwn.net/Articles/132196/[KProbe] (symbol or address).
5282 . The entry and return points of a Linux function (symbol or address).
5286 |First positional argument.
5289 Tracepoint or system call name. In the case of a Linux KProbe or
5290 function, this is a custom name given to the event rule. With the
5291 JUL, log4j, and Python domains, this is a logger name.
5293 With a tracepoint, logger, or system call name, the last character
5294 can be `*` to match anything that remains.
5301 . +--loglevel=__LEVEL__+
5302 . +--loglevel-only=__LEVEL__+
5305 . Match only tracepoints or log statements with a logging level at
5306 least as severe as +__LEVEL__+.
5307 . Match only tracepoints or log statements with a logging level
5308 equal to +__LEVEL__+.
5310 You can get the list of available logging level names with
5311 `lttng enable-event --help`.
5313 |User space, JUL, log4j, and Python.
5315 |+--exclude=__EXCLUSIONS__+
5318 When you use a `*` character at the end of the tracepoint or logger
5319 name (first positional argument), exclude the specific names in the
5320 comma-delimited list +__EXCLUSIONS__+.
5323 User space, JUL, log4j, and Python.
5325 |+--filter=__EXPR__+
5328 Match only events which satisfy the expression +__EXPR__+.
5330 +__EXPR__+ is a C-like logical expression where identifiers are event
5331 fields (preceded with `$ctx.` for context fields). Nested expressions
5332 with `(` and `)`, and all the logical and comparison operators of the C
5333 language are supported. The precedence rules of those operators are the
5334 same as in the C language.
5336 When a comparison includes a non-existent event field, the whole filter
5337 expression evaluates to false.
5339 C integer and floating point number constants are supported, as well as
5340 literal strings between double quotes (`"`). Literal strings can
5341 contain a wildcard character (`*`) at the end to match anything that
5342 remains. This wildcard can be escaped using `\*`.
5344 Note that, although it is possible to use this option with the JUL,
5345 log4j, and Python tracing domains, the tracer evalutes the expression
5346 against the equivalent user space event.
5353 for more details about those command-line options.
5355 You attach an event rule to a <<channel,channel>> on creation. If you
5356 do not specify the channel with the `--channel` option, and if the event
5357 rule to create is the first in its <<domain,tracing domain>> for a given
5358 tracing session, then LTTng creates a _default channel_ for you. This
5359 default channel is reused in subsequent invocations of the
5360 `enable-event` command for the same tracing domain.
5362 An event rule is always enabled at creation time.
5364 The following examples show how you can combine the previous
5365 command-line options to create simple to more complex event rules.
5367 .Create an event rule targetting a Linux kernel tracepoint (default channel).
5371 lttng enable-event --kernel sched_switch
5375 .Create an event rule matching four Linux kernel system calls (default channel).
5379 lttng enable-event --kernel --syscall open,write,read,close
5383 .Create an event rule matching a Linux kernel tracepoint with a filter expression (default channel).
5387 lttng enable-event --kernel sched_switch --filter='prev_comm == "bash"'
5390 IMPORTANT: Make sure to always quote the filter string when you
5391 use man:lttng(1) from a shell.
5394 .Create an event rule matching any user space tracepoint of a given tracepoint provider with a log level range (default channel).
5398 lttng enable-event --userspace my_app:'*' --loglevel=TRACE_INFO
5401 IMPORTANT: Make sure to always quote the wildcard character when you
5402 use man:lttng(1) from a shell.
5405 .Create an event rule matching multiple Python loggers with a wildcard and with exclusions (default channel).
5409 lttng enable-event --python my-app.'*' \
5410 --exclude='my-app.module,my-app.hello'
5414 .Create an event rule matching any Apache log4j logger with a specific log level (default channel).
5418 lttng enable-event --log4j --all --loglevel-only=LOG4J_WARN
5422 .Create an event rule attached to a specific channel matching a specific user space tracepoint provider and tracepoint.
5426 lttng enable-event --userspace my_app:my_tracepoint --channel=my-channel
5430 The event rules of a given channel form a whitelist: as soon as an
5431 emitted event passes one of them, LTTng can record the event. For
5432 example, an event named `my_app:my_tracepoint` emitted from a user space
5433 tracepoint with a `TRACE_ERROR` log level passes both of the following
5438 lttng enable-event --userspace my_app:my_tracepoint
5439 lttng enable-event --userspace my_app:my_tracepoint \
5440 --loglevel=TRACE_INFO
5443 The second event rule is redundant: the first one includes
5447 [[disable-event-rule]]
5448 === Disable an event rule
5450 To disable an event rule that you <<enabling-disabling-events,created>>
5451 previously, use the `disable-event` command. This command disables _all_
5452 the event rules (of a given tracing domain and channel) which match an
5453 instrumentation point. The other conditions are not supported as of
5454 LTTng{nbsp}{revision}.
5456 The LTTng tracer does not record an emitted event which passes
5457 a _disabled_ event rule.
5459 .Disable an event rule matching a Python logger (default channel).
5463 lttng disable-event --python my-logger
5467 .Disable an event rule matching all `java.util.logging` loggers (default channel).
5471 lttng disable-event --jul '*'
5475 .Disable _all_ the event rules of the default channel.
5477 The `--all-events` option is not, like the `--all` option of
5478 `enable-event`, the equivalent of the event name `*` (wildcard): it
5479 disables _all_ the event rules of a given channel.
5483 lttng disable-event --jul --all-events
5487 NOTE: You cannot delete an event rule once you create it.
5491 === Get the status of a tracing session
5493 To get the status of a tracing session, that is, its channels, event
5494 rules, and their attributes:
5496 * Use the `list` command with the tracing session's name:
5501 lttng list my-session
5505 Replace `my-session` with your tracing session's name.
5508 [[basic-tracing-session-control]]
5509 === Start and stop a tracing session
5511 Once you <<creating-destroying-tracing-sessions,create a tracing
5513 <<enabling-disabling-events,create one or more event rules>>,
5514 you can start and stop the tracers for this tracing session.
5516 To start tracing in the current tracing session:
5518 * Use the `start` command:
5527 To stop tracing in the current tracing session:
5529 * Use the `stop` command:
5538 LTTng is very flexible: you can launch user applications before
5539 or after the you start the tracers. The tracers only record the events
5540 if they pass enabled event rules and if they occur while the tracers are
5544 [[enabling-disabling-channels]]
5545 === Create a channel
5547 Once you create a tracing session, you can create a <<channel,channel>>
5548 with the `enable-channel` command.
5550 Note that LTTng automatically creates a default channel when, for a
5551 given <<domain,tracing domain>>, no channels exist and you
5552 <<enabling-disabling-events,create>> the first event rule. This default
5553 channel is named `channel0` and its attributes are set to reasonable
5554 values. Therefore, you only need to create a channel when you need
5555 non-default attributes.
5557 You specify each non-default channel attribute with a command-line
5558 option when you use the `enable-channel` command. The available
5559 command-line options are:
5561 [role="growable",cols="asciidoc,asciidoc"]
5562 .Command-line options for the `enable-channel` command.
5564 |Option |Description
5570 <<channel-overwrite-mode-vs-discard-mode,event loss mode>> instead of
5571 the default _discard_ mode.
5573 |`--buffers-pid` (user space tracing domain only)
5576 Use the per-process <<channel-buffering-schemes,buffering scheme>>
5577 instead of the default per-user buffering scheme.
5579 |+--subbuf-size=__SIZE__+
5582 Allocate sub-buffers of +__SIZE__+ bytes (power of two), for each CPU,
5583 either for each Unix user (default), or for each instrumented process.
5585 See <<channel-subbuf-size-vs-subbuf-count,Sub-buffer count and size>>.
5587 |+--num-subbuf=__COUNT__+
5590 Allocate +__COUNT__+ sub-buffers (power of two), for each CPU, either
5591 for each Unix user (default), or for each instrumented process.
5593 See <<channel-subbuf-size-vs-subbuf-count,Sub-buffer count and size>>.
5595 |+--tracefile-size=__SIZE__+
5598 Set the maximum size of each trace file that this channel writes within
5599 a stream to +__SIZE__+ bytes instead of no maximum.
5601 See <<tracefile-rotation,Trace file count and size>>.
5603 |+--tracefile-count=__COUNT__+
5606 Limit the number of trace files that this channel creates to
5607 +__COUNT__+ channels instead of no limit.
5609 See <<tracefile-rotation,Trace file count and size>>.
5611 |+--switch-timer=__PERIODUS__+
5614 Set the <<channel-switch-timer,switch timer period>>
5615 to +__PERIODUS__+{nbsp}µs.
5617 |+--read-timer=__PERIODUS__+
5620 Set the <<channel-read-timer,read timer period>>
5621 to +__PERIODUS__+{nbsp}µs.
5623 |+--output=__TYPE__+ (Linux kernel tracing domain only)
5626 Set the channel's output type to +__TYPE__+, either `mmap` or `splice`.
5631 for more details about those command-line options.
5633 You can only create a channel in the Linux kernel and user space
5634 <<domain,tracing domains>>: other tracing domains have their own
5635 channel created on the fly when
5636 <<enabling-disabling-events,creating event rules>>.
5640 Because of a current LTTng limitation, you must create all channels
5641 _before_ you <<basic-tracing-session-control,start tracing>> in a given
5642 tracing session, that is, before the first time you run `lttng start`.
5644 Since LTTng automatically creates a default channel when you use the
5645 `enable-event` command with a specific tracing domain, you cannot, for
5646 example, create a Linux kernel event rule, start tracing, and then
5647 create a user space event rule, because no user space channel exists yet
5648 and it's too late to create one.
5650 For this reason, make sure to configure your channels properly
5651 before starting the tracers for the first time!
5654 The following examples show how you can combine the previous
5655 command-line options to create simple to more complex channels.
5657 .Create a Linux kernel channel with default attributes.
5661 lttng enable-channel --kernel my-channel
5665 .Create a user space channel with 4 sub-buffers or 1{nbsp}MiB each, per CPU, per instrumented process.
5669 lttng enable-channel --userspace --num-subbuf=4 --subbuf-size=1M \
5670 --buffers-pid my-channel
5674 .Create a Linux kernel channel which rotates 8 trace files of 4{nbsp}MiB each for each stream
5678 lttng enable-channel --kernel --tracefile-count=8 \
5679 --tracefile-size=4194304 my-channel
5683 .Create a user space channel in overwrite (or _flight recorder_) mode.
5687 lttng enable-channel --userspace --overwrite my-channel
5691 You can <<enabling-disabling-events,create>> the same event rule in
5692 two different channels:
5696 lttng enable-event --userspace --channel=my-channel app:tp
5697 lttng enable-event --userspace --channel=other-channel app:tp
5700 If both channels are enabled, when a tracepoint named `app:tp` is
5701 reached, LTTng records two events, one for each channel.
5705 === Disable a channel
5707 To disable a specific channel that you <<enabling-disabling-channels,created>>
5708 previously, use the `disable-channel` command.
5710 .Disable a specific Linux kernel channel.
5714 lttng disable-channel --kernel my-channel
5718 The state of a channel precedes the individual states of event rules
5719 attached to it: event rules which belong to a disabled channel, even if
5720 they are enabled, are also considered disabled.
5724 === Add context fields to a channel
5726 Event record fields in trace files provide important information about
5727 events that occured previously, but sometimes some external context may
5728 help you solve a problem faster. Examples of context fields are:
5730 * The **process ID**, **thread ID**, **process name**, and
5731 **process priority** of the thread in which the event occurs.
5732 * The **hostname** of the system on which the event occurs.
5733 * The current values of many possible **performance counters** using
5735 ** CPU cycles, stalled cycles, idle cycles, and the other cycle types.
5737 ** Branch instructions, misses, and loads.
5740 To get the full list of available context fields, see
5741 `lttng add-context --help`. Some context fields are reserved for a
5742 specific <<domain,tracing domain>> (Linux kernel or user space).
5744 You add context fields to <<channel,channels>>. All the events
5745 that a channel with added context fields records contain those fields.
5747 To add context fields to one or all the channels of a given tracing
5748 session, use the `add-context` command.
5750 .Add context fields to all the channels of the current tracing session.
5752 The following command line adds the virtual process identifier and
5753 the per-thread CPU cycles count fields to all the user space channels
5754 of the current tracing session.
5758 lttng add-context --userspace --type=vpid --type=perf:thread:cpu-cycles
5762 .Add a context field to a specific channel.
5764 The following command line adds the thread identifier context field
5765 to the Linux kernel channel named `my-channel` in the current
5770 lttng add-context --kernel --channel=my-channel --type=tid
5774 NOTE: You cannot remove context fields from a channel once you add it.
5779 === Track process IDs
5781 It's often useful to allow only specific process IDs (PIDs) to emit
5782 events. For example, you may wish to record all the system calls made by
5783 a given process (Ă la http://linux.die.net/man/1/strace[strace]).
5785 The `track` and `untrack` commands serve this purpose. Both commands
5786 operate on a whitelist of process IDs. You _add_ entries to this
5787 whitelist with the `track` command and remove entries with the `untrack`
5788 command. Any process which has one of the PIDs in the whitelist is
5789 allowed to emit LTTng events which pass an enabled <<event,event rule>>.
5791 NOTE: The PID tracker tracks the _numeric process IDs_. Should a
5792 process with a given tracked ID exit and another process be given this
5793 ID, then the latter would also be allowed to emit events.
5795 .Track and untrack process IDs.
5797 For the sake of the following example, assume the target system has 16
5801 <<creating-destroying-tracing-sessions,create a tracing session>>,
5802 the whitelist contains all the possible PIDs:
5805 .All PIDs are tracked.
5806 image::track-all.png[]
5808 When the whitelist is full and you use the `track` command to specify
5809 some PIDs to track, LTTng first clears the whitelist, then it tracks
5810 the specific PIDs. After:
5814 lttng track --pid=3,4,7,10,13
5820 .PIDs 3, 4, 7, 10, and 13 are tracked.
5821 image::track-3-4-7-10-13.png[]
5823 You can add more PIDs to the whitelist afterwards:
5827 lttng track --pid=1,15,16
5833 .PIDs 1, 15, and 16 are added to the whitelist.
5834 image::track-1-3-4-7-10-13-15-16.png[]
5836 The `untrack` command removes entries from the PID tracker's whitelist.
5837 Given the previous example, the following command:
5841 lttng untrack --pid=3,7,10,13
5844 leads to this whitelist:
5847 .PIDs 3, 7, 10, and 13 are removed from the whitelist.
5848 image::track-1-4-15-16.png[]
5850 LTTng can track all possible PIDs again using the `--all` option:
5854 lttng track --pid --all
5857 The result is, again:
5860 .All PIDs are tracked.
5861 image::track-all.png[]
5864 .Track only specific PIDs
5866 A very typical use case with PID tracking is to start with an empty
5867 whitelist, then <<basic-tracing-session-control,start the tracers>>,
5868 and then add PIDs manually while tracers are active. You can accomplish
5869 this by using the `--all` option of the `untrack` command to clear the
5870 whitelist after you create a tracing session:
5874 lttng untrack --pid --all
5880 .No PIDs are tracked.
5881 image::untrack-all.png[]
5883 If you trace with this whitelist configuration, the tracer records no
5884 events for this <<domain,tracing domain>> because no processes are
5885 tracked. You can use the `track` command as usual to track specific
5890 lttng track --pid=6,11
5896 .PIDs 6 and 11 are tracked.
5897 image::track-6-11.png[]
5902 [[saving-loading-tracing-session]]
5903 === Save and load tracing session configurations
5905 Configuring a <<tracing-session,tracing session>> can be long. Some of
5906 the tasks involved are:
5908 * <<enabling-disabling-channels,Create channels>> with
5909 specific attributes.
5910 * <<adding-context,Add context fields>> to specific channels.
5911 * <<enabling-disabling-events,Create event rules>> with specific log
5912 level and filter conditions.
5914 If you use LTTng to solve real world problems, chances are you have to
5915 record events using the same tracing session setup over and over,
5916 modifying a few variables each time in your instrumented program
5917 or environment. To avoid constant tracing session reconfiguration,
5918 the cmd:lttng command-line tool can save and load tracing session
5919 configurations to/from XML files.
5921 To save a given tracing session configuration:
5923 * Use the `save` command:
5928 lttng save my-session
5932 Replace `my-session` with the name of the tracing session to save.
5934 LTTng saves tracing session configurations to
5935 dir:{$LTTNG_HOME/.lttng/sessions} by default. Note that the
5936 env:LTTNG_HOME environment variable defaults to `$HOME` if not set. Use
5937 the `--output-path` option to change this destination directory.
5939 LTTng saves all configuration parameters, for example:
5941 * The tracing session name.
5942 * The trace data output path.
5943 * The channels with their state and all their attributes.
5944 * The context fields you added to channels.
5945 * The event rules with their state, log level and filter conditions.
5947 To load a tracing session:
5949 * Use the `load` command:
5954 lttng load my-session
5958 Replace `my-session` with the name of the tracing session to load.
5960 When LTTng loads a configuration, it restores your saved tracing session
5961 as if you just configured it manually.
5963 See man:lttng(1) for the complete list of command-line options. You
5964 can also save and load all many sessions at a time, and decide in which
5965 directory to output the XML files.
5968 [[sending-trace-data-over-the-network]]
5969 === Send trace data over the network
5971 LTTng can send the recorded trace data to a remote system over the
5972 network instead of writing it to the local file system.
5974 To send the trace data over the network:
5976 . On the _remote_ system (which can also be the target system),
5977 start an LTTng <<lttng-relayd,relay daemon>>:
5986 . On the _target_ system, create a tracing session configured to
5987 send trace data over the network:
5992 lttng create my-session --set-url=net://remote-system
5996 Replace `remote-system` by the host name or IP address of the
5997 remote system. See `lttng create --help` for the exact URL format.
5999 . On the target system, use the cmd:lttng command-line tool as usual.
6000 When tracing is active, the target's consumer daemon sends sub-buffers
6001 to the relay daemon running on the remote system intead of flushing
6002 them to the local file system. The relay daemon writes the received
6003 packets to the local file system.
6005 The relay daemon writes trace files to
6006 +$LTTNG_HOME/lttng-traces/__hostname__/__session__+ by default, where
6007 +__hostname__+ is the host name of the target system and +__session__+
6008 is the tracing session name. Note that the env:LTTNG_HOME environment
6009 variable defaults to `$HOME` if not set. Use the `--output` option of
6010 cmd:lttng-relayd to write trace files to another base directory.
6015 === View events as LTTng emits them (noch:{LTTng} live)
6017 LTTng live is a network protocol implemented by the
6018 <<lttng-relayd,relay daemon>> to allow compatible trace viewers to
6019 display events as LTTng emits them on the target system while tracing
6022 The relay daemon creates a _tee_: it forwards the trace data to both
6023 the local file system and to connected live viewers:
6026 .The relay daemon creates a _tee_, forwarding the trace data to both trace files and a connected live viewer.
6031 . On the _target system_, create a <<tracing-session,tracing session>>
6037 lttng create --live my-session
6041 This spawns a local relay daemon.
6043 . Start the live viewer and configure it to connect to the relay
6044 daemon. For example, with http://diamon.org/babeltrace[Babeltrace]:
6049 babeltrace --input-format=lttng-live net://localhost/host/hostname/my-session
6056 * `hostname` with the host name of the target system.
6057 * `my-session` with the name of the tracing session to view.
6060 . Configure the tracing session as usual with the cmd:lttng
6061 command-line tool, and <<basic-tracing-session-control,start tracing>>.
6063 You can list the available live tracing sessions with Babeltrace:
6067 babeltrace --input-format=lttng-live net://localhost
6070 You can start the relay daemon on another system. In this case, you need
6071 to specify the relay daemon's URL when you create the tracing session
6072 with the `--set-url` option. You also need to replace `localhost`
6073 in the procedure above with the host name of the system on which the
6074 relay daemon is running.
6076 See man:lttng(1) and man:lttng-relayd(8) for the complete list of
6077 command-line options.
6081 [[taking-a-snapshot]]
6082 === Take a snapshot of the current sub-buffers of a tracing session
6084 The normal behavior of LTTng is to append full sub-buffers to growing
6085 trace data files. This is ideal to keep a full history of the events
6086 that occurred on the target system, but it can
6087 represent too much data in some situations. For example, you may wish
6088 to trace your application continuously until some critical situation
6089 happens, in which case you only need the latest few recorded
6090 events to perform the desired analysis, not multi-gigabyte trace files.
6092 With the `snapshot` command, you can take a snapshot of the current
6093 sub-buffers of a given <<tracing-session,tracing session>>. LTTng can
6094 write the snapshot to the local file system or send it over the network.
6098 . Create a tracing session in _snapshot mode_:
6103 lttng create --snapshot my-session
6107 The <<channel-overwrite-mode-vs-discard-mode,event loss mode>> of
6108 <<channel,channels>> created in this mode is automatically set to
6109 _overwrite_ (flight recorder mode).
6111 . Configure the tracing session as usual with the cmd:lttng
6112 command-line tool, and <<basic-tracing-session-control,start tracing>>.
6114 . **Optional**: When you need to take a snapshot, stop tracing.
6116 You can take a snapshot when the tracers are active, but if you stop
6117 them first, you are sure that the data in the sub-buffers does not
6118 change before you actually take the snapshot.
6125 lttng snapshot record --name=my-first-snapshot
6129 LTTng writes the current sub-buffers of all the current tracing
6130 session's channels to trace files on the local file system. Those trace
6131 files have `my-first-snapshot` in their name.
6133 There is no difference between the format of a normal trace file and the
6134 format of a snapshot: viewers of LTTng traces also support LTTng
6137 By default, LTTng writes snapshot files to the path shown by
6138 `lttng snapshot list-output`. You can change this path or decide to send
6139 snapshots over the network using either:
6141 . An output path or URL that you specify when you create the
6143 . An snapshot output path or URL that you add using
6144 `lttng snapshot add-output`
6145 . An output path or URL that you provide directly to the
6146 `lttng snapshot record` command.
6148 Method 3 overrides method 2, which overrides method 1. When you
6149 specify a URL, a relay daemon must listen on a remote system (see
6150 <<sending-trace-data-over-the-network,Send trace data over the network>>).
6155 === Use the machine interface
6157 With any command of the cmd:lttng command-line tool, you can use the
6158 `--mi=xml` argument (before the command name) to get an XML machine
6159 interface output, for example:
6163 lttng --mi=xml enable-event --kernel --syscall open
6166 A schema definition (XSD) is
6167 https://github.com/lttng/lttng-tools/blob/stable-{revision}/src/common/mi_lttng.xsd[available]
6168 to ease the integration with external tools as much as possible.
6172 [[persistent-memory-file-systems]]
6173 === Record trace data on persistent memory file systems
6175 https://en.wikipedia.org/wiki/Non-volatile_random-access_memory[Non-volatile random-access memory]
6176 (NVRAM) is random-access memory that retains its information when power
6177 is turned off (non-volatile). Systems with such memory can store data
6178 structures in RAM and retrieve them after a reboot, without flushing
6179 to typical _storage_.
6181 Linux supports NVRAM file systems thanks to either
6182 http://pramfs.sourceforge.net/[PRAMFS] or
6183 https://www.kernel.org/doc/Documentation/filesystems/dax.txt[DAX]{nbsp}+{nbsp}http://lkml.iu.edu/hypermail/linux/kernel/1504.1/03463.html[pmem]
6184 (requires Linux 4.1+).
6186 This section does not describe how to operate such file systems;
6187 we assume that you have a working persistent memory file system.
6189 When you create a <<tracing-session,tracing session>>, you can specify
6190 the path of the shared memory holding the sub-buffers. If you specify a
6191 location on an NVRAM file system, then you can retrieve the latest
6192 recorded trace data when the system reboots after a crash.
6194 To record trace data on a persistent memory file system and retrieve the
6195 trace data after a system crash:
6197 . Create a tracing session with a sub-buffer shared memory path located
6198 on an NVRAM file system:
6203 lttng create --shm-path=/path/to/shm
6207 . Configure the tracing session as usual with the cmd:lttng
6208 command-line tool, and <<basic-tracing-session-control,start tracing>>.
6210 . After a system crash, use the cmd:lttng-crash command-line tool to
6211 view the trace data recorded on the NVRAM file system:
6216 lttng-crash /path/to/shm
6220 The binary layout of the ring buffer files is not exactly the same as
6221 the trace files layout. This is why you need to use the cmd:lttng-crash
6222 utility instead of your preferred trace viewer directly.
6224 To convert the ring buffer files to LTTng trace files:
6226 * Use the `--extract` option of cmd:lttng-crash:
6231 lttng-crash --extract=/path/to/trace /path/to/shm
6235 See man:lttng-crash(1) for the complete list of command-line options.
6241 This section presents various references for LTTng packages such as
6242 links to online manpages, tables that the rest of the text needs,
6243 descriptions of library functions, and more.
6246 [[online-lttng-manpages]]
6247 === Online noch:{LTTng} manpages
6249 LTTng packages currently install the following link:/man[man pages],
6250 available online using the links below:
6254 ** man:lttng-crash(1)
6255 ** man:lttng-sessiond(8)
6256 ** man:lttng-relayd(8)
6258 ** man:lttng-gen-tp(1)
6260 ** man:lttng-ust-cyg-profile(3)
6261 ** man:lttng-ust-dl(3)
6265 === noch:{LTTng-UST}
6267 This section presents references of the LTTng-UST package.
6271 ==== noch:{LTTng-UST} library (+liblttng‑ust+)
6273 The LTTng-UST library, or `liblttng-ust`, is the main shared object
6274 against which user applications are linked to make LTTng user space
6277 The <<c-application,C application>> guide shows the complete
6278 process to instrument, build and run a C/$$C++$$ application using
6279 LTTng-UST, while this section contains a few important tables.
6282 [[liblttng-ust-tp-fields]]
6283 ===== Tracepoint fields macros (for `TP_FIELDS()`)
6285 The available macros to define tracepoint fields, which you must use
6286 within `TP_FIELDS()` in `TRACEPOINT_EVENT()`, are:
6288 [role="func-desc growable",cols="asciidoc,asciidoc"]
6289 .Available macros to define LTTng-UST tracepoint fields
6291 |Macro |Description and parameters
6294 +ctf_integer(__t__, __n__, __e__)+
6296 +ctf_integer_nowrite(__t__, __n__, __e__)+
6298 Standard integer, displayed in base 10.
6301 Integer C type (`int`, `long`, `size_t`, ...).
6307 Argument expression.
6309 |+ctf_integer_hex(__t__, __n__, __e__)+
6311 Standard integer, displayed in base 16.
6320 Argument expression.
6322 |+ctf_integer_network(__t__, __n__, __e__)+
6324 Integer in network byte order (big-endian), displayed in base 10.
6333 Argument expression.
6335 |+ctf_integer_network_hex(__t__, __n__, __e__)+
6337 Integer in network byte order, displayed in base 16.
6346 Argument expression.
6349 +ctf_float(__t__, __n__, __e__)+
6351 +ctf_float_nowrite(__t__, __n__, __e__)+
6353 Floating point number.
6356 Floating point number C type (`float` or `double`).
6362 Argument expression.
6365 +ctf_string(__n__, __e__)+
6367 +ctf_string_nowrite(__n__, __e__)+
6369 Null-terminated string; undefined behavior if +__e__+ is `NULL`.
6375 Argument expression.
6378 +ctf_array(__t__, __n__, __e__, __s__)+
6380 +ctf_array_nowrite(__t__, __n__, __e__, __s__)+
6382 Statically-sized array of integers
6385 Array element C type.
6391 Argument expression.
6397 +ctf_array_text(__t__, __n__, __e__, __s__)+
6399 +ctf_array_text_nowrite(__t__, __n__, __e__, __s__)+
6401 Statically-sized array, printed as text.
6403 The string does not need to be null-terminated.
6406 Array element C type (always `char`).
6412 Argument expression.
6418 +ctf_sequence(__t__, __n__, __e__, __T__, __E__)+
6420 +ctf_sequence_nowrite(__t__, __n__, __e__, __T__, __E__)+
6422 Dynamically-sized array of integers.
6424 The type of +__E__+ must be unsigned.
6427 Array element C type.
6433 Argument expression.
6436 Length expression C type.
6442 +ctf_sequence_text(__t__, __n__, __e__, __T__, __E__)+
6444 +ctf_sequence_text_nowrite(__t__, __n__, __e__, __T__, __E__)+
6446 Dynamically-sized array, displayed as text.
6448 The string does not need to be null-terminated.
6450 The type of +__E__+ must be unsigned.
6452 The behaviour is undefined if +__e__+ is `NULL`.
6455 Sequence element C type (always `char`).
6461 Argument expression.
6464 Length expression C type.
6470 The `_nowrite` versions omit themselves from the session trace, but are
6471 otherwise identical. This means the tracer does not write the `_nowrite`
6472 fields to the trace. Their primary purpose is to make some of the event
6473 context available to the <<enabling-disabling-events,event filters>>
6474 without having to commit the data to sub-buffers.
6477 [[liblttng-ust-tracepoint-loglevel]]
6478 ===== Tracepoint log levels (for `TRACEPOINT_LOGLEVEL()`)
6480 The following table shows the available log level values for the
6481 `TRACEPOINT_LOGLEVEL()` macro:
6487 Action must be taken immediately.
6490 Critical conditions.
6499 Normal, but significant, condition.
6502 Informational message.
6504 `TRACE_DEBUG_SYSTEM`::
6505 Debug information with system-level scope (set of programs).
6507 `TRACE_DEBUG_PROGRAM`::
6508 Debug information with program-level scope (set of processes).
6510 `TRACE_DEBUG_PROCESS`::
6511 Debug information with process-level scope (set of modules).
6513 `TRACE_DEBUG_MODULE`::
6514 Debug information with module (executable/library) scope (set of units).
6516 `TRACE_DEBUG_UNIT`::
6517 Debug information with compilation unit scope (set of functions).
6519 `TRACE_DEBUG_FUNCTION`::
6520 Debug information with function-level scope.
6522 `TRACE_DEBUG_LINE`::
6523 Debug information with line-level scope (TRACEPOINT_EVENT default).
6526 Debug-level message.
6528 Log levels `TRACE_EMERG` through `TRACE_INFO` and `TRACE_DEBUG` match
6529 http://man7.org/linux/man-pages/man3/syslog.3.html[syslog]
6530 level semantics. Log levels `TRACE_DEBUG_SYSTEM` through `TRACE_DEBUG`
6531 offer more fine-grained selection of debug information.
6534 [[lttng-modules-ref]]
6535 === noch:{LTTng-modules}
6537 This section presents references of the LTTng-modules package.
6541 [[lttng-modules-tp-fields]]
6542 ==== Tracepoint fields macros (for `TP_FIELDS()`)
6544 [[tp-fast-assign]][[tp-struct-entry]]The available macros to define
6545 tracepoint fields, which must be listed within `TP_FIELDS()` in
6546 `LTTNG_TRACEPOINT_EVENT()`, are:
6548 [role="func-desc growable",cols="asciidoc,asciidoc"]
6549 .Available macros to define LTTng-modules tracepoint fields
6551 |Macro |Description and parameters
6554 +ctf_integer(__t__, __n__, __e__)+
6556 +ctf_integer_nowrite(__t__, __n__, __e__)+
6558 +ctf_user_integer(__t__, __n__, __e__)+
6560 +ctf_user_integer_nowrite(__t__, __n__, __e__)+
6562 Standard integer, displayed in base 10.
6565 Integer C type (`int`, `long`, `size_t`, ...).
6571 Argument expression.
6574 +ctf_integer_hex(__t__, __n__, __e__)+
6576 +ctf_user_integer_hex(__t__, __n__, __e__)+
6578 Standard integer, displayed in base 16.
6587 Argument expression.
6589 |+ctf_integer_oct(__t__, __n__, __e__)+
6591 Standard integer, displayed in base 8.
6600 Argument expression.
6603 +ctf_integer_network(__t__, __n__, __e__)+
6605 +ctf_user_integer_network(__t__, __n__, __e__)+
6607 Integer in network byte order (big-endian), displayed in base 10.
6616 Argument expression.
6619 +ctf_integer_network_hex(__t__, __n__, __e__)+
6621 +ctf_user_integer_network_hex(__t__, __n__, __e__)+
6623 Integer in network byte order, displayed in base 16.
6632 Argument expression.
6635 +ctf_string(__n__, __e__)+
6637 +ctf_string_nowrite(__n__, __e__)+
6639 +ctf_user_string(__n__, __e__)+
6641 +ctf_user_string_nowrite(__n__, __e__)+
6643 Null-terminated string; undefined behavior if +__e__+ is `NULL`.
6649 Argument expression.
6652 +ctf_array(__t__, __n__, __e__, __s__)+
6654 +ctf_array_nowrite(__t__, __n__, __e__, __s__)+
6656 +ctf_user_array(__t__, __n__, __e__, __s__)+
6658 +ctf_user_array_nowrite(__t__, __n__, __e__, __s__)+
6660 Statically-sized array of integers
6663 Array element C type.
6669 Argument expression.
6675 +ctf_array_text(__t__, __n__, __e__, __s__)+
6677 +ctf_array_text_nowrite(__t__, __n__, __e__, __s__)+
6679 +ctf_user_array_text(__t__, __n__, __e__, __s__)+
6681 +ctf_user_array_text_nowrite(__t__, __n__, __e__, __s__)+
6683 Statically-sized array, printed as text.
6685 The string does not need to be null-terminated.
6688 Array element C type (always `char`).
6694 Argument expression.
6700 +ctf_sequence(__t__, __n__, __e__, __T__, __E__)+
6702 +ctf_sequence_nowrite(__t__, __n__, __e__, __T__, __E__)+
6704 +ctf_user_sequence(__t__, __n__, __e__, __T__, __E__)+
6706 +ctf_user_sequence_nowrite(__t__, __n__, __e__, __T__, __E__)+
6708 Dynamically-sized array of integers.
6710 The type of +__E__+ must be unsigned.
6713 Array element C type.
6719 Argument expression.
6722 Length expression C type.
6727 |+ctf_sequence_hex(__t__, __n__, __e__, __T__, __E__)+
6729 Dynamically-sized array of integers, displayed in base 16.
6731 The type of +__E__+ must be unsigned.
6734 Array element C type.
6740 Argument expression.
6743 Length expression C type.
6748 |+ctf_sequence_network(__t__, __n__, __e__, __T__, __E__)+
6750 Dynamically-sized array of integers in network byte order (big-endian),
6751 displayed in base 10.
6753 The type of +__E__+ must be unsigned.
6756 Array element C type.
6762 Argument expression.
6765 Length expression C type.
6771 +ctf_sequence_text(__t__, __n__, __e__, __T__, __E__)+
6773 +ctf_sequence_text_nowrite(__t__, __n__, __e__, __T__, __E__)+
6775 +ctf_user_sequence_text(__t__, __n__, __e__, __T__, __E__)+
6777 +ctf_user_sequence_text_nowrite(__t__, __n__, __e__, __T__, __E__)+
6779 Dynamically-sized array, displayed as text.
6781 The string does not need to be null-terminated.
6783 The type of +__E__+ must be unsigned.
6785 The behaviour is undefined if +__e__+ is `NULL`.
6788 Sequence element C type (always `char`).
6794 Argument expression.
6797 Length expression C type.
6803 Use the `_user` versions when the argument expression, `e`, is
6804 a user space address. In the cases of `ctf_user_integer*()` and
6805 `ctf_user_float*()`, `&e` must be a user space address, thus `e` must
6808 The `_nowrite` versions omit themselves from the session trace, but are
6809 otherwise identical. This means the `_nowrite` fields won't be written
6810 in the recorded trace. Their primary purpose is to make some
6811 of the event context available to the
6812 <<enabling-disabling-events,event filters>> without having to
6813 commit the data to sub-buffers.
6819 Terms related to LTTng and to tracing in general:
6822 The http://diamon.org/babeltrace[Babeltrace] project, which includes
6823 the cmd:babeltrace command, some libraries, and Python bindings.
6825 <<channel-buffering-schemes,buffering scheme>>::
6826 A layout of sub-buffers applied to a given channel.
6828 <<channel,channel>>::
6829 An entity which is responsible for a set of ring buffers.
6831 <<event,Event rules>> are always attached to a specific channel.
6834 A reference of time for a tracer.
6836 <<lttng-consumerd,consumer daemon>>::
6837 A process which is responsible for consuming the full sub-buffers
6838 and write them to a file system or send them over the network.
6840 <<channel-overwrite-mode-vs-discard-mode,discard mode>>:: The event loss
6841 mode in which the tracer _discards_ new event records when there's no
6842 sub-buffer space left to store them.
6845 The consequence of the execution of an instrumentation
6846 point, like a tracepoint that you manually place in some source code,
6847 or a Linux kernel KProbe.
6849 An event is said to _occur_ at a specific time. Different actions can
6850 be taken upon the occurance of an event, like record the event's payload
6853 <<channel-overwrite-mode-vs-discard-mode,event loss mode>>::
6854 The mechanism by which event records of a given channel are lost
6855 (not recorded) when there is no sub-buffer space left to store them.
6857 [[def-event-name]]event name::
6858 The name of an event, which is also the name of the event record.
6859 This is also called the _instrumentation point name_.
6862 A record, in a trace, of the payload of an event which occured.
6864 <<event,event rule>>::
6865 Set of conditions which must be satisfied for one or more occuring
6866 events to be recorded.
6868 `java.util.logging`::
6870 https://docs.oracle.com/javase/7/docs/api/java/util/logging/package-summary.html[core logging facilities].
6872 <<instrumenting,instrumentation>>::
6873 The use of LTTng probes to make a piece of software traceable.
6875 instrumentation point::
6876 A point in the execution path of a piece of software that, when
6877 reached by this execution, can emit an event.
6879 instrumentation point name::
6880 See _<<def-event-name,event name>>_.
6883 A http://logging.apache.org/log4j/1.2/[logging library] for Java
6884 developed by the Apache Software Foundation.
6887 Level of severity of a log statement or user space
6888 instrumentation point.
6891 The _Linux Trace Toolkit: next generation_ project.
6893 <<lttng-cli,cmd:lttng>>::
6894 A command-line tool provided by the LTTng-tools project which you
6895 can use to send and receive control messages to and from a
6899 The https://github.com/lttng/lttng-analyses[LTTng analyses] project,
6900 which is a set of analyzing programs that are used to obtain a
6901 higher level view of an LTTng trace.
6903 cmd:lttng-consumerd::
6904 The name of the consumer daemon program.
6907 A utility provided by the LTTng-tools project which can convert
6908 ring buffer files (usually
6909 <<persistent-memory-file-systems,saved on a persistent memory file system>>)
6912 LTTng Documentation::
6915 <<lttng-live,LTTng live>>::
6916 A communication protocol between the relay daemon and live viewers
6917 which makes it possible to see events "live", as they are received by
6920 <<lttng-modules,LTTng-modules>>::
6921 The https://github.com/lttng/lttng-modules[LTTng-modules] project,
6922 which contains the Linux kernel modules to make the Linux kernel
6923 instrumentation points available for LTTng tracing.
6926 The name of the relay daemon program.
6928 cmd:lttng-sessiond::
6929 The name of the session daemon program.
6932 The https://github.com/lttng/lttng-tools[LTTng-tools] project, which
6933 contains the various programs and libraries used to
6934 <<controlling-tracing,control tracing>>.
6936 <<lttng-ust,LTTng-UST>>::
6937 The https://github.com/lttng/lttng-ust[LTTng-UST] project, which
6938 contains libraries to instrument user applications.
6940 <<lttng-ust-agents,LTTng-UST Java agent>>::
6941 A Java package provided by the LTTng-UST project to allow the
6942 LTTng instrumentation of `java.util.logging` and Apache log4j 1.2
6945 <<lttng-ust-agents,LTTng-UST Python agent>>::
6946 A Python package provided by the LTTng-UST project to allow the
6947 LTTng instrumentation of Python logging statements.
6949 <<channel-overwrite-mode-vs-discard-mode,overwrite mode>>::
6950 The event loss mode in which new event records overwrite older
6951 event records when there's no sub-buffer space left to store them.
6953 <<channel-buffering-schemes,per-process buffering>>::
6954 A buffering scheme in which each instrumented process has its own
6955 sub-buffers for a given user space channel.
6957 <<channel-buffering-schemes,per-user buffering>>::
6958 A buffering scheme in which all the processes of a Unix user share the
6959 same sub-buffer for a given user space channel.
6961 <<lttng-relayd,relay daemon>>::
6962 A process which is responsible for receiving the trace data sent by
6963 a distant consumer daemon.
6966 A set of sub-buffers.
6968 <<lttng-sessiond,session daemon>>::
6969 A process which receives control commands from you and orchestrates
6970 the tracers and various LTTng daemons.
6972 <<taking-a-snapshot,snapshot>>::
6973 A copy of the current data of all the sub-buffers of a given tracing
6974 session, saved as trace files.
6977 One part of an LTTng ring buffer which contains event records.
6980 The time information attached to an event when it is emitted.
6983 A set of files which are the concatenations of one or more
6984 flushed sub-buffers.
6987 The action of recording the events emitted by an application
6988 or by a system, or to initiate such recording by controlling
6992 The http://tracecompass.org[Trace Compass] project and application.
6995 An instrumentation point using the tracepoint mechanism of the Linux
6996 kernel or of LTTng-UST.
6998 tracepoint definition::
6999 The definition of a single tracepoint.
7002 The name of a tracepoint.
7004 tracepoint provider::
7005 A set of functions providing tracepoints to an instrumented user
7008 Not to be confused with a _tracepoint provider package_: many tracepoint
7009 providers can exist within a tracepoint provider package.
7011 tracepoint provider package::
7012 One or more tracepoint providers compiled as an object file or as
7016 A software which records emitted events.
7018 <<domain,tracing domain>>::
7019 A namespace for event sources.
7022 The Unix group in which a Unix user can be to be allowed to trace the
7025 <<tracing-session,tracing session>>::
7026 A stateful dialogue between you and a <<lttng-sessiond,session
7030 An application running in user space, as opposed to a Linux kernel
7031 module, for example.