1 The LTTng Documentation
2 =======================
3 Philippe Proulx <pproulx@efficios.com>
4 v2.12, 28 November 2023
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 Skip this section if you’re familiar with software tracing and with the
28 * **<<installing-lttng,Installation>>** describes the steps to
29 install the LTTng packages on common Linux distributions and from
32 Skip this section if you already properly installed LTTng on your target
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 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 Skip this section if you don't have a programming background.
60 * **<<controlling-tracing,Tracing control>>** is divided into topics
61 which demonstrate how to use the vast array of features that
62 LTTng{nbsp}{revision} offers.
63 * **<<reference,Reference>>** contains reference tables.
64 * **<<glossary,Glossary>>** is a specialized dictionary of terms related
65 to LTTng or to the field of software tracing.
68 include::../common/convention.txt[]
71 include::../common/acknowledgements.txt[]
75 == What's new in LTTng{nbsp}{revision}?
77 LTTng{nbsp}{revision} bears the name _Ta Meilleure_, a Northeast IPA
78 beer brewed by https://lagabiere.com/[Lagabière]. Translating to ``Your
79 best one'', this beer gives out strong aromas of passion fruit, lemon,
80 and peaches. Tastewise, expect a lot of fruit, a creamy texture, and a
81 smooth lingering hop bitterness.
83 New features and changes in LTTng{nbsp}{revision}:
87 * Clear the contents of one or more <<tracing-session,tracing sessions>>
88 without having to destroy and reconfigure them
89 with the new man:lttng-clear(1) command.
91 This is especially useful to clear the tracing data of a tracing session
92 between attempts to reproduce a problem.
94 See <<clear,Clear a tracing session>>.
96 * Before LTTng{nbsp}{revision}, the man:lttng-track(1) and
97 man:lttng-untrack(1) commands used to add and remove process IDs
98 (PIDs) to a whitelist so that LTTng would only trace processes with
101 LTTng{nbsp}{revision} adds Unix user IDs (UIDs) and Unix group IDs
102 (GIDs) to the available <<pid-tracking,process attributes to track>>.
103 You can specify numeric user/group IDs and user/group names to track,
108 $ lttng track --userspace --vuid=http,999 --vgid=mysql,9
111 While you can also track UIDs and GIDs with the
112 opt:lttng-enable-event(1):--filter option of the `enable-event` command,
113 this dedicated process attribute tracking approach reduces tracing
114 overhead and prevents the creation of <<def-sub-buffer,sub-buffers>> for
115 the users and groups which LTTng doesn't track.
117 In the command manual pages, the term ``whitelist'' is renamed to
118 ``inclusion set'' to clarify the concept.
120 * The <<lttng-relayd,relay daemon>> can now maintain many files
121 virtually opened without using as many file descriptors (FD). It does
122 so by closing and reopening FDs as needed.
124 This feature is meant as a workaround for users who can't bump the
125 system limit because of permission restrictions.
127 The new opt:lttng-relayd(8):--fd-pool-size relay daemon option
128 sets the maximum number of simultaneously opened file descriptors
129 (using the soft `RLIMIT_NOFILE` resource limit of the process by
130 default; see man:getrlimit(2)).
132 * By default, the relay daemon writes its traces under a predefined
134 +$LTTNG_HOME/lttng-traces/__host__/__session__/__domain__+, with:
141 <<tracing-session,Tracing session>> name.
144 <<domain,Tracing domain>> name (`ust` or `kernel`).
147 Change this hierarchy to group traces by tracing session name rather
149 (+$LTTNG_HOME/lttng-traces/__session__/__host__/__domain__+) with the
150 new opt:lttng-relayd(8):--group-output-by-session option of the
153 This feature is especially useful if you're tracing two or more hosts,
154 having different hostnames, which share the same tracing session name as
155 part of their configuration. In this scenario, you can use a descriptive
156 tracing session name (for example, `connection-hang`) across a fleet of
157 machines streaming to a single relay daemon.
159 * The relay daemon has a new opt:lttng-relayd(8):--working-directory
160 option to override its working directory.
162 Linux kernel tracing::
164 * New instrumentation hooks to trace the entry and exit tracepoints of
165 the network reception code paths of the Linux kernel.
167 Use the resulting event records to identify the bounds of a network
168 reception and link the events that occur in the interim (for example,
169 wake-ups) to a specific network reception instance. You can also
170 analyze the latency of the network stack thanks to those event records.
172 * The `thread` field of the `irqaction` structure, which specifies the
173 process to wake up when a threaded interrupt request (IRQ) occurs, is
174 now part of the `lttng_statedump_interrupt` event record.
176 Use this information to discover which processes handle the various
177 IRQs. You can also associate the events occurring in the context of
178 those processes with their respective IRQ.
180 * New `lttng_statedump_cpu_topology` tracepoint to record the active
183 Use this information to discover which CPUs are SMT siblings or part of
184 the same socket. As of LTTng{nbsp}{revision}, only the x86 architecture
185 is supported since all architectures describe their topologies
188 The `architecture` field of the tracepoint is statically defined and
189 exists for all architecture implementations. Analysis tools can
190 therefore anticipate the layout of the event record.
192 Event record example:
196 lttng_statedump_cpu_topology:
202 model_name: Intel(R) Core(TM) i7-7600U CPU @ 2.80GHz
208 * New product UUID metadata environment field, `product_uuid`,
209 which LTTng copies from the
210 https://en.wikipedia.org/wiki/Desktop_Management_Interface[Desktop
211 Management Interface] (DMI).
213 Use this environment field to uniquely identify a machine (virtual or
214 physical) in order to correlate traces from multiple virtual machines.
220 What is LTTng? As its name suggests, the _Linux Trace Toolkit: next
221 generation_ is a modern toolkit for tracing Linux systems and
222 applications. So your first question might be:
229 As the history of software engineering progressed and led to what
230 we now take for granted--complex, numerous and
231 interdependent software applications running in parallel on
232 sophisticated operating systems like Linux--the authors of such
233 components, software developers, began feeling a natural
234 urge to have tools that would ensure the robustness and good performance
235 of their masterpieces.
237 One major achievement in this field is, inarguably, the
238 https://www.gnu.org/software/gdb/[GNU debugger (GDB)],
239 an essential tool for developers to find and fix bugs. But even the best
240 debugger won't help make your software run faster, and nowadays, faster
241 software means either more work done by the same hardware, or cheaper
242 hardware for the same work.
244 A _profiler_ is often the tool of choice to identify performance
245 bottlenecks. Profiling is suitable to identify _where_ performance is
246 lost in a given software. The profiler outputs a profile, a statistical
247 summary of observed events, which you may use to discover which
248 functions took the most time to execute. However, a profiler won't
249 report _why_ some identified functions are the bottleneck. Bottlenecks
250 might only occur when specific conditions are met, conditions that are
251 sometimes impossible to capture by a statistical profiler, or impossible
252 to reproduce with an application altered by the overhead of an
253 event-based profiler. For a thorough investigation of software
254 performance issues, a history of execution is essential, with the
255 recorded values of variables and context fields you choose, and
256 with as little influence as possible on the instrumented software. This
257 is where tracing comes in handy.
259 _Tracing_ is a technique used to understand what goes on in a running
260 software system. The software used for tracing is called a _tracer_,
261 which is conceptually similar to a tape recorder. When recording,
262 specific instrumentation points placed in the software source code
263 generate events that are saved on a giant tape: a _trace_ file. You
264 can trace user applications and the operating system at the same time,
265 opening the possibility of resolving a wide range of problems that would
266 otherwise be extremely challenging.
268 Tracing is often compared to _logging_. However, tracers and loggers are
269 two different tools, serving two different purposes. Tracers are
270 designed to record much lower-level events that occur much more
271 frequently than log messages, often in the range of thousands per
272 second, with very little execution overhead. Logging is more appropriate
273 for a very high-level analysis of less frequent events: user accesses,
274 exceptional conditions (errors and warnings, for example), database
275 transactions, instant messaging communications, and such. Simply put,
276 logging is one of the many use cases that can be satisfied with tracing.
278 The list of recorded events inside a trace file can be read manually
279 like a log file for the maximum level of detail, but it is generally
280 much more interesting to perform application-specific analyses to
281 produce reduced statistics and graphs that are useful to resolve a
282 given problem. Trace viewers and analyzers are specialized tools
285 In the end, this is what LTTng is: a powerful, open source set of
286 tools to trace the Linux kernel and user applications at the same time.
287 LTTng is composed of several components actively maintained and
288 developed by its link:/community/#where[community].
291 [[lttng-alternatives]]
292 === Alternatives to noch:{LTTng}
294 Excluding proprietary solutions, a few competing software tracers
297 https://github.com/dtrace4linux/linux[dtrace4linux]::
298 A port of Sun Microsystems' DTrace to Linux.
300 The cmd:dtrace tool interprets user scripts and is responsible for
301 loading code into the Linux kernel for further execution and collecting
304 https://en.wikipedia.org/wiki/Berkeley_Packet_Filter[eBPF]::
305 A subsystem in the Linux kernel in which a virtual machine can
306 execute programs passed from the user space to the kernel.
308 You can attach such programs to tracepoints and kprobes thanks to a
309 system call, and they can output data to the user space when executed
310 thanks to different mechanisms (pipe, VM register values, and eBPF maps,
313 https://www.kernel.org/doc/Documentation/trace/ftrace.txt[ftrace]::
314 The de facto function tracer of the Linux kernel.
316 Its user interface is a set of special files in sysfs.
318 https://perf.wiki.kernel.org/[perf]::
319 A performance analysis tool for Linux which supports hardware
320 performance counters, tracepoints, as well as other counters and
323 The controlling utility of perf is the cmd:perf command line/text UI
326 http://linux.die.net/man/1/strace[strace]::
327 A command-line utility which records system calls made by a
328 user process, as well as signal deliveries and changes of process
331 strace makes use of https://en.wikipedia.org/wiki/Ptrace[ptrace] to
332 fulfill its function.
334 http://www.sysdig.org/[sysdig]::
335 Like SystemTap, uses scripts to analyze Linux kernel events.
337 You write scripts, or _chisels_ in the jargon of sysdig, in Lua and
338 sysdig executes them while it traces the system or afterwards. The
339 interface of sysdig is the cmd:sysdig command-line tool as well as the
340 text UI-based cmd:csysdig tool.
342 https://sourceware.org/systemtap/[SystemTap]::
343 A Linux kernel and user space tracer which uses custom user scripts
344 to produce plain text traces.
346 SystemTap converts the scripts to the C language, and then compiles them
347 as Linux kernel modules which are loaded to produce trace data. The
348 primary user interface of SystemTap is the cmd:stap command-line tool.
350 The main distinctive features of LTTng is that it produces correlated
351 kernel and user space traces, as well as doing so with the lowest
352 overhead amongst other solutions. It produces trace files in the
353 http://diamon.org/ctf[CTF] format, a file format optimized
354 for the production and analyses of multi-gigabyte data.
356 LTTng is the result of more than 10{nbsp}years of active open source
357 development by a community of passionate developers. LTTng is currently
358 available on major desktop and server Linux distributions.
360 The main interface for tracing control is a single command-line tool
361 named cmd:lttng. The latter can create several tracing sessions, enable
362 and disable events on the fly, filter events efficiently with custom
363 user expressions, start and stop tracing, and much more. LTTng can
364 record the traces on the file system or send them over the network, and
365 keep them totally or partially. You can view the traces once tracing
366 becomes inactive or in real-time.
368 <<installing-lttng,Install LTTng now>> and
369 <<getting-started,start tracing>>!
375 **LTTng** is a set of software <<plumbing,components>> which interact to
376 <<instrumenting,instrument>> the Linux kernel and user applications, and
377 to <<controlling-tracing,control tracing>> (start and stop
378 tracing, enable and disable event rules, and the rest). Those
379 components are bundled into the following packages:
382 Libraries and command-line interface to control tracing.
385 Linux kernel modules to instrument and trace the kernel.
388 Libraries and Java/Python packages to instrument and trace user
391 Most distributions mark the LTTng-modules and LTTng-UST packages as
392 optional when installing LTTng-tools (which is always required). In the
393 following sections, we always provide the steps to install all three,
396 * You only need to install LTTng-modules if you intend to trace the
398 * You only need to install LTTng-UST if you intend to trace user
402 .Availability of LTTng{nbsp}{revision} for major Linux distributions as of 3{nbsp}November{nbsp}2023.
404 |Distribution |Available in releases
406 |https://www.ubuntu.com/[Ubuntu]
407 |Ubuntu{nbsp}16.04 _Xenial Xerus_, Ubuntu{nbsp}18.04 _Bionic Beaver_,
408 Ubuntu{nbsp}20.04 _Focal Fossa_, and Ubuntu{nbsp}22.04 _Jammy Jellyfish_:
409 <<ubuntu-ppa,use the LTTng Stable{nbsp}{revision} PPA>>.
411 |https://www.debian.org/[Debian]
412 |<<debian,Debian{nbsp}11 _bullseye_>>.
414 |https://www.redhat.com/[RHEL] and https://www.suse.com/[SLES]
415 |See http://packages.efficios.com/[EfficiOS Enterprise Packages].
417 |https://alpinelinux.org/[Alpine Linux]
418 |xref:alpine-linux[Alpine Linux{nbsp}3.12, Alpine Linux{nbsp}3.13,
419 Alpine Linux{nbsp}3.14, and Alpine Linux{nbsp}3.15].
421 |https://buildroot.org/[Buildroot]
422 |xref:buildroot[Buildroot{nbsp}2020.08, Buildroot{nbsp}2020.11,
423 Builroot{nbsp}2021.02, Buildroot{nbsp}2021.05, Buildroot{nbsp}2021.08,
424 and Builroot{nbsp}2021.11].
426 |https://www.openembedded.org/wiki/Main_Page[OpenEmbedded] and
427 https://www.yoctoproject.org/[Yocto]
428 |xref:oe-yocto[Yocto Project{nbsp}3.2 _Gatesgarth_ and Yocto Project{nbsp}3.3 _Hardknott_].
433 === Ubuntu: noch:{LTTng} Stable {revision} PPA
435 The https://launchpad.net/~lttng/+archive/ubuntu/stable-{revision}[LTTng
436 Stable{nbsp}{revision} PPA] offers the latest stable
437 LTTng{nbsp}{revision} packages for Ubuntu{nbsp}16.04 _Xenial Xerus_,
438 Ubuntu{nbsp}18.04 _Bionic Beaver_, Ubuntu{nbsp}20.04 _Focal Fossa_,
439 and Ubuntu{nbsp}22.04 _Jammy Jellyfish_.
441 To install LTTng{nbsp}{revision} from the LTTng Stable{nbsp}{revision}
444 . Add the LTTng Stable{nbsp}{revision} PPA repository and update the
450 # apt-add-repository ppa:lttng/stable-2.12
455 . Install the main LTTng{nbsp}{revision} packages:
460 # apt-get install lttng-tools
461 # apt-get install lttng-modules-dkms
462 # apt-get install liblttng-ust-dev
466 . **If you need to instrument and trace
467 <<java-application,Java applications>>**, install the LTTng-UST
473 # apt-get install liblttng-ust-agent-java
477 . **If you need to instrument and trace
478 <<python-application,Python{nbsp}3 applications>>**, install the
479 LTTng-UST Python agent:
484 # apt-get install python3-lttngust
492 To install LTTng{nbsp}{revision} on Debian{nbsp}11 _bullseye_:
494 . Install the main LTTng{nbsp}{revision} packages:
499 # apt-get install lttng-modules-dkms
500 # apt-get install liblttng-ust-dev
501 # apt-get install lttng-tools
505 . **If you need to instrument and trace <<java-application,Java
506 applications>>**, install the LTTng-UST Java agent:
511 # apt-get install liblttng-ust-agent-java
515 . **If you need to instrument and trace <<python-application,Python
516 applications>>**, install the LTTng-UST Python agent:
521 # apt-get install python3-lttngust
528 To install LTTng-tools{nbsp}{revision} and LTTng-UST{nbsp}{revision} on
529 Alpine Linux{nbsp}3.12, Alpine Linux{nbsp}3.13, Alpine Linux{nbsp}3.14,
530 or Alpine Linux{nbsp}3.15:
532 . Add the LTTng packages:
537 # apk add lttng-tools
538 # apk add lttng-ust-dev
542 . Download, build, and install the latest LTTng-modules{nbsp}{revision}:
548 wget http://lttng.org/files/lttng-modules/lttng-modules-latest-2.12.tar.bz2 &&
549 tar -xf lttng-modules-latest-2.12.tar.bz2 &&
550 cd lttng-modules-2.12.* &&
552 sudo make modules_install &&
561 To install LTTng{nbsp}{revision} on Buildroot{nbsp}2020.08, Buildroot{nbsp}2020.11,
562 Buildroot{nbsp}2021.02, Buildroot{nbsp}2021.05, Buildroot{nbsp}2021.08, or
563 Buildroot{nbsp}2021.11:
565 . Launch the Buildroot configuration tool:
574 . In **Kernel**, check **Linux kernel**.
575 . In **Toolchain**, check **Enable WCHAR support**.
576 . In **Target packages**{nbsp}→ **Debugging, profiling and benchmark**,
577 check **lttng-modules** and **lttng-tools**.
578 . In **Target packages**{nbsp}→ **Libraries**{nbsp}→
579 **Other**, check **lttng-libust**.
583 === OpenEmbedded and Yocto
585 LTTng{nbsp}{revision} recipes are available in the
586 https://layers.openembedded.org/layerindex/branch/master/layer/openembedded-core/[`openembedded-core`]
587 layer for Yocto Project{nbsp}3.2 _Gatesgarth_ and Yocto Project{nbsp}3.3 _Hardknott_
588 under the following names:
594 With BitBake, the simplest way to include LTTng recipes in your target
595 image is to add them to `IMAGE_INSTALL_append` in path:{conf/local.conf}:
598 IMAGE_INSTALL_append = " lttng-tools lttng-modules lttng-ust"
603 . Select a machine and an image recipe.
604 . Click **Edit image recipe**.
605 . Under the **All recipes** tab, search for **lttng**.
606 . Check the desired LTTng recipes.
609 [[building-from-source]]
610 === Build from source
612 To build and install LTTng{nbsp}{revision} from source:
614 . Using the package manager of your distribution, or from source,
615 install the following dependencies of LTTng-tools and LTTng-UST:
618 * https://sourceforge.net/projects/libuuid/[libuuid]
619 * http://directory.fsf.org/wiki/Popt[popt]
620 * http://liburcu.org/[Userspace RCU]
621 * http://www.xmlsoft.org/[libxml2]
622 * **Optional**: https://github.com/numactl/numactl[numactl]
625 . Download, build, and install the latest LTTng-modules{nbsp}{revision}:
631 wget http://lttng.org/files/lttng-modules/lttng-modules-latest-2.12.tar.bz2 &&
632 tar -xf lttng-modules-latest-2.12.tar.bz2 &&
633 cd lttng-modules-2.12.* &&
635 sudo make modules_install &&
640 . Download, build, and install the latest LTTng-UST{nbsp}{revision}:
646 wget http://lttng.org/files/lttng-ust/lttng-ust-latest-2.12.tar.bz2 &&
647 tar -xf lttng-ust-latest-2.12.tar.bz2 &&
648 cd lttng-ust-2.12.* &&
656 Add `--disable-numa` to `./configure` if you don't have
657 https://github.com/numactl/numactl[numactl].
661 .Java and Python application tracing
663 If you need to instrument and trace <<java-application,Java
664 applications>>, pass the `--enable-java-agent-jul`,
665 `--enable-java-agent-log4j`, or `--enable-java-agent-all` options to the
666 `configure` script, depending on which Java logging framework you use.
668 If you need to instrument and trace <<python-application,Python
669 applications>>, pass the `--enable-python-agent` option to the
670 `configure` script. You can set the `PYTHON` environment variable to the
671 path to the Python interpreter for which to install the LTTng-UST Python
679 By default, LTTng-UST libraries are installed to
680 dir:{/usr/local/lib}, which is the de facto directory in which to
681 keep self-compiled and third-party libraries.
683 When <<building-tracepoint-providers-and-user-application,linking an
684 instrumented user application with `liblttng-ust`>>:
686 * Append `/usr/local/lib` to the env:LD_LIBRARY_PATH environment
688 * Pass the `-L/usr/local/lib` and `-Wl,-rpath,/usr/local/lib` options to
689 man:gcc(1), man:g++(1), or man:clang(1).
693 . Download, build, and install the latest LTTng-tools{nbsp}{revision}:
699 wget http://lttng.org/files/lttng-tools/lttng-tools-latest-2.12.tar.bz2 &&
700 tar -xf lttng-tools-latest-2.12.tar.bz2 &&
701 cd lttng-tools-2.12.* &&
709 TIP: The https://github.com/eepp/vlttng[vlttng tool] can do all the
710 previous steps automatically for a given version of LTTng and confine
711 the installed files in a specific directory. This can be useful to test
712 LTTng without installing it on your system.
716 === Linux kernel module signature
718 Linux kernel modules require trusted signatures in order to be loaded
719 when any of the following is true:
721 * The system boots with
722 https://uefi.org/specs/UEFI/2.10/32_Secure_Boot_and_Driver_Signing.html#secure-boot-and-driver-signing[Secure Boot]
725 * The Linux kernel which boots is configured with
726 `CONFIG_MODULE_SIG_FORCE`.
728 * The Linux kernel boots with a command line containing
729 `module.sig_enforce=1`.
731 .`root` user running <<lttng-sessiond,`lttng-sessiond`>> which fails to load a required <<lttng-modules,kernel module>> due to the signature enforcement policies.
736 Warning: No tracing group detected
737 modprobe: ERROR: could not insert 'lttng_ring_buffer_client_discard': Key was rejected by service
738 Error: Unable to load required module lttng-ring-buffer-client-discard
739 Warning: No kernel tracer available
743 There are several methods to enroll trusted keys for signing modules
744 that are built from source. The precise details vary from one Linux
745 version to another, and distributions may have their own mechanisms. For
746 example, https://github.com/dell/dkms[DKMS] may autogenerate a key and
747 sign modules, but the key isn't automatically enrolled.
750 https://www.kernel.org/doc/html/latest/admin-guide/module-signing.html[Kernel
751 module signing facility] and the documentation of your distribution
752 to learn more about signing Linux kernel modules.
757 This is a short guide to get started quickly with LTTng kernel and user
760 Before you follow this guide, make sure to <<installing-lttng,install>>
763 This tutorial walks you through the steps to:
765 . <<tracing-the-linux-kernel,Trace the Linux kernel>>.
766 . <<tracing-your-own-user-application,Trace a user application>> written
768 . <<viewing-and-analyzing-your-traces,View and analyze the
772 [[tracing-the-linux-kernel]]
773 === Trace the Linux kernel
775 The following command lines start with the `#` prompt because you need
776 root privileges to trace the Linux kernel. You can also trace the kernel
777 as a regular user if your Unix user is a member of the
778 <<tracing-group,tracing group>>.
780 . Create a <<tracing-session,tracing session>> which writes its traces
781 to dir:{/tmp/my-kernel-trace}:
786 # lttng create my-kernel-session --output=/tmp/my-kernel-trace
790 . List the available kernel tracepoints and system calls:
795 # lttng list --kernel
796 # lttng list --kernel --syscall
800 . Create <<event,event rules>> which match the desired instrumentation
801 point names, for example the `sched_switch` and `sched_process_fork`
802 tracepoints, and the man:open(2) and man:close(2) system calls:
807 # lttng enable-event --kernel sched_switch,sched_process_fork
808 # lttng enable-event --kernel --syscall open,close
812 Create an event rule which matches _all_ the Linux kernel
813 tracepoints with the opt:lttng-enable-event(1):--all option
814 (this will generate a lot of data when tracing):
819 # lttng enable-event --kernel --all
823 . <<basic-tracing-session-control,Start tracing>>:
832 . Do some operation on your system for a few seconds. For example,
833 load a website, or list the files of a directory.
834 . <<creating-destroying-tracing-sessions,Destroy>> the current
844 The man:lttng-destroy(1) command doesn't destroy the trace data; it
845 only destroys the state of the tracing session.
847 The man:lttng-destroy(1) command also runs the man:lttng-stop(1) command
848 implicitly (see <<basic-tracing-session-control,Start and stop a tracing
849 session>>). You need to stop tracing to make LTTng flush the remaining
850 trace data and make the trace readable.
852 . For the sake of this example, make the recorded trace accessible to
858 # chown -R $(whoami) /tmp/my-kernel-trace
862 See <<viewing-and-analyzing-your-traces,View and analyze the
863 recorded events>> to view the recorded events.
866 [[tracing-your-own-user-application]]
867 === Trace a user application
869 This section steps you through a simple example to trace a
870 _Hello world_ program written in C.
872 To create the traceable user application:
874 . Create the tracepoint provider header file, which defines the
875 tracepoints and the events they can generate:
881 #undef TRACEPOINT_PROVIDER
882 #define TRACEPOINT_PROVIDER hello_world
884 #undef TRACEPOINT_INCLUDE
885 #define TRACEPOINT_INCLUDE "./hello-tp.h"
887 #if !defined(_HELLO_TP_H) || defined(TRACEPOINT_HEADER_MULTI_READ)
890 #include <lttng/tracepoint.h>
900 ctf_string(my_string_field, my_string_arg)
901 ctf_integer(int, my_integer_field, my_integer_arg)
905 #endif /* _HELLO_TP_H */
907 #include <lttng/tracepoint-event.h>
911 . Create the tracepoint provider package source file:
917 #define TRACEPOINT_CREATE_PROBES
918 #define TRACEPOINT_DEFINE
920 #include "hello-tp.h"
924 . Build the tracepoint provider package:
929 $ gcc -c -I. hello-tp.c
933 . Create the _Hello World_ application source file:
940 #include "hello-tp.h"
942 int main(int argc, char *argv[])
946 puts("Hello, World!\nPress Enter to continue...");
949 * The following getchar() call is only placed here for the purpose
950 * of this demonstration, to pause the application in order for
951 * you to have time to list its tracepoints. It's not needed
957 * A tracepoint() call.
959 * Arguments, as defined in hello-tp.h:
961 * 1. Tracepoint provider name (required)
962 * 2. Tracepoint name (required)
963 * 3. my_integer_arg (first user-defined argument)
964 * 4. my_string_arg (second user-defined argument)
966 * Notice the tracepoint provider and tracepoint names are
967 * NOT strings: they are in fact parts of variables that the
968 * macros in hello-tp.h create.
970 tracepoint(hello_world, my_first_tracepoint, 23, "hi there!");
972 for (x = 0; x < argc; ++x) {
973 tracepoint(hello_world, my_first_tracepoint, x, argv[x]);
976 puts("Quitting now!");
977 tracepoint(hello_world, my_first_tracepoint, x * x, "x^2");
984 . Build the application:
993 . Link the application with the tracepoint provider package,
994 `liblttng-ust`, and `libdl`:
999 $ gcc -o hello hello.o hello-tp.o -llttng-ust -ldl
1003 Here's the whole build process:
1006 .Build steps of the user space tracing tutorial.
1007 image::ust-flow.png[]
1009 To trace the user application:
1011 . Run the application with a few arguments:
1016 $ ./hello world and beyond
1025 Press Enter to continue...
1029 . Start an LTTng <<lttng-sessiond,session daemon>>:
1034 $ lttng-sessiond --daemonize
1038 Note that a session daemon might already be running, for example as a
1039 service that the service manager of the distribution started.
1041 . List the available user space tracepoints:
1046 $ lttng list --userspace
1050 You see the `hello_world:my_first_tracepoint` tracepoint listed
1051 under the `./hello` process.
1053 . Create a <<tracing-session,tracing session>>:
1058 $ lttng create my-user-space-session
1062 . Create an <<event,event rule>> which matches the
1063 `hello_world:my_first_tracepoint` event name:
1068 $ lttng enable-event --userspace hello_world:my_first_tracepoint
1072 . <<basic-tracing-session-control,Start tracing>>:
1081 . Go back to the running `hello` application and press Enter. The
1082 program executes all `tracepoint()` instrumentation points and exits.
1083 . <<creating-destroying-tracing-sessions,Destroy>> the current
1093 The man:lttng-destroy(1) command doesn't destroy the trace data; it
1094 only destroys the state of the tracing session.
1096 The man:lttng-destroy(1) command also runs the man:lttng-stop(1) command
1097 implicitly (see <<basic-tracing-session-control,Start and stop a tracing
1098 session>>). You need to stop tracing to make LTTng flush the remaining
1099 trace data and make the trace readable.
1101 By default, LTTng saves the traces in
1102 +$LTTNG_HOME/lttng-traces/__name__-__date__-__time__+,
1103 where +__name__+ is the tracing session name. The
1104 env:LTTNG_HOME environment variable defaults to `$HOME` if not set.
1106 See <<viewing-and-analyzing-your-traces,View and analyze the
1107 recorded events>> to view the recorded events.
1110 [[viewing-and-analyzing-your-traces]]
1111 === View and analyze the recorded events
1113 Once you have completed the <<tracing-the-linux-kernel,Trace the Linux
1114 kernel>> and <<tracing-your-own-user-application,Trace a user
1115 application>> tutorials, you can inspect the recorded events.
1117 There are many tools you can use to read LTTng traces:
1119 https://babeltrace.org/[Babeltrace{nbsp}2]::
1120 A rich, flexible trace manipulation toolkit which includes
1121 a versatile command-line interface
1122 (https://babeltrace.org/docs/v2.0/man1/babeltrace2.1/[cmd:babeltrace2]),
1123 a https://babeltrace.org/docs/v2.0/libbabeltrace2/[C library],
1124 and https://babeltrace.org/docs/v2.0/python/bt2/[Python{nbsp}3 bindings]
1125 so that you can easily process or convert an LTTng trace with
1128 The Babeltrace{nbsp}2 project ships with a
1129 https://babeltrace.org/docs/v2.0/man7/babeltrace2-plugin-ctf.7/[plugin]
1130 which supports the format of the traces which LTTng produces,
1131 https://diamon.org/ctf/[CTF].
1133 http://tracecompass.org/[Trace Compass]::
1134 A graphical user interface for viewing and analyzing any type of
1135 logs or traces, including those of LTTng.
1137 https://github.com/lttng/lttng-analyses[**LTTng analyses**]::
1138 An experimental project which includes many high-level analyses of
1139 LTTng kernel traces, like scheduling statistics, interrupt
1140 frequency distribution, top CPU usage, and more.
1142 NOTE: This section assumes that LTTng saved the traces it recorded
1143 during the previous tutorials to their default location, in the
1144 dir:{$LTTNG_HOME/lttng-traces} directory. The env:LTTNG_HOME
1145 environment variable defaults to `$HOME` if not set.
1148 [[viewing-and-analyzing-your-traces-bt]]
1149 ==== Use the cmd:babeltrace2 command-line tool
1151 The simplest way to list all the recorded events of an LTTng trace is to
1153 https://babeltrace.org/docs/v2.0/man1/babeltrace2.1/[cmd:babeltrace2]
1158 $ babeltrace2 ~/lttng-traces/my-user-space-session*
1161 cmd:babeltrace2 finds all traces recursively within the given path and
1162 prints all their events, sorting them chronologically.
1164 Pipe the output of cmd:babeltrace2 into a tool like man:grep(1) for
1169 $ babeltrace2 /tmp/my-kernel-trace | grep _switch
1172 Pipe the output of cmd:babeltrace2 into a tool like man:wc(1) to count
1173 the recorded events:
1177 $ babeltrace2 /tmp/my-kernel-trace | grep _open | wc --lines
1181 [[viewing-and-analyzing-your-traces-bt-python]]
1182 ==== Use the Babeltrace{nbsp}2 Python bindings
1184 The <<viewing-and-analyzing-your-traces-bt,text output of
1185 cmd:babeltrace2>> is useful to isolate events by simple matching using
1186 man:grep(1) and similar utilities. However, more elaborate filters, such
1187 as keeping only event records with a field value falling within a
1188 specific range, are not trivial to write using a shell. Moreover,
1189 reductions and even the most basic computations involving multiple event
1190 records are virtually impossible to implement.
1192 Fortunately, Babeltrace{nbsp}2 ships with
1193 https://babeltrace.org/docs/v2.0/python/bt2/[Python{nbsp}3 bindings]
1194 which make it easy to read the event records of an LTTng trace
1195 sequentially and compute the desired information.
1197 The following script accepts an LTTng Linux kernel trace path as its
1198 first argument and prints the short names of the top five running
1199 processes on CPU{nbsp}0 during the whole trace:
1209 # Get the trace path from the first command-line argument.
1210 it = bt2.TraceCollectionMessageIterator(sys.argv[1])
1212 # This counter dictionary will hold execution times:
1214 # Task command name -> Total execution time (ns)
1215 exec_times = collections.Counter()
1217 # This holds the last `sched_switch` timestamp.
1221 # We only care about event messages.
1222 if type(msg) is not bt2._EventMessageConst:
1225 # Event of the event message.
1228 # Keep only `sched_switch` events.
1229 if event.cls.name != 'sched_switch':
1232 # Keep only events which occurred on CPU 0.
1233 if event.packet.context_field['cpu_id'] != 0:
1236 # Event timestamp (ns).
1237 cur_ts = msg.default_clock_snapshot.ns_from_origin
1243 # (Short) name of the previous task command.
1244 prev_comm = str(event.payload_field['prev_comm'])
1246 # Initialize an entry in our dictionary if not yet done.
1247 if prev_comm not in exec_times:
1248 exec_times[prev_comm] = 0
1250 # Compute previous command execution time.
1251 diff = cur_ts - last_ts
1253 # Update execution time of this command.
1254 exec_times[prev_comm] += diff
1256 # Update last timestamp.
1260 for name, ns in exec_times.most_common(5):
1261 print('{:20}{} s'.format(name, ns / 1e9))
1264 if __name__ == '__main__':
1272 $ python3 top5proc.py /tmp/my-kernel-trace/kernel
1278 swapper/0 48.607245889 s
1279 chromium 7.192738188 s
1280 pavucontrol 0.709894415 s
1281 Compositor 0.660867933 s
1282 Xorg.bin 0.616753786 s
1285 Note that `swapper/0` is the ``idle'' process of CPU{nbsp}0 on Linux;
1286 since we weren't using the CPU that much when tracing, its first
1287 position in the list makes sense.
1291 == [[understanding-lttng]]Core concepts
1293 From a user's perspective, the LTTng system is built on a few concepts,
1294 or objects, on which the <<lttng-cli,cmd:lttng command-line tool>>
1295 operates by sending commands to the <<lttng-sessiond,session daemon>>.
1296 Understanding how those objects relate to eachother is key in mastering
1299 The core concepts are:
1301 * <<tracing-session,Tracing session>>
1302 * <<domain,Tracing domain>>
1303 * <<channel,Channel and ring buffer>>
1304 * <<"event","Instrumentation point, event rule, event, and event record">>
1310 A _tracing session_ is a stateful dialogue between you and
1311 a <<lttng-sessiond,session daemon>>. You can
1312 <<creating-destroying-tracing-sessions,create a new tracing
1313 session>> with the `lttng create` command.
1315 Most of what you do when you control LTTng tracers happens within a
1316 tracing session. In particular, a tracing session:
1319 * Has its own set of trace files.
1320 * Has its own state of activity (started or stopped).
1321 * Has its own <<tracing-session-mode,mode>> (local, network streaming,
1323 * Has its own <<channel,channels>> to which are associated their own
1324 <<event,event rules>>.
1325 * Has its own <<pid-tracking,process attribute tracking>> inclusion
1329 .A _tracing session_ contains <<channel,channels>> that are members of <<domain,tracing domains>> and contain <<event,event rules>>.
1330 image::concepts.png[]
1332 Those attributes and objects are completely isolated between different
1335 A tracing session is analogous to a cash machine session:
1336 the operations you do on the banking system through the cash machine do
1337 not alter the data of other users of the same system. In the case of
1338 the cash machine, a session lasts as long as your bank card is inside.
1339 In the case of LTTng, a tracing session lasts from the `lttng create`
1340 command to the `lttng destroy` command.
1343 .Each Unix user has its own set of tracing sessions.
1344 image::many-sessions.png[]
1347 [[tracing-session-mode]]
1348 ==== Tracing session mode
1350 LTTng can send the generated trace data to different locations. The
1351 _tracing session mode_ dictates where to send it. The following modes
1352 are available in LTTng{nbsp}{revision}:
1354 [[local-mode]]Local mode::
1355 LTTng writes the traces to the file system of the machine it traces
1358 [[net-streaming-mode]]Network streaming mode::
1359 LTTng sends the traces over the network to a
1360 <<lttng-relayd,relay daemon>> running on a remote system.
1363 LTTng doesn't write the traces by default.
1365 Instead, you can request LTTng to <<taking-a-snapshot,take a snapshot>>,
1366 that is, a copy of the current sub-buffers of the tracing session, and
1367 to write it to the file system of the target or to send it over the
1368 network to a <<lttng-relayd,relay daemon>> running on a remote system.
1370 [[live-mode]]Live mode::
1371 This mode is similar to the network streaming mode, but a live
1372 trace viewer can connect to the distant relay daemon to
1373 <<lttng-live,view event records as LTTng generates them>>.
1379 A _tracing domain_ is a namespace for event sources. A tracing domain
1380 has its own properties and features.
1382 There are currently five available tracing domains:
1386 * `java.util.logging` (JUL)
1390 You must specify a tracing domain when using some commands to avoid
1391 ambiguity. For example, since all the domains support named tracepoints
1392 as event sources (instrumentation points that you manually insert in the
1393 source code), you need to specify a tracing domain when
1394 <<enabling-disabling-events,creating an event rule>> because all the
1395 tracing domains could have tracepoints with the same names.
1397 You can create <<channel,channels>> in the Linux kernel and user space
1398 tracing domains. The other tracing domains have a single default
1403 === Channel and ring buffer
1405 A _channel_ is an object which is responsible for a set of ring buffers.
1406 Each ring buffer is divided into multiple sub-buffers. When an LTTng
1407 tracer emits an event, it can record it to one or more
1408 sub-buffers. The attributes of a channel determine what to do when
1409 there's no space left for a new event record because all sub-buffers
1410 are full, where to send a full sub-buffer, and other behaviours.
1412 A channel is always associated to a <<domain,tracing domain>>. The
1413 `java.util.logging` (JUL), log4j, and Python tracing domains each have
1414 a default channel which you can't configure.
1416 A channel also owns <<event,event rules>>. When an LTTng tracer emits
1417 an event, it records it to the sub-buffers of all
1418 the enabled channels with a satisfied event rule, as long as those
1419 channels are part of active <<tracing-session,tracing sessions>>.
1422 [[channel-buffering-schemes]]
1423 ==== Per-user vs. per-process buffering schemes
1425 A channel has at least one ring buffer _per CPU_. LTTng always
1426 records an event to the ring buffer associated to the CPU on which it
1429 Two _buffering schemes_ are available when you
1430 <<enabling-disabling-channels,create a channel>> in the
1431 user space <<domain,tracing domain>>:
1433 Per-user buffering::
1434 Allocate one set of ring buffers--one per CPU--shared by all the
1435 instrumented processes of each Unix user.
1439 .Per-user buffering scheme.
1440 image::per-user-buffering.png[]
1443 Per-process buffering::
1444 Allocate one set of ring buffers--one per CPU--for each
1445 instrumented process.
1449 .Per-process buffering scheme.
1450 image::per-process-buffering.png[]
1453 The per-process buffering scheme tends to consume more memory than the
1454 per-user option because systems generally have more instrumented
1455 processes than Unix users running instrumented processes. However, the
1456 per-process buffering scheme ensures that one process having a high
1457 event throughput won't fill all the shared sub-buffers of the same
1460 The Linux kernel tracing domain has only one available buffering scheme
1461 which is to allocate a single set of ring buffers for the whole system.
1462 This scheme is similar to the per-user option, but with a single, global
1463 user ``running'' the kernel.
1466 [[channel-overwrite-mode-vs-discard-mode]]
1467 ==== Overwrite vs. discard event record loss modes
1469 When an event occurs, LTTng records it to a specific sub-buffer (yellow
1470 arc in the following animations) of the ring buffer of a specific
1471 channel. When there's no space left in a sub-buffer, the tracer marks it
1472 as consumable (red) and another, empty sub-buffer starts receiving the
1473 following event records. A <<lttng-consumerd,consumer daemon>>
1474 eventually consumes the marked sub-buffer (returns to white).
1477 [role="docsvg-channel-subbuf-anim"]
1482 In an ideal world, sub-buffers are consumed faster than they are filled,
1483 as it is the case in the previous animation. In the real world,
1484 however, all sub-buffers can be full at some point, leaving no space to
1485 record the following events.
1487 By default, LTTng-modules and LTTng-UST are _non-blocking_ tracers: when
1488 no empty sub-buffer is available, it is acceptable to lose event records
1489 when the alternative would be to cause substantial delays in the
1490 execution of the instrumented application. LTTng privileges performance
1491 over integrity; it aims at perturbing the target system as little as
1492 possible in order to make tracing of subtle race conditions and rare
1493 interrupt cascades possible.
1495 Since LTTng{nbsp}2.10, the LTTng user space tracer, LTTng-UST, supports
1496 a _blocking mode_. See the <<blocking-timeout-example,blocking timeout
1497 example>> to learn how to use the blocking mode.
1499 When it comes to losing event records because no empty sub-buffer is
1500 available, or because the <<opt-blocking-timeout,blocking timeout>> is
1501 reached, the _event record loss mode_ of the channel determines what to
1502 do. The available event record loss modes are:
1505 Drop the newest event records until the tracer releases a sub-buffer.
1507 This is the only available mode when you specify a
1508 <<opt-blocking-timeout,blocking timeout>>.
1511 Clear the sub-buffer containing the oldest event records and start
1512 writing the newest event records there.
1514 This mode is sometimes called _flight recorder mode_ because it's
1516 https://en.wikipedia.org/wiki/Flight_recorder[flight recorder]:
1517 always keep a fixed amount of the latest data.
1519 Which mechanism you should choose depends on your context: prioritize
1520 the newest or the oldest event records in the ring buffer?
1522 Beware that, in overwrite mode, the tracer abandons a _whole sub-buffer_
1523 as soon as a there's no space left for a new event record, whereas in
1524 discard mode, the tracer only discards the event record that doesn't
1527 In discard mode, LTTng increments a count of lost event records when an
1528 event record is lost and saves this count to the trace. Since
1529 LTTng{nbsp}2.8, in overwrite mode, LTTng writes to a given sub-buffer
1530 its sequence number within its data stream. With a <<local-mode,local>>,
1531 <<net-streaming-mode,network streaming>>, or <<live-mode,live>>
1532 <<tracing-session,tracing session>>, a trace reader can use such
1533 sequence numbers to report lost packets. In overwrite mode, LTTng
1534 doesn't write to the trace the exact number of lost event records in
1535 those lost sub-buffers.
1537 Trace analyses can use saved discarded event record and sub-buffer
1538 (packet) counts of the trace to decide whether or not to perform the
1539 analyses even if trace data is known to be missing.
1541 There are a few ways to decrease your probability of losing event
1543 <<channel-subbuf-size-vs-subbuf-count,Sub-buffer count and size>> shows
1544 how to fine-tune the sub-buffer count and size of a channel to virtually
1545 stop losing event records, though at the cost of greater memory usage.
1548 [[channel-subbuf-size-vs-subbuf-count]]
1549 ==== Sub-buffer count and size
1551 When you <<enabling-disabling-channels,create a channel>>, you can
1552 set its number of sub-buffers and their size.
1554 Note that there is noticeable CPU overhead introduced when
1555 switching sub-buffers (marking a full one as consumable and switching
1556 to an empty one for the following events to be recorded). Knowing this,
1557 the following list presents a few practical situations along with how
1558 to configure the sub-buffer count and size for them:
1560 * **High event throughput**: In general, prefer bigger sub-buffers to
1561 lower the risk of losing event records.
1563 Having bigger sub-buffers also ensures a lower
1564 <<channel-switch-timer,sub-buffer switching frequency>>.
1566 The number of sub-buffers is only meaningful if you create the channel
1567 in overwrite mode: in this case, if a sub-buffer overwrite happens, the
1568 other sub-buffers are left unaltered.
1570 * **Low event throughput**: In general, prefer smaller sub-buffers
1571 since the risk of losing event records is low.
1573 Because events occur less frequently, the sub-buffer switching frequency
1574 should remain low and thus the overhead of the tracer shouldn't be a
1577 * **Low memory system**: If your target system has a low memory
1578 limit, prefer fewer first, then smaller sub-buffers.
1580 Even if the system is limited in memory, you want to keep the
1581 sub-buffers as big as possible to avoid a high sub-buffer switching
1584 Note that LTTng uses http://diamon.org/ctf/[CTF] as its trace format,
1585 which means event data is very compact. For example, the average
1586 LTTng kernel event record weights about 32{nbsp}bytes. Thus, a
1587 sub-buffer size of 1{nbsp}MiB is considered big.
1589 The previous situations highlight the major trade-off between a few big
1590 sub-buffers and more, smaller sub-buffers: sub-buffer switching
1591 frequency vs. how much data is lost in overwrite mode. Assuming a
1592 constant event throughput and using the overwrite mode, the two
1593 following configurations have the same ring buffer total size:
1596 [role="docsvg-channel-subbuf-size-vs-count-anim"]
1601 * **Two sub-buffers of 4{nbsp}MiB each**: Expect a very low sub-buffer
1602 switching frequency, but if a sub-buffer overwrite happens, half of
1603 the event records so far (4{nbsp}MiB) are definitely lost.
1604 * **Eight sub-buffers of 1{nbsp}MiB each**: Expect four times the
1605 overhead of the tracer as the previous configuration, but if a
1606 sub-buffer overwrite happens, only the eighth of event records so far
1607 are definitely lost.
1609 In discard mode, the sub-buffers count parameter is pointless: use two
1610 sub-buffers and set their size according to the requirements of your
1614 [[channel-switch-timer]]
1615 ==== Switch timer period
1617 The _switch timer period_ is an important configurable attribute of
1618 a channel to ensure periodic sub-buffer flushing.
1620 When the _switch timer_ expires, a sub-buffer switch happens. Set
1621 the switch timer period attribute when you
1622 <<enabling-disabling-channels,create a channel>> to ensure that LTTng
1623 consumes and commits trace data to trace files or to a distant relay
1624 daemon periodically in case of a low event throughput.
1627 [role="docsvg-channel-switch-timer"]
1632 This attribute is also convenient when you use big sub-buffers to cope
1633 with a sporadic high event throughput, even if the throughput is
1637 [[channel-read-timer]]
1638 ==== Read timer period
1640 By default, the LTTng tracers use a notification mechanism to signal a
1641 full sub-buffer so that a consumer daemon can consume it. When such
1642 notifications must be avoided, for example in real-time applications,
1643 use the _read timer_ of the channel instead. When the read timer fires,
1644 the <<lttng-consumerd,consumer daemon>> checks for full, consumable
1648 [[tracefile-rotation]]
1649 ==== Trace file count and size
1651 By default, trace files can grow as large as needed. Set the maximum
1652 size of each trace file that a channel writes when you
1653 <<enabling-disabling-channels,create a channel>>. When the size of a
1654 trace file reaches the fixed maximum size of the channel, LTTng creates
1655 another file to contain the next event records. LTTng appends a file
1656 count to each trace file name in this case.
1658 If you set the trace file size attribute when you create a channel, the
1659 maximum number of trace files that LTTng creates is _unlimited_ by
1660 default. To limit them, set a maximum number of trace files. When the
1661 number of trace files reaches the fixed maximum count of the channel,
1662 the oldest trace file is overwritten. This mechanism is called _trace
1667 Even if you don't limit the trace file count, you can't assume that
1668 LTTng doesn't manage any trace file.
1670 In other words, there is no safe way to know if LTTng still holds a
1671 given trace file open with the trace file rotation feature.
1673 The only way to obtain an unmanaged, self-contained LTTng trace before
1674 you <<creating-destroying-tracing-sessions,destroy>> the tracing session
1675 is with the <<session-rotation,tracing session rotation>> feature
1676 (available since LTTng{nbsp}2.11).
1681 === Instrumentation point, event rule, event, and event record
1683 An _event rule_ is a set of conditions which must be **all** satisfied
1684 for LTTng to record an occuring event.
1686 You set the conditions when you <<enabling-disabling-events,create
1689 You always attach an event rule to a <<channel,channel>> when you create
1692 When an event passes the conditions of an event rule, LTTng records it
1693 in one of the sub-buffers of the attached channel.
1695 The available conditions, as of LTTng{nbsp}{revision}, are:
1697 * The event rule _is enabled_.
1698 * The type of the instrumentation point _is{nbsp}T_.
1699 * The name of the instrumentation point (sometimes called _event name_)
1700 _matches{nbsp}N_, but _isn't{nbsp}E_.
1701 * The log level of the instrumentation point _is as severe as{nbsp}L_, or
1702 _is exactly{nbsp}L_.
1703 * The fields of the payload of the event _satisfy_ a filter
1704 expression{nbsp}__F__.
1706 As you can see, all the conditions but the dynamic filter are related to
1707 the status of the event rule or to the instrumentation point, not to the
1708 occurring events. This is why, without a filter, checking if an event
1709 passes an event rule isn't a dynamic task: when you create or modify an
1710 event rule, all the tracers of its tracing domain enable or disable the
1711 instrumentation points themselves once. This is possible because the
1712 attributes of an instrumentation point (type, name, and log level) are
1713 defined statically. In other words, without a dynamic filter, the tracer
1714 _doesn't evaluate_ the arguments of an instrumentation point unless it
1715 matches an enabled event rule.
1717 Note that, for LTTng to record an event, the <<channel,channel>> to
1718 which a matching event rule is attached must also be enabled, and the
1719 <<tracing-session,tracing session>> owning this channel must be active
1723 .Logical path from an instrumentation point to an event record.
1724 image::event-rule.png[]
1726 .Event, event record, or event rule?
1728 With so many similar terms, it's easy to get confused.
1730 An **event** is the consequence of the execution of an _instrumentation
1731 point_, like a tracepoint that you manually place in some source code,
1732 or a Linux kernel kprobe. An event is said to _occur_ at a specific
1733 time. Different actions can be taken upon the occurrence of an event,
1734 like record the payload of the event to a buffer.
1736 An **event record** is the representation of an event in a sub-buffer. A
1737 tracer is responsible for capturing the payload of an event, current
1738 context variables, the ID of the event, and its timestamp. LTTng
1739 can append this sub-buffer to a trace file.
1741 An **event rule** is a set of conditions which must _all_ be satisfied
1742 for LTTng to record an occuring event. Events still occur without
1743 satisfying event rules, but LTTng doesn't record them.
1748 == Components of noch:{LTTng}
1750 The second _T_ in _LTTng_ stands for _toolkit_: it would be wrong
1751 to call LTTng a simple _tool_ since it is composed of multiple
1752 interacting components. This section describes those components,
1753 explains their respective roles, and shows how they connect together to
1754 form the LTTng ecosystem.
1756 The following diagram shows how the most important components of LTTng
1757 interact with user applications, the Linux kernel, and you:
1760 .Control and trace data paths between LTTng components.
1761 image::plumbing.png[]
1763 The LTTng project incorporates:
1765 * **LTTng-tools**: Libraries and command-line interface to
1766 control tracing sessions.
1767 ** <<lttng-sessiond,Session daemon>> (man:lttng-sessiond(8)).
1768 ** <<lttng-consumerd,Consumer daemon>> (cmd:lttng-consumerd).
1769 ** <<lttng-relayd,Relay daemon>> (man:lttng-relayd(8)).
1770 ** <<liblttng-ctl-lttng,Tracing control library>> (`liblttng-ctl`).
1771 ** <<lttng-cli,Tracing control command-line tool>> (man:lttng(1)).
1772 * **LTTng-UST**: Libraries and Java/Python packages to trace user
1774 ** <<lttng-ust,User space tracing library>> (`liblttng-ust`) and its
1775 headers to instrument and trace any native user application.
1776 ** <<prebuilt-ust-helpers,Preloadable user space tracing helpers>>:
1777 *** `liblttng-ust-libc-wrapper`
1778 *** `liblttng-ust-pthread-wrapper`
1779 *** `liblttng-ust-cyg-profile`
1780 *** `liblttng-ust-cyg-profile-fast`
1781 *** `liblttng-ust-dl`
1782 ** User space tracepoint provider source files generator command-line
1783 tool (man:lttng-gen-tp(1)).
1784 ** <<lttng-ust-agents,LTTng-UST Java agent>> to instrument and trace
1785 Java applications using `java.util.logging` or
1786 Apache log4j{nbsp}1.2 logging.
1787 ** <<lttng-ust-agents,LTTng-UST Python agent>> to instrument
1788 Python applications using the standard `logging` package.
1789 * **LTTng-modules**: <<lttng-modules,Linux kernel modules>> to trace
1791 ** LTTng kernel tracer module.
1792 ** Tracing ring buffer kernel modules.
1793 ** Probe kernel modules.
1794 ** LTTng logger kernel module.
1798 === Tracing control command-line interface
1801 .The tracing control command-line interface.
1802 image::plumbing-lttng-cli.png[]
1804 The _man:lttng(1) command-line tool_ is the standard user interface to
1805 control LTTng <<tracing-session,tracing sessions>>. The cmd:lttng tool
1806 is part of LTTng-tools.
1808 The cmd:lttng tool is linked with
1809 <<liblttng-ctl-lttng,`liblttng-ctl`>> to communicate with
1810 one or more <<lttng-sessiond,session daemons>> behind the scenes.
1812 The cmd:lttng tool has a Git-like interface:
1816 $ lttng <GENERAL OPTIONS> <COMMAND> <COMMAND OPTIONS>
1819 The <<controlling-tracing,Tracing control>> section explores the
1820 available features of LTTng using the cmd:lttng tool.
1823 [[liblttng-ctl-lttng]]
1824 === Tracing control library
1827 .The tracing control library.
1828 image::plumbing-liblttng-ctl.png[]
1830 The _LTTng control library_, `liblttng-ctl`, is used to communicate
1831 with a <<lttng-sessiond,session daemon>> using a C API that hides the
1832 underlying details of the protocol. `liblttng-ctl` is part of LTTng-tools.
1834 The <<lttng-cli,cmd:lttng command-line tool>>
1835 is linked with `liblttng-ctl`.
1837 Use `liblttng-ctl` in C or $$C++$$ source code by including its
1842 #include <lttng/lttng.h>
1845 Some objects are referenced by name (C string), such as tracing
1846 sessions, but most of them require to create a handle first using
1847 `lttng_create_handle()`.
1849 As of LTTng{nbsp}{revision}, the best available developer documentation for
1850 `liblttng-ctl` is its installed header files. Every function and structure is
1851 thoroughly documented.
1855 === User space tracing library
1858 .The user space tracing library.
1859 image::plumbing-liblttng-ust.png[]
1861 The _user space tracing library_, `liblttng-ust` (see man:lttng-ust(3)),
1862 is the LTTng user space tracer. It receives commands from a
1863 <<lttng-sessiond,session daemon>>, for example to
1864 enable and disable specific instrumentation points, and writes event
1865 records to ring buffers shared with a
1866 <<lttng-consumerd,consumer daemon>>.
1867 `liblttng-ust` is part of LTTng-UST.
1869 Public C header files are installed beside `liblttng-ust` to
1870 instrument any <<c-application,C or $$C++$$ application>>.
1872 <<lttng-ust-agents,LTTng-UST agents>>, which are regular Java and Python
1873 packages, use their own library providing tracepoints which is
1874 linked with `liblttng-ust`.
1876 An application or library doesn't have to initialize `liblttng-ust`
1877 manually: its constructor does the necessary tasks to properly register
1878 to a session daemon. The initialization phase also enables the
1879 instrumentation points matching the <<event,event rules>> that you
1883 [[lttng-ust-agents]]
1884 === User space tracing agents
1887 .The user space tracing agents.
1888 image::plumbing-lttng-ust-agents.png[]
1890 The _LTTng-UST Java and Python agents_ are regular Java and Python
1891 packages which add LTTng tracing capabilities to the
1892 native logging frameworks. The LTTng-UST agents are part of LTTng-UST.
1894 In the case of Java, the
1895 https://docs.oracle.com/javase/7/docs/api/java/util/logging/package-summary.html[`java.util.logging`
1896 core logging facilities] and
1897 https://logging.apache.org/log4j/1.2/[Apache log4j{nbsp}1.2] are supported.
1898 Note that Apache Log4{nbsp}2 isn't supported.
1900 In the case of Python, the standard
1901 https://docs.python.org/3/library/logging.html[`logging`] package
1902 is supported. Both Python{nbsp}2 and Python{nbsp}3 modules can import the
1903 LTTng-UST Python agent package.
1905 The applications using the LTTng-UST agents are in the
1906 `java.util.logging` (JUL),
1907 log4j, and Python <<domain,tracing domains>>.
1909 Both agents use the same mechanism to trace the log statements. When an
1910 agent initializes, it creates a log handler that attaches to the root
1911 logger. The agent also registers to a <<lttng-sessiond,session daemon>>.
1912 When the application executes a log statement, the root logger passes it
1913 to the log handler of the agent. The log handler of the agent calls a
1914 native function in a tracepoint provider package shared library linked
1915 with <<lttng-ust,`liblttng-ust`>>, passing the formatted log message and
1916 other fields, like its logger name and its log level. This native
1917 function contains a user space instrumentation point, hence tracing the
1920 The log level condition of an
1921 <<event,event rule>> is considered when tracing
1922 a Java or a Python application, and it's compatible with the standard
1923 JUL, log4j, and Python log levels.
1927 === LTTng kernel modules
1930 .The LTTng kernel modules.
1931 image::plumbing-lttng-modules.png[]
1933 The _LTTng kernel modules_ are a set of Linux kernel modules
1934 which implement the kernel tracer of the LTTng project. The LTTng
1935 kernel modules are part of LTTng-modules.
1937 The LTTng kernel modules include:
1939 * A set of _probe_ modules.
1941 Each module attaches to a specific subsystem
1942 of the Linux kernel using its tracepoint instrument points. There are
1943 also modules to attach to the entry and return points of the Linux
1944 system call functions.
1946 * _Ring buffer_ modules.
1948 A ring buffer implementation is provided as kernel modules. The LTTng
1949 kernel tracer writes to the ring buffer; a
1950 <<lttng-consumerd,consumer daemon>> reads from the ring buffer.
1952 * The _LTTng kernel tracer_ module.
1953 * The _LTTng logger_ module.
1955 The LTTng logger module implements the special path:{/proc/lttng-logger}
1956 (and path:{/dev/lttng-logger} since LTTng{nbsp}2.11) files so that any
1957 executable can generate LTTng events by opening and writing to those
1960 See <<proc-lttng-logger-abi,LTTng logger>>.
1962 Generally, you don't have to load the LTTng kernel modules manually
1963 (using man:modprobe(8), for example): a root <<lttng-sessiond,session
1964 daemon>> loads the necessary modules when starting. If you have extra
1965 probe modules, you can specify to load them to the session daemon on the
1966 command line. See also
1967 <<linux-kernel-sig,Linux kernel module signature>>.
1969 The LTTng kernel modules are installed in
1970 +/usr/lib/modules/__release__/extra+ by default, where +__release__+ is
1971 the kernel release (see `uname --kernel-release`).
1978 .The session daemon.
1979 image::plumbing-sessiond.png[]
1981 The _session daemon_, man:lttng-sessiond(8), is a daemon responsible for
1982 managing tracing sessions and for controlling the various components of
1983 LTTng. The session daemon is part of LTTng-tools.
1985 The session daemon sends control requests to and receives control
1988 * The <<lttng-ust,user space tracing library>>.
1990 Any instance of the user space tracing library first registers to
1991 a session daemon. Then, the session daemon can send requests to
1992 this instance, such as:
1995 ** Get the list of tracepoints.
1996 ** Share an <<event,event rule>> so that the user space tracing library
1997 can enable or disable tracepoints. Amongst the possible conditions
1998 of an event rule is a filter expression which `liblttng-ust` evalutes
1999 when an event occurs.
2000 ** Share <<channel,channel>> attributes and ring buffer locations.
2003 The session daemon and the user space tracing library use a Unix
2004 domain socket for their communication.
2006 * The <<lttng-ust-agents,user space tracing agents>>.
2008 Any instance of a user space tracing agent first registers to
2009 a session daemon. Then, the session daemon can send requests to
2010 this instance, such as:
2013 ** Get the list of loggers.
2014 ** Enable or disable a specific logger.
2017 The session daemon and the user space tracing agent use a TCP connection
2018 for their communication.
2020 * The <<lttng-modules,LTTng kernel tracer>>.
2021 * The <<lttng-consumerd,consumer daemon>>.
2023 The session daemon sends requests to the consumer daemon to instruct
2024 it where to send the trace data streams, amongst other information.
2026 * The <<lttng-relayd,relay daemon>>.
2028 The session daemon receives commands from the
2029 <<liblttng-ctl-lttng,tracing control library>>.
2031 The root session daemon loads the appropriate
2032 <<lttng-modules,LTTng kernel modules>> on startup. It also spawns
2033 a <<lttng-consumerd,consumer daemon>> as soon as you create
2034 an <<event,event rule>>.
2036 The session daemon doesn't send and receive trace data: this is the
2037 role of the <<lttng-consumerd,consumer daemon>> and
2038 <<lttng-relayd,relay daemon>>. It does, however, generate the
2039 http://diamon.org/ctf/[CTF] metadata stream.
2041 Each Unix user can have its own session daemon instance. The
2042 tracing sessions which different session daemons manage are completely
2045 The root user's session daemon is the only one which is
2046 allowed to control the LTTng kernel tracer, and its spawned consumer
2047 daemon is the only one which is allowed to consume trace data from the
2048 LTTng kernel tracer. Note, however, that any Unix user which is a member
2049 of the <<tracing-group,tracing group>> is allowed
2050 to create <<channel,channels>> in the
2051 Linux kernel <<domain,tracing domain>>, and thus to trace the Linux
2054 The <<lttng-cli,cmd:lttng command-line tool>> automatically starts a
2055 session daemon when using its `create` command if none is currently
2056 running. You can also start the session daemon manually.
2063 .The consumer daemon.
2064 image::plumbing-consumerd.png[]
2066 The _consumer daemon_, cmd:lttng-consumerd, is a daemon which shares
2067 ring buffers with user applications or with the LTTng kernel modules to
2068 collect trace data and send it to some location (on disk or to a
2069 <<lttng-relayd,relay daemon>> over the network). The consumer daemon
2070 is part of LTTng-tools.
2072 You don't start a consumer daemon manually: a consumer daemon is always
2073 spawned by a <<lttng-sessiond,session daemon>> as soon as you create an
2074 <<event,event rule>>, that is, before you start tracing. When you kill
2075 its owner session daemon, the consumer daemon also exits because it is
2076 the child process of the session daemon. Command-line options of
2077 man:lttng-sessiond(8) target the consumer daemon process.
2079 There are up to two running consumer daemons per Unix user, whereas only
2080 one session daemon can run per user. This is because each process can be
2081 either 32-bit or 64-bit: if the target system runs a mixture of 32-bit
2082 and 64-bit processes, it is more efficient to have separate
2083 corresponding 32-bit and 64-bit consumer daemons. The root user is an
2084 exception: it can have up to _three_ running consumer daemons: 32-bit
2085 and 64-bit instances for its user applications, and one more
2086 reserved for collecting kernel trace data.
2094 image::plumbing-relayd.png[]
2096 The _relay daemon_, man:lttng-relayd(8), is a daemon acting as a bridge
2097 between remote session and consumer daemons, local trace files, and a
2098 remote live trace viewer. The relay daemon is part of LTTng-tools.
2100 The main purpose of the relay daemon is to implement a receiver of
2101 <<sending-trace-data-over-the-network,trace data over the network>>.
2102 This is useful when the target system doesn't have much file system
2103 space to record trace files locally.
2105 The relay daemon is also a server to which a
2106 <<lttng-live,live trace viewer>> can
2107 connect. The live trace viewer sends requests to the relay daemon to
2108 receive trace data as the target system emits events. The
2109 communication protocol is named _LTTng live_; it is used over TCP
2112 Note that you can start the relay daemon on the target system directly.
2113 This is the setup of choice when the use case is to view events as
2114 the target system emits them without the need of a remote system.
2118 == [[using-lttng]]Instrumentation
2120 There are many examples of tracing and monitoring in our everyday life:
2122 * You have access to real-time and historical weather reports and
2123 forecasts thanks to weather stations installed around the country.
2124 * You know your heart is safe thanks to an electrocardiogram.
2125 * You make sure not to drive your car too fast and to have enough fuel
2126 to reach your destination thanks to gauges visible on your dashboard.
2128 All the previous examples have something in common: they rely on
2129 **instruments**. Without the electrodes attached to the surface of your
2130 body skin, cardiac monitoring is futile.
2132 LTTng, as a tracer, is no different from those real life examples. If
2133 you're about to trace a software system or, in other words, record its
2134 history of execution, you better have **instrumentation points** in the
2135 subject you're tracing, that is, the actual software.
2137 Various ways were developed to instrument a piece of software for LTTng
2138 tracing. The most straightforward one is to manually place
2139 instrumentation points, called _tracepoints_, in the source code of the
2140 software. It is also possible to add instrumentation points dynamically
2141 in the Linux kernel <<domain,tracing domain>>.
2143 If you're only interested in tracing the Linux kernel, your
2144 instrumentation needs are probably already covered by the built-in
2145 <<lttng-modules,Linux kernel tracepoints>> of LTTng. You may also wish
2146 to trace a user application which is already instrumented for LTTng
2147 tracing. In such cases, skip this whole section and read the topics of
2148 the <<controlling-tracing,Tracing control>> section.
2150 Many methods are available to instrument a piece of software for LTTng
2153 * <<c-application,User space instrumentation for C and $$C++$$
2155 * <<prebuilt-ust-helpers,Prebuilt user space tracing helpers>>.
2156 * <<java-application,User space Java agent>>.
2157 * <<python-application,User space Python agent>>.
2158 * <<proc-lttng-logger-abi,LTTng logger>>.
2159 * <<instrumenting-linux-kernel,LTTng kernel tracepoints>>.
2163 === [[cxx-application]]User space instrumentation for C and $$C++$$ applications
2165 The procedure to instrument a C or $$C++$$ user application with
2166 the <<lttng-ust,LTTng user space tracing library>>, `liblttng-ust`, is:
2168 . <<tracepoint-provider,Create the source files of a tracepoint provider
2170 . <<probing-the-application-source-code,Add tracepoints to
2171 the source code of the application>>.
2172 . <<building-tracepoint-providers-and-user-application,Build and link
2173 a tracepoint provider package and the user application>>.
2175 If you need quick, man:printf(3)-like instrumentation, skip
2176 those steps and use <<tracef,`tracef()`>> or <<tracelog,`tracelog()`>>
2179 IMPORTANT: You need to <<installing-lttng,install>> LTTng-UST to
2180 instrument a user application with `liblttng-ust`.
2183 [[tracepoint-provider]]
2184 ==== Create the source files of a tracepoint provider package
2186 A _tracepoint provider_ is a set of compiled functions which provide
2187 **tracepoints** to an application, the type of instrumentation point
2188 supported by LTTng-UST. Those functions can emit events with
2189 user-defined fields and serialize those events as event records to one
2190 or more LTTng-UST <<channel,channel>> sub-buffers. The `tracepoint()`
2191 macro, which you <<probing-the-application-source-code,insert in the
2192 source code of a user application>>, calls those functions.
2194 A _tracepoint provider package_ is an object file (`.o`) or a shared
2195 library (`.so`) which contains one or more tracepoint providers.
2196 Its source files are:
2198 * One or more <<tpp-header,tracepoint provider header>> (`.h`).
2199 * A <<tpp-source,tracepoint provider package source>> (`.c`).
2201 A tracepoint provider package is dynamically linked with `liblttng-ust`,
2202 the LTTng user space tracer, at run time.
2205 .User application linked with `liblttng-ust` and containing a tracepoint provider.
2206 image::ust-app.png[]
2208 NOTE: If you need quick, man:printf(3)-like instrumentation,
2209 skip creating and using a tracepoint provider and use
2210 <<tracef,`tracef()`>> or <<tracelog,`tracelog()`>> instead.
2214 ===== Create a tracepoint provider header file template
2216 A _tracepoint provider header file_ contains the tracepoint
2217 definitions of a tracepoint provider.
2219 To create a tracepoint provider header file:
2221 . Start from this template:
2225 .Tracepoint provider header file template (`.h` file extension).
2227 #undef TRACEPOINT_PROVIDER
2228 #define TRACEPOINT_PROVIDER provider_name
2230 #undef TRACEPOINT_INCLUDE
2231 #define TRACEPOINT_INCLUDE "./tp.h"
2233 #if !defined(_TP_H) || defined(TRACEPOINT_HEADER_MULTI_READ)
2236 #include <lttng/tracepoint.h>
2239 * Use TRACEPOINT_EVENT(), TRACEPOINT_EVENT_CLASS(),
2240 * TRACEPOINT_EVENT_INSTANCE(), and TRACEPOINT_LOGLEVEL() here.
2245 #include <lttng/tracepoint-event.h>
2251 * `provider_name` with the name of your tracepoint provider.
2252 * `"tp.h"` with the name of your tracepoint provider header file.
2254 . Below the `#include <lttng/tracepoint.h>` line, put your
2255 <<defining-tracepoints,tracepoint definitions>>.
2257 Your tracepoint provider name must be unique amongst all the possible
2258 tracepoint provider names used on the same target system. We
2259 suggest to include the name of your project or company in the name,
2260 for example, `org_lttng_my_project_tpp`.
2262 TIP: [[lttng-gen-tp]]Use the man:lttng-gen-tp(1) tool to create
2263 this boilerplate for you. When using cmd:lttng-gen-tp, all you need to
2264 write are the <<defining-tracepoints,tracepoint definitions>>.
2267 [[defining-tracepoints]]
2268 ===== Create a tracepoint definition
2270 A _tracepoint definition_ defines, for a given tracepoint:
2272 * Its **input arguments**. They are the macro parameters that the
2273 `tracepoint()` macro accepts for this particular tracepoint
2274 in the source code of the user application.
2275 * Its **output event fields**. They are the sources of event fields
2276 that form the payload of any event that the execution of the
2277 `tracepoint()` macro emits for this particular tracepoint.
2279 Create a tracepoint definition by using the
2280 `TRACEPOINT_EVENT()` macro below the `#include <lttng/tracepoint.h>`
2282 <<tpp-header,tracepoint provider header file template>>.
2284 The syntax of the `TRACEPOINT_EVENT()` macro is:
2287 .`TRACEPOINT_EVENT()` macro syntax.
2290 /* Tracepoint provider name */
2293 /* Tracepoint name */
2296 /* Input arguments */
2301 /* Output event fields */
2310 * `provider_name` with your tracepoint provider name.
2311 * `tracepoint_name` with your tracepoint name.
2312 * `arguments` with the <<tpp-def-input-args,input arguments>>.
2313 * `fields` with the <<tpp-def-output-fields,output event field>>
2316 This tracepoint emits events named `provider_name:tracepoint_name`.
2319 .Event name length limitation
2321 The concatenation of the tracepoint provider name and the
2322 tracepoint name must not exceed **254{nbsp}characters**. If it does, the
2323 instrumented application compiles and runs, but LTTng throws multiple
2324 warnings and you could experience serious issues.
2327 [[tpp-def-input-args]]The syntax of the `TP_ARGS()` macro is:
2330 .`TP_ARGS()` macro syntax.
2339 * `type` with the C type of the argument.
2340 * `arg_name` with the argument name.
2342 You can repeat `type` and `arg_name` up to 10{nbsp}times to have more
2345 .`TP_ARGS()` usage with three arguments.
2357 The `TP_ARGS()` and `TP_ARGS(void)` forms are valid to create a
2358 tracepoint definition with no input arguments.
2360 [[tpp-def-output-fields]]The `TP_FIELDS()` macro contains a list of
2361 `ctf_*()` macros. Each `ctf_*()` macro defines one event field. See
2362 man:lttng-ust(3) for a complete description of the available `ctf_*()`
2363 macros. A `ctf_*()` macro specifies the type, size, and byte order of
2366 Each `ctf_*()` macro takes an _argument expression_ parameter. This is a
2367 C expression that the tracer evalutes at the `tracepoint()` macro site
2368 in the source code of the application. This expression provides the
2369 source of data of a field. The argument expression can include input
2370 argument names listed in the `TP_ARGS()` macro.
2372 Each `ctf_*()` macro also takes a _field name_ parameter. Field names
2373 must be unique within a given tracepoint definition.
2375 Here's a complete tracepoint definition example:
2377 .Tracepoint definition.
2379 The following tracepoint definition defines a tracepoint which takes
2380 three input arguments and has four output event fields.
2384 #include "my-custom-structure.h"
2390 const struct my_custom_structure*, my_custom_structure,
2395 ctf_string(query_field, query)
2396 ctf_float(double, ratio_field, ratio)
2397 ctf_integer(int, recv_size, my_custom_structure->recv_size)
2398 ctf_integer(int, send_size, my_custom_structure->send_size)
2403 Refer to this tracepoint definition with the `tracepoint()` macro in
2404 the source code of your application like this:
2408 tracepoint(my_provider, my_tracepoint,
2409 my_structure, some_ratio, the_query);
2413 NOTE: The LTTng tracer only evaluates tracepoint arguments at run time
2414 if they satisfy an enabled <<event,event rule>>.
2417 [[using-tracepoint-classes]]
2418 ===== Use a tracepoint class
2420 A _tracepoint class_ is a class of tracepoints which share the same
2421 output event field definitions. A _tracepoint instance_ is one
2422 instance of such a defined tracepoint class, with its own tracepoint
2425 The <<defining-tracepoints,`TRACEPOINT_EVENT()` macro>> is actually a
2426 shorthand which defines both a tracepoint class and a tracepoint
2427 instance at the same time.
2429 When you build a tracepoint provider package, the C or $$C++$$ compiler
2430 creates one serialization function for each **tracepoint class**. A
2431 serialization function is responsible for serializing the event fields
2432 of a tracepoint to a sub-buffer when tracing.
2434 For various performance reasons, when your situation requires multiple
2435 tracepoint definitions with different names, but with the same event
2436 fields, we recommend that you manually create a tracepoint class
2437 and instantiate as many tracepoint instances as needed. One positive
2438 effect of such a design, amongst other advantages, is that all
2439 tracepoint instances of the same tracepoint class reuse the same
2440 serialization function, thus reducing
2441 https://en.wikipedia.org/wiki/Cache_pollution[cache pollution].
2443 .Use a tracepoint class and tracepoint instances.
2445 Consider the following three tracepoint definitions:
2457 ctf_integer(int, userid, userid)
2458 ctf_integer(size_t, len, len)
2470 ctf_integer(int, userid, userid)
2471 ctf_integer(size_t, len, len)
2483 ctf_integer(int, userid, userid)
2484 ctf_integer(size_t, len, len)
2489 In this case, we create three tracepoint classes, with one implicit
2490 tracepoint instance for each of them: `get_account`, `get_settings`, and
2491 `get_transaction`. However, they all share the same event field names
2492 and types. Hence three identical, yet independent serialization
2493 functions are created when you build the tracepoint provider package.
2495 A better design choice is to define a single tracepoint class and three
2496 tracepoint instances:
2500 /* The tracepoint class */
2501 TRACEPOINT_EVENT_CLASS(
2502 /* Tracepoint provider name */
2505 /* Tracepoint class name */
2508 /* Input arguments */
2514 /* Output event fields */
2516 ctf_integer(int, userid, userid)
2517 ctf_integer(size_t, len, len)
2521 /* The tracepoint instances */
2522 TRACEPOINT_EVENT_INSTANCE(
2523 /* Tracepoint provider name */
2526 /* Tracepoint class name */
2529 /* Tracepoint name */
2532 /* Input arguments */
2538 TRACEPOINT_EVENT_INSTANCE(
2547 TRACEPOINT_EVENT_INSTANCE(
2560 [[assigning-log-levels]]
2561 ===== Assign a log level to a tracepoint definition
2563 Assign a _log level_ to a <<defining-tracepoints,tracepoint definition>>
2564 with the `TRACEPOINT_LOGLEVEL()` macro.
2566 Assigning different levels of severity to tracepoint definitions can
2567 be useful: when you <<enabling-disabling-events,create an event rule>>,
2568 you can target tracepoints having a log level as severe as a specific
2571 The concept of LTTng-UST log levels is similar to the levels found
2572 in typical logging frameworks:
2574 * In a logging framework, the log level is given by the function
2575 or method name you use at the log statement site: `debug()`,
2576 `info()`, `warn()`, `error()`, and so on.
2577 * In LTTng-UST, you statically assign the log level to a tracepoint
2578 definition; any `tracepoint()` macro invocation which refers to
2579 this definition has this log level.
2581 You must use `TRACEPOINT_LOGLEVEL()` _after_ the
2582 <<defining-tracepoints,`TRACEPOINT_EVENT()`>> or
2583 <<using-tracepoint-classes,`TRACEPOINT_INSTANCE()`>> macro for a given
2586 The syntax of the `TRACEPOINT_LOGLEVEL()` macro is:
2589 .`TRACEPOINT_LOGLEVEL()` macro syntax.
2591 TRACEPOINT_LOGLEVEL(provider_name, tracepoint_name, log_level)
2596 * `provider_name` with the tracepoint provider name.
2597 * `tracepoint_name` with the tracepoint name.
2598 * `log_level` with the log level to assign to the tracepoint
2599 definition named `tracepoint_name` in the `provider_name`
2600 tracepoint provider.
2602 See man:lttng-ust(3) for a list of available log level names.
2604 .Assign the `TRACE_DEBUG_UNIT` log level to a tracepoint definition.
2608 /* Tracepoint definition */
2617 ctf_integer(int, userid, userid)
2618 ctf_integer(size_t, len, len)
2622 /* Log level assignment */
2623 TRACEPOINT_LOGLEVEL(my_app, get_transaction, TRACE_DEBUG_UNIT)
2629 ===== Create a tracepoint provider package source file
2631 A _tracepoint provider package source file_ is a C source file which
2632 includes a <<tpp-header,tracepoint provider header file>> to expand its
2633 macros into event serialization and other functions.
2635 Use the following tracepoint provider package source file template:
2638 .Tracepoint provider package source file template.
2640 #define TRACEPOINT_CREATE_PROBES
2645 Replace `tp.h` with the name of your <<tpp-header,tracepoint provider
2646 header file>> name. You may also include more than one tracepoint
2647 provider header file here to create a tracepoint provider package
2648 holding more than one tracepoint providers.
2651 [[probing-the-application-source-code]]
2652 ==== Add tracepoints to the source code of an application
2654 Once you <<tpp-header,create a tracepoint provider header file>>, use
2655 the `tracepoint()` macro in the source code of your application to
2656 insert the tracepoints that this header
2657 <<defining-tracepoints,defines>>.
2659 The `tracepoint()` macro takes at least two parameters: the tracepoint
2660 provider name and the tracepoint name. The corresponding tracepoint
2661 definition defines the other parameters.
2663 .`tracepoint()` usage.
2665 The following <<defining-tracepoints,tracepoint definition>> defines a
2666 tracepoint which takes two input arguments and has two output event
2670 .Tracepoint provider header file.
2672 #include "my-custom-structure.h"
2679 const char*, cmd_name
2682 ctf_string(cmd_name, cmd_name)
2683 ctf_integer(int, number_of_args, argc)
2688 Refer to this tracepoint definition with the `tracepoint()` macro in
2689 the source code of your application like this:
2692 .Application source file.
2696 int main(int argc, char* argv[])
2698 tracepoint(my_provider, my_tracepoint, argc, argv[0]);
2704 Note how the source code of the application includes
2705 the tracepoint provider header file containing the tracepoint
2706 definitions to use, path:{tp.h}.
2709 .`tracepoint()` usage with a complex tracepoint definition.
2711 Consider this complex tracepoint definition, where multiple event
2712 fields refer to the same input arguments in their argument expression
2716 .Tracepoint provider header file.
2718 /* For `struct stat` */
2719 #include <sys/types.h>
2720 #include <sys/stat.h>
2732 ctf_integer(int, my_constant_field, 23 + 17)
2733 ctf_integer(int, my_int_arg_field, my_int_arg)
2734 ctf_integer(int, my_int_arg_field2, my_int_arg * my_int_arg)
2735 ctf_integer(int, sum4_field, my_str_arg[0] + my_str_arg[1] +
2736 my_str_arg[2] + my_str_arg[3])
2737 ctf_string(my_str_arg_field, my_str_arg)
2738 ctf_integer_hex(off_t, size_field, st->st_size)
2739 ctf_float(double, size_dbl_field, (double) st->st_size)
2740 ctf_sequence_text(char, half_my_str_arg_field, my_str_arg,
2741 size_t, strlen(my_str_arg) / 2)
2746 Refer to this tracepoint definition with the `tracepoint()` macro in
2747 the source code of your application like this:
2750 .Application source file.
2752 #define TRACEPOINT_DEFINE
2759 stat("/etc/fstab", &s);
2760 tracepoint(my_provider, my_tracepoint, 23, "Hello, World!", &s);
2766 If you look at the event record that LTTng writes when tracing this
2767 program, assuming the file size of path:{/etc/fstab} is 301{nbsp}bytes,
2768 it should look like this:
2770 .Event record fields
2772 |Field name |Field value
2773 |`my_constant_field` |40
2774 |`my_int_arg_field` |23
2775 |`my_int_arg_field2` |529
2777 |`my_str_arg_field` |`Hello, World!`
2778 |`size_field` |0x12d
2779 |`size_dbl_field` |301.0
2780 |`half_my_str_arg_field` |`Hello,`
2784 Sometimes, the arguments you pass to `tracepoint()` are expensive to
2785 compute--they use the call stack, for example. To avoid this computation
2786 when the tracepoint is disabled, use the `tracepoint_enabled()` and
2787 `do_tracepoint()` macros.
2789 The syntax of the `tracepoint_enabled()` and `do_tracepoint()` macros
2793 .`tracepoint_enabled()` and `do_tracepoint()` macros syntax.
2795 tracepoint_enabled(provider_name, tracepoint_name)
2796 do_tracepoint(provider_name, tracepoint_name, ...)
2801 * `provider_name` with the tracepoint provider name.
2802 * `tracepoint_name` with the tracepoint name.
2804 `tracepoint_enabled()` returns a non-zero value if the tracepoint named
2805 `tracepoint_name` from the provider named `provider_name` is enabled
2808 `do_tracepoint()` is like `tracepoint()`, except that it doesn't check
2809 if the tracepoint is enabled. Using `tracepoint()` with
2810 `tracepoint_enabled()` is dangerous since `tracepoint()` also contains
2811 the `tracepoint_enabled()` check, thus a race condition is
2812 possible in this situation:
2815 .Possible race condition when using `tracepoint_enabled()` with `tracepoint()`.
2817 if (tracepoint_enabled(my_provider, my_tracepoint)) {
2818 stuff = prepare_stuff();
2821 tracepoint(my_provider, my_tracepoint, stuff);
2824 If the tracepoint is enabled after the condition, then `stuff` isn't
2825 prepared: the emitted event will either contain wrong data, or the whole
2826 application could crash (segmentation fault, for example).
2828 NOTE: Neither `tracepoint_enabled()` nor `do_tracepoint()` have an
2829 `STAP_PROBEV()` call. If you need it, you must emit
2833 [[building-tracepoint-providers-and-user-application]]
2834 ==== Build and link a tracepoint provider package and an application
2836 Once you have one or more <<tpp-header,tracepoint provider header
2837 files>> and a <<tpp-source,tracepoint provider package source file>>,
2838 create the tracepoint provider package by compiling its source
2839 file. From here, multiple build and run scenarios are possible. The
2840 following table shows common application and library configurations
2841 along with the required command lines to achieve them.
2843 In the following diagrams, we use the following file names:
2846 Executable application.
2849 Application object file.
2852 Tracepoint provider package object file.
2855 Tracepoint provider package archive file.
2858 Tracepoint provider package shared object file.
2861 User library object file.
2864 User library shared object file.
2866 We use the following symbols in the diagrams of table below:
2869 .Symbols used in the build scenario diagrams.
2870 image::ust-sit-symbols.png[]
2872 We assume that path:{.} is part of the env:LD_LIBRARY_PATH environment
2873 variable in the following instructions.
2875 [role="growable ust-scenarios",cols="asciidoc,asciidoc"]
2876 .Common tracepoint provider package scenarios.
2878 |Scenario |Instructions
2881 The instrumented application is statically linked with
2882 the tracepoint provider package object.
2884 image::ust-sit+app-linked-with-tp-o+app-instrumented.png[]
2887 include::../common/ust-sit-step-tp-o.txt[]
2889 To build the instrumented application:
2891 . In path:{app.c}, before including path:{tpp.h}, add the following line:
2896 #define TRACEPOINT_DEFINE
2900 . Compile the application source file:
2909 . Build the application:
2914 $ gcc -o app app.o tpp.o -llttng-ust -ldl
2918 To run the instrumented application:
2920 * Start the application:
2930 The instrumented application is statically linked with the
2931 tracepoint provider package archive file.
2933 image::ust-sit+app-linked-with-tp-a+app-instrumented.png[]
2936 To create the tracepoint provider package archive file:
2938 . Compile the <<tpp-source,tracepoint provider package source file>>:
2947 . Create the tracepoint provider package archive file:
2952 $ ar rcs tpp.a tpp.o
2956 To build the instrumented application:
2958 . In path:{app.c}, before including path:{tpp.h}, add the following line:
2963 #define TRACEPOINT_DEFINE
2967 . Compile the application source file:
2976 . Build the application:
2981 $ gcc -o app app.o tpp.a -llttng-ust -ldl
2985 To run the instrumented application:
2987 * Start the application:
2997 The instrumented application is linked with the tracepoint provider
2998 package shared object.
3000 image::ust-sit+app-linked-with-tp-so+app-instrumented.png[]
3003 include::../common/ust-sit-step-tp-so.txt[]
3005 To build the instrumented application:
3007 . In path:{app.c}, before including path:{tpp.h}, add the following line:
3012 #define TRACEPOINT_DEFINE
3016 . Compile the application source file:
3025 . Build the application:
3030 $ gcc -o app app.o -ldl -L. -ltpp
3034 To run the instrumented application:
3036 * Start the application:
3046 The tracepoint provider package shared object is preloaded before the
3047 instrumented application starts.
3049 image::ust-sit+tp-so-preloaded+app-instrumented.png[]
3052 include::../common/ust-sit-step-tp-so.txt[]
3054 To build the instrumented application:
3056 . In path:{app.c}, before including path:{tpp.h}, add the
3062 #define TRACEPOINT_DEFINE
3063 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3067 . Compile the application source file:
3076 . Build the application:
3081 $ gcc -o app app.o -ldl
3085 To run the instrumented application with tracing support:
3087 * Preload the tracepoint provider package shared object and
3088 start the application:
3093 $ LD_PRELOAD=./libtpp.so ./app
3097 To run the instrumented application without tracing support:
3099 * Start the application:
3109 The instrumented application dynamically loads the tracepoint provider
3110 package shared object.
3112 image::ust-sit+app-dlopens-tp-so+app-instrumented.png[]
3115 include::../common/ust-sit-step-tp-so.txt[]
3117 To build the instrumented application:
3119 . In path:{app.c}, before including path:{tpp.h}, add the
3125 #define TRACEPOINT_DEFINE
3126 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3130 . Compile the application source file:
3139 . Build the application:
3144 $ gcc -o app app.o -ldl
3148 To run the instrumented application:
3150 * Start the application:
3160 The application is linked with the instrumented user library.
3162 The instrumented user library is statically linked with the tracepoint
3163 provider package object file.
3165 image::ust-sit+app-linked-with-lib+lib-linked-with-tp-o+lib-instrumented.png[]
3168 include::../common/ust-sit-step-tp-o-fpic.txt[]
3170 To build the instrumented user library:
3172 . In path:{emon.c}, before including path:{tpp.h}, add the
3178 #define TRACEPOINT_DEFINE
3182 . Compile the user library source file:
3187 $ gcc -I. -fpic -c emon.c
3191 . Build the user library shared object:
3196 $ gcc -shared -o libemon.so emon.o tpp.o -llttng-ust -ldl
3200 To build the application:
3202 . Compile the application source file:
3211 . Build the application:
3216 $ gcc -o app app.o -L. -lemon
3220 To run the application:
3222 * Start the application:
3232 The application is linked with the instrumented user library.
3234 The instrumented user library is linked with the tracepoint provider
3235 package shared object.
3237 image::ust-sit+app-linked-with-lib+lib-linked-with-tp-so+lib-instrumented.png[]
3240 include::../common/ust-sit-step-tp-so.txt[]
3242 To build the instrumented user library:
3244 . In path:{emon.c}, before including path:{tpp.h}, add the
3250 #define TRACEPOINT_DEFINE
3254 . Compile the user library source file:
3259 $ gcc -I. -fpic -c emon.c
3263 . Build the user library shared object:
3268 $ gcc -shared -o libemon.so emon.o -ldl -L. -ltpp
3272 To build the application:
3274 . Compile the application source file:
3283 . Build the application:
3288 $ gcc -o app app.o -L. -lemon
3292 To run the application:
3294 * Start the application:
3304 The tracepoint provider package shared object is preloaded before the
3307 The application is linked with the instrumented user library.
3309 image::ust-sit+tp-so-preloaded+app-linked-with-lib+lib-instrumented.png[]
3312 include::../common/ust-sit-step-tp-so.txt[]
3314 To build the instrumented user library:
3316 . In path:{emon.c}, before including path:{tpp.h}, add the
3322 #define TRACEPOINT_DEFINE
3323 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3327 . Compile the user library source file:
3332 $ gcc -I. -fpic -c emon.c
3336 . Build the user library shared object:
3341 $ gcc -shared -o libemon.so emon.o -ldl
3345 To build the application:
3347 . Compile the application source file:
3356 . Build the application:
3361 $ gcc -o app app.o -L. -lemon
3365 To run the application with tracing support:
3367 * Preload the tracepoint provider package shared object and
3368 start the application:
3373 $ LD_PRELOAD=./libtpp.so ./app
3377 To run the application without tracing support:
3379 * Start the application:
3389 The application is linked with the instrumented user library.
3391 The instrumented user library dynamically loads the tracepoint provider
3392 package shared object.
3394 image::ust-sit+app-linked-with-lib+lib-dlopens-tp-so+lib-instrumented.png[]
3397 include::../common/ust-sit-step-tp-so.txt[]
3399 To build the instrumented user library:
3401 . In path:{emon.c}, before including path:{tpp.h}, add the
3407 #define TRACEPOINT_DEFINE
3408 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3412 . Compile the user library source file:
3417 $ gcc -I. -fpic -c emon.c
3421 . Build the user library shared object:
3426 $ gcc -shared -o libemon.so emon.o -ldl
3430 To build the application:
3432 . Compile the application source file:
3441 . Build the application:
3446 $ gcc -o app app.o -L. -lemon
3450 To run the application:
3452 * Start the application:
3462 The application dynamically loads the instrumented user library.
3464 The instrumented user library is linked with the tracepoint provider
3465 package shared object.
3467 image::ust-sit+app-dlopens-lib+lib-linked-with-tp-so+lib-instrumented.png[]
3470 include::../common/ust-sit-step-tp-so.txt[]
3472 To build the instrumented user library:
3474 . In path:{emon.c}, before including path:{tpp.h}, add the
3480 #define TRACEPOINT_DEFINE
3484 . Compile the user library source file:
3489 $ gcc -I. -fpic -c emon.c
3493 . Build the user library shared object:
3498 $ gcc -shared -o libemon.so emon.o -ldl -L. -ltpp
3502 To build the application:
3504 . Compile the application source file:
3513 . Build the application:
3518 $ gcc -o app app.o -ldl -L. -lemon
3522 To run the application:
3524 * Start the application:
3534 The application dynamically loads the instrumented user library.
3536 The instrumented user library dynamically loads the tracepoint provider
3537 package shared object.
3539 image::ust-sit+app-dlopens-lib+lib-dlopens-tp-so+lib-instrumented.png[]
3542 include::../common/ust-sit-step-tp-so.txt[]
3544 To build the instrumented user library:
3546 . In path:{emon.c}, before including path:{tpp.h}, add the
3552 #define TRACEPOINT_DEFINE
3553 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3557 . Compile the user library source file:
3562 $ gcc -I. -fpic -c emon.c
3566 . Build the user library shared object:
3571 $ gcc -shared -o libemon.so emon.o -ldl
3575 To build the application:
3577 . Compile the application source file:
3586 . Build the application:
3591 $ gcc -o app app.o -ldl -L. -lemon
3595 To run the application:
3597 * Start the application:
3607 The tracepoint provider package shared object is preloaded before the
3610 The application dynamically loads the instrumented user library.
3612 image::ust-sit+tp-so-preloaded+app-dlopens-lib+lib-instrumented.png[]
3615 include::../common/ust-sit-step-tp-so.txt[]
3617 To build the instrumented user library:
3619 . In path:{emon.c}, before including path:{tpp.h}, add the
3625 #define TRACEPOINT_DEFINE
3626 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3630 . Compile the user library source file:
3635 $ gcc -I. -fpic -c emon.c
3639 . Build the user library shared object:
3644 $ gcc -shared -o libemon.so emon.o -ldl
3648 To build the application:
3650 . Compile the application source file:
3659 . Build the application:
3664 $ gcc -o app app.o -L. -lemon
3668 To run the application with tracing support:
3670 * Preload the tracepoint provider package shared object and
3671 start the application:
3676 $ LD_PRELOAD=./libtpp.so ./app
3680 To run the application without tracing support:
3682 * Start the application:
3692 The application is statically linked with the tracepoint provider
3693 package object file.
3695 The application is linked with the instrumented user library.
3697 image::ust-sit+app-linked-with-tp-o+app-linked-with-lib+lib-instrumented.png[]
3700 include::../common/ust-sit-step-tp-o.txt[]
3702 To build the instrumented user library:
3704 . In path:{emon.c}, before including path:{tpp.h}, add the
3710 #define TRACEPOINT_DEFINE
3714 . Compile the user library source file:
3719 $ gcc -I. -fpic -c emon.c
3723 . Build the user library shared object:
3728 $ gcc -shared -o libemon.so emon.o
3732 To build the application:
3734 . Compile the application source file:
3743 . Build the application:
3748 $ gcc -o app app.o tpp.o -llttng-ust -ldl -L. -lemon
3752 To run the instrumented application:
3754 * Start the application:
3764 The application is statically linked with the tracepoint provider
3765 package object file.
3767 The application dynamically loads the instrumented user library.
3769 image::ust-sit+app-linked-with-tp-o+app-dlopens-lib+lib-instrumented.png[]
3772 include::../common/ust-sit-step-tp-o.txt[]
3774 To build the application:
3776 . In path:{app.c}, before including path:{tpp.h}, add the following line:
3781 #define TRACEPOINT_DEFINE
3785 . Compile the application source file:
3794 . Build the application:
3799 $ gcc -Wl,--export-dynamic -o app app.o tpp.o \
3804 The `--export-dynamic` option passed to the linker is necessary for the
3805 dynamically loaded library to ``see'' the tracepoint symbols defined in
3808 To build the instrumented user library:
3810 . Compile the user library source file:
3815 $ gcc -I. -fpic -c emon.c
3819 . Build the user library shared object:
3824 $ gcc -shared -o libemon.so emon.o
3828 To run the application:
3830 * Start the application:
3841 [[using-lttng-ust-with-daemons]]
3842 ===== Use noch:{LTTng-UST} with daemons
3844 If your instrumented application calls man:fork(2), man:clone(2),
3845 or BSD's man:rfork(2), without a following man:exec(3)-family
3846 system call, you must preload the path:{liblttng-ust-fork.so} shared
3847 object when you start the application.
3851 $ LD_PRELOAD=liblttng-ust-fork.so ./my-app
3854 If your tracepoint provider package is
3855 a shared library which you also preload, you must put both
3856 shared objects in env:LD_PRELOAD:
3860 $ LD_PRELOAD=liblttng-ust-fork.so:/path/to/tp.so ./my-app
3866 ===== Use noch:{LTTng-UST} with applications which close file descriptors that don't belong to them
3868 If your instrumented application closes one or more file descriptors
3869 which it did not open itself, you must preload the
3870 path:{liblttng-ust-fd.so} shared object when you start the application:
3874 $ LD_PRELOAD=liblttng-ust-fd.so ./my-app
3877 Typical use cases include closing all the file descriptors after
3878 man:fork(2) or man:rfork(2) and buggy applications doing
3882 [[lttng-ust-pkg-config]]
3883 ===== Use noch:{pkg-config}
3885 On some distributions, LTTng-UST ships with a
3886 https://www.freedesktop.org/wiki/Software/pkg-config/[pkg-config]
3887 metadata file. If this is your case, then use cmd:pkg-config to
3888 build an application on the command line:
3892 $ gcc -o my-app my-app.o tp.o $(pkg-config --cflags --libs lttng-ust)
3896 [[instrumenting-32-bit-app-on-64-bit-system]]
3897 ===== [[advanced-instrumenting-techniques]]Build a 32-bit instrumented application for a 64-bit target system
3899 In order to trace a 32-bit application running on a 64-bit system,
3900 LTTng must use a dedicated 32-bit
3901 <<lttng-consumerd,consumer daemon>>.
3903 The following steps show how to build and install a 32-bit consumer
3904 daemon, which is _not_ part of the default 64-bit LTTng build, how to
3905 build and install the 32-bit LTTng-UST libraries, and how to build and
3906 link an instrumented 32-bit application in that context.
3908 To build a 32-bit instrumented application for a 64-bit target system,
3909 assuming you have a fresh target system with no installed Userspace RCU
3912 . Download, build, and install a 32-bit version of Userspace RCU:
3917 $ cd $(mktemp -d) &&
3918 wget http://lttng.org/files/urcu/userspace-rcu-latest-0.9.tar.bz2 &&
3919 tar -xf userspace-rcu-latest-0.9.tar.bz2 &&
3920 cd userspace-rcu-0.9.* &&
3921 ./configure --libdir=/usr/local/lib32 CFLAGS=-m32 &&
3923 sudo make install &&
3928 . Using the package manager of your distribution, or from source,
3929 install the following 32-bit versions of the following dependencies of
3930 LTTng-tools and LTTng-UST:
3933 * https://sourceforge.net/projects/libuuid/[libuuid]
3934 * http://directory.fsf.org/wiki/Popt[popt]
3935 * http://www.xmlsoft.org/[libxml2]
3938 . Download, build, and install a 32-bit version of the latest
3939 LTTng-UST{nbsp}{revision}:
3944 $ cd $(mktemp -d) &&
3945 wget http://lttng.org/files/lttng-ust/lttng-ust-latest-2.12.tar.bz2 &&
3946 tar -xf lttng-ust-latest-2.12.tar.bz2 &&
3947 cd lttng-ust-2.12.* &&
3948 ./configure --libdir=/usr/local/lib32 \
3949 CFLAGS=-m32 CXXFLAGS=-m32 \
3950 LDFLAGS='-L/usr/local/lib32 -L/usr/lib32' &&
3952 sudo make install &&
3959 Depending on your distribution,
3960 32-bit libraries could be installed at a different location than
3961 `/usr/lib32`. For example, Debian is known to install
3962 some 32-bit libraries in `/usr/lib/i386-linux-gnu`.
3964 In this case, make sure to set `LDFLAGS` to all the
3965 relevant 32-bit library paths, for example:
3969 $ LDFLAGS='-L/usr/lib/i386-linux-gnu -L/usr/lib32'
3973 . Download the latest LTTng-tools{nbsp}{revision}, build, and install
3974 the 32-bit consumer daemon:
3979 $ cd $(mktemp -d) &&
3980 wget http://lttng.org/files/lttng-tools/lttng-tools-latest-2.12.tar.bz2 &&
3981 tar -xf lttng-tools-latest-2.12.tar.bz2 &&
3982 cd lttng-tools-2.12.* &&
3983 ./configure --libdir=/usr/local/lib32 CFLAGS=-m32 CXXFLAGS=-m32 \
3984 LDFLAGS='-L/usr/local/lib32 -L/usr/lib32' \
3985 --disable-bin-lttng --disable-bin-lttng-crash \
3986 --disable-bin-lttng-relayd --disable-bin-lttng-sessiond &&
3988 cd src/bin/lttng-consumerd &&
3989 sudo make install &&
3994 . From your distribution or from source,
3995 <<installing-lttng,install>> the 64-bit versions of
3996 LTTng-UST and Userspace RCU.
3997 . Download, build, and install the 64-bit version of the
3998 latest LTTng-tools{nbsp}{revision}:
4003 $ cd $(mktemp -d) &&
4004 wget http://lttng.org/files/lttng-tools/lttng-tools-latest-2.12.tar.bz2 &&
4005 tar -xf lttng-tools-latest-2.12.tar.bz2 &&
4006 cd lttng-tools-2.12.* &&
4007 ./configure --with-consumerd32-libdir=/usr/local/lib32 \
4008 --with-consumerd32-bin=/usr/local/lib32/lttng/libexec/lttng-consumerd &&
4010 sudo make install &&
4015 . Pass the following options to man:gcc(1), man:g++(1), or man:clang(1)
4016 when linking your 32-bit application:
4019 -m32 -L/usr/lib32 -L/usr/local/lib32 \
4020 -Wl,-rpath,/usr/lib32,-rpath,/usr/local/lib32
4023 For example, let's rebuild the quick start example in
4024 <<tracing-your-own-user-application,Trace a user application>> as an
4025 instrumented 32-bit application:
4030 $ gcc -m32 -c -I. hello-tp.c
4031 $ gcc -m32 -c hello.c
4032 $ gcc -m32 -o hello hello.o hello-tp.o \
4033 -L/usr/lib32 -L/usr/local/lib32 \
4034 -Wl,-rpath,/usr/lib32,-rpath,/usr/local/lib32 \
4039 No special action is required to execute the 32-bit application and
4040 to trace it: use the command-line man:lttng(1) tool as usual.
4047 man:tracef(3) is a small LTTng-UST API designed for quick,
4048 man:printf(3)-like instrumentation without the burden of
4049 <<tracepoint-provider,creating>> and
4050 <<building-tracepoint-providers-and-user-application,building>>
4051 a tracepoint provider package.
4053 To use `tracef()` in your application:
4055 . In the C or C++ source files where you need to use `tracef()`,
4056 include `<lttng/tracef.h>`:
4061 #include <lttng/tracef.h>
4065 . In the source code of the application, use `tracef()` like you would
4073 tracef("my message: %d (%s)", my_integer, my_string);
4079 . Link your application with `liblttng-ust`:
4084 $ gcc -o app app.c -llttng-ust
4088 To trace the events that `tracef()` calls emit:
4090 * <<enabling-disabling-events,Create an event rule>> which matches the
4091 `lttng_ust_tracef:*` event name:
4096 $ lttng enable-event --userspace 'lttng_ust_tracef:*'
4101 .Limitations of `tracef()`
4103 The `tracef()` utility function was developed to make user space tracing
4104 super simple, albeit with notable disadvantages compared to
4105 <<defining-tracepoints,user-defined tracepoints>>:
4107 * All the emitted events have the same tracepoint provider and
4108 tracepoint names, respectively `lttng_ust_tracef` and `event`.
4109 * There is no static type checking.
4110 * The only event record field you actually get, named `msg`, is a string
4111 potentially containing the values you passed to `tracef()`
4112 using your own format string. This also means that you can't filter
4113 events with a custom expression at run time because there are no
4115 * Since `tracef()` uses the man:vasprintf(3) function of the
4116 C{nbsp}standard library behind the scenes to format the strings at run
4117 time, its expected performance is lower than with user-defined
4118 tracepoints, which don't require a conversion to a string.
4120 Taking this into consideration, `tracef()` is useful for some quick
4121 prototyping and debugging, but you shouldn't consider it for any
4122 permanent and serious applicative instrumentation.
4128 ==== Use `tracelog()`
4130 The man:tracelog(3) API is very similar to <<tracef,`tracef()`>>, with
4131 the difference that it accepts an additional log level parameter.
4133 The goal of `tracelog()` is to ease the migration from logging to
4136 To use `tracelog()` in your application:
4138 . In the C or C++ source files where you need to use `tracelog()`,
4139 include `<lttng/tracelog.h>`:
4144 #include <lttng/tracelog.h>
4148 . In the source code of the application, use `tracelog()` like you would
4149 use man:printf(3), except for the first parameter which is the log
4157 tracelog(TRACE_WARNING, "my message: %d (%s)",
4158 my_integer, my_string);
4164 See man:lttng-ust(3) for a list of available log level names.
4166 . Link your application with `liblttng-ust`:
4171 $ gcc -o app app.c -llttng-ust
4175 To trace the events that `tracelog()` calls emit with a log level
4176 _as severe as_ a specific log level:
4178 * <<enabling-disabling-events,Create an event rule>> which matches the
4179 `lttng_ust_tracelog:*` event name and a minimum level
4185 $ lttng enable-event --userspace 'lttng_ust_tracelog:*'
4186 --loglevel=TRACE_WARNING
4190 To trace the events that `tracelog()` calls emit with a
4191 _specific log level_:
4193 * Create an event rule which matches the `lttng_ust_tracelog:*`
4194 event name and a specific log level:
4199 $ lttng enable-event --userspace 'lttng_ust_tracelog:*'
4200 --loglevel-only=TRACE_INFO
4205 [[prebuilt-ust-helpers]]
4206 === Prebuilt user space tracing helpers
4208 The LTTng-UST package provides a few helpers in the form of preloadable
4209 shared objects which automatically instrument system functions and
4212 The helper shared objects are normally found in dir:{/usr/lib}. If you
4213 built LTTng-UST <<building-from-source,from source>>, they are probably
4214 located in dir:{/usr/local/lib}.
4216 The installed user space tracing helpers in LTTng-UST{nbsp}{revision}
4219 path:{liblttng-ust-libc-wrapper.so}::
4220 path:{liblttng-ust-pthread-wrapper.so}::
4221 <<liblttng-ust-libc-pthread-wrapper,C{nbsp}standard library
4222 memory and POSIX threads function tracing>>.
4224 path:{liblttng-ust-cyg-profile.so}::
4225 path:{liblttng-ust-cyg-profile-fast.so}::
4226 <<liblttng-ust-cyg-profile,Function entry and exit tracing>>.
4228 path:{liblttng-ust-dl.so}::
4229 <<liblttng-ust-dl,Dynamic linker tracing>>.
4231 To use a user space tracing helper with any user application:
4233 * Preload the helper shared object when you start the application:
4238 $ LD_PRELOAD=liblttng-ust-libc-wrapper.so my-app
4242 You can preload more than one helper:
4247 $ LD_PRELOAD=liblttng-ust-libc-wrapper.so:liblttng-ust-dl.so my-app
4253 [[liblttng-ust-libc-pthread-wrapper]]
4254 ==== Instrument C standard library memory and POSIX threads functions
4256 The path:{liblttng-ust-libc-wrapper.so} and
4257 path:{liblttng-ust-pthread-wrapper.so} helpers
4258 add instrumentation to some C standard library and POSIX
4262 .Functions instrumented by preloading path:{liblttng-ust-libc-wrapper.so}.
4264 |TP provider name |TP name |Instrumented function
4266 .6+|`lttng_ust_libc` |`malloc` |man:malloc(3)
4267 |`calloc` |man:calloc(3)
4268 |`realloc` |man:realloc(3)
4269 |`free` |man:free(3)
4270 |`memalign` |man:memalign(3)
4271 |`posix_memalign` |man:posix_memalign(3)
4275 .Functions instrumented by preloading path:{liblttng-ust-pthread-wrapper.so}.
4277 |TP provider name |TP name |Instrumented function
4279 .4+|`lttng_ust_pthread` |`pthread_mutex_lock_req` |man:pthread_mutex_lock(3p) (request time)
4280 |`pthread_mutex_lock_acq` |man:pthread_mutex_lock(3p) (acquire time)
4281 |`pthread_mutex_trylock` |man:pthread_mutex_trylock(3p)
4282 |`pthread_mutex_unlock` |man:pthread_mutex_unlock(3p)
4285 When you preload the shared object, it replaces the functions listed
4286 in the previous tables by wrappers which contain tracepoints and call
4287 the replaced functions.
4290 [[liblttng-ust-cyg-profile]]
4291 ==== Instrument function entry and exit
4293 The path:{liblttng-ust-cyg-profile*.so} helpers can add instrumentation
4294 to the entry and exit points of functions.
4296 man:gcc(1) and man:clang(1) have an option named
4297 https://gcc.gnu.org/onlinedocs/gcc/Instrumentation-Options.html[`-finstrument-functions`]
4298 which generates instrumentation calls for entry and exit to functions.
4299 The LTTng-UST function tracing helpers,
4300 path:{liblttng-ust-cyg-profile.so} and
4301 path:{liblttng-ust-cyg-profile-fast.so}, take advantage of this feature
4302 to add tracepoints to the two generated functions (which contain
4303 `cyg_profile` in their names, hence the name of the helper).
4305 To use the LTTng-UST function tracing helper, the source files to
4306 instrument must be built using the `-finstrument-functions` compiler
4309 There are two versions of the LTTng-UST function tracing helper:
4311 * **path:{liblttng-ust-cyg-profile-fast.so}** is a lightweight variant
4312 that you should only use when it can be _guaranteed_ that the
4313 complete event stream is recorded without any lost event record.
4314 Any kind of duplicate information is left out.
4316 Assuming no event record is lost, having only the function addresses on
4317 entry is enough to create a call graph, since an event record always
4318 contains the ID of the CPU that generated it.
4320 Use a tool like man:addr2line(1) to convert function addresses back to
4321 source file names and line numbers.
4323 * **path:{liblttng-ust-cyg-profile.so}** is a more robust variant
4324 which also works in use cases where event records might get discarded or
4325 not recorded from application startup.
4326 In these cases, the trace analyzer needs more information to be
4327 able to reconstruct the program flow.
4329 See man:lttng-ust-cyg-profile(3) to learn more about the instrumentation
4330 points of this helper.
4332 All the tracepoints that this helper provides have the
4333 log level `TRACE_DEBUG_FUNCTION` (see man:lttng-ust(3)).
4335 TIP: It's sometimes a good idea to limit the number of source files that
4336 you compile with the `-finstrument-functions` option to prevent LTTng
4337 from writing an excessive amount of trace data at run time. When using
4339 `-finstrument-functions-exclude-function-list` option to avoid
4340 instrument entries and exits of specific function names.
4345 ==== Instrument the dynamic linker
4347 The path:{liblttng-ust-dl.so} helper adds instrumentation to the
4348 man:dlopen(3) and man:dlclose(3) function calls.
4350 See man:lttng-ust-dl(3) to learn more about the instrumentation points
4355 [[java-application]]
4356 === User space Java agent
4358 You can instrument any Java application which uses one of the following
4361 * The https://docs.oracle.com/javase/7/docs/api/java/util/logging/package-summary.html[**`java.util.logging`**]
4362 (JUL) core logging facilities.
4363 * http://logging.apache.org/log4j/1.2/[**Apache log4j{nbsp}1.2**], since
4364 LTTng{nbsp}2.6. Note that Apache Log4j{nbsp}2 isn't supported.
4367 .LTTng-UST Java agent imported by a Java application.
4368 image::java-app.png[]
4370 Note that the methods described below are new in LTTng{nbsp}2.8.
4371 Previous LTTng versions use another technique.
4373 NOTE: We use http://openjdk.java.net/[OpenJDK]{nbsp}8 for development
4374 and https://ci.lttng.org/[continuous integration], thus this version is
4375 directly supported. However, the LTTng-UST Java agent is also tested
4376 with OpenJDK{nbsp}7.
4381 ==== Use the LTTng-UST Java agent for `java.util.logging`
4383 To use the LTTng-UST Java agent in a Java application which uses
4384 `java.util.logging` (JUL):
4386 . In the source code of the Java application, import the LTTng-UST log
4387 handler package for `java.util.logging`:
4392 import org.lttng.ust.agent.jul.LttngLogHandler;
4396 . Create an LTTng-UST JUL log handler:
4401 Handler lttngUstLogHandler = new LttngLogHandler();
4405 . Add this handler to the JUL loggers which should emit LTTng events:
4410 Logger myLogger = Logger.getLogger("some-logger");
4412 myLogger.addHandler(lttngUstLogHandler);
4416 . Use `java.util.logging` log statements and configuration as usual.
4417 The loggers with an attached LTTng-UST log handler can emit
4420 . Before exiting the application, remove the LTTng-UST log handler from
4421 the loggers attached to it and call its `close()` method:
4426 myLogger.removeHandler(lttngUstLogHandler);
4427 lttngUstLogHandler.close();
4431 This isn't strictly necessary, but it is recommended for a clean
4432 disposal of the resources of the handler.
4434 . Include the common and JUL-specific JAR files of the LTTng-UST Java agent,
4435 path:{lttng-ust-agent-common.jar} and path:{lttng-ust-agent-jul.jar},
4437 https://docs.oracle.com/javase/tutorial/essential/environment/paths.html[class
4438 path] when you build the Java application.
4440 The JAR files are typically located in dir:{/usr/share/java}.
4442 IMPORTANT: The LTTng-UST Java agent must be
4443 <<installing-lttng,installed>> for the logging framework your
4446 .Use the LTTng-UST Java agent for `java.util.logging`.
4451 import java.io.IOException;
4452 import java.util.logging.Handler;
4453 import java.util.logging.Logger;
4454 import org.lttng.ust.agent.jul.LttngLogHandler;
4458 private static final int answer = 42;
4460 public static void main(String[] argv) throws Exception
4463 Logger logger = Logger.getLogger("jello");
4465 // Create an LTTng-UST log handler
4466 Handler lttngUstLogHandler = new LttngLogHandler();
4468 // Add the LTTng-UST log handler to our logger
4469 logger.addHandler(lttngUstLogHandler);
4472 logger.info("some info");
4473 logger.warning("some warning");
4475 logger.finer("finer information; the answer is " + answer);
4477 logger.severe("error!");
4479 // Not mandatory, but cleaner
4480 logger.removeHandler(lttngUstLogHandler);
4481 lttngUstLogHandler.close();
4490 $ javac -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-jul.jar Test.java
4493 <<creating-destroying-tracing-sessions,Create a tracing session>>,
4494 <<enabling-disabling-events,create an event rule>> matching the
4495 `jello` JUL logger, and <<basic-tracing-session-control,start tracing>>:
4500 $ lttng enable-event --jul jello
4504 Run the compiled class:
4508 $ java -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-jul.jar:. Test
4511 <<basic-tracing-session-control,Stop tracing>> and inspect the
4521 In the resulting trace, an <<event,event record>> generated by a Java
4522 application using `java.util.logging` is named `lttng_jul:event` and
4523 has the following fields:
4532 Name of the class in which the log statement was executed.
4535 Name of the method in which the log statement was executed.
4538 Logging time (timestamp in milliseconds).
4541 Log level integer value.
4544 ID of the thread in which the log statement was executed.
4546 Use the opt:lttng-enable-event(1):--loglevel or
4547 opt:lttng-enable-event(1):--loglevel-only option of the
4548 man:lttng-enable-event(1) command to target a range of JUL log levels
4549 or a specific JUL log level.
4554 ==== Use the LTTng-UST Java agent for Apache log4j
4556 To use the LTTng-UST Java agent in a Java application which uses
4557 Apache log4j{nbsp}1.2:
4559 . In the source code of the Java application, import the LTTng-UST log
4560 appender package for Apache log4j:
4565 import org.lttng.ust.agent.log4j.LttngLogAppender;
4569 . Create an LTTng-UST log4j log appender:
4574 Appender lttngUstLogAppender = new LttngLogAppender();
4578 . Add this appender to the log4j loggers which should emit LTTng events:
4583 Logger myLogger = Logger.getLogger("some-logger");
4585 myLogger.addAppender(lttngUstLogAppender);
4589 . Use Apache log4j log statements and configuration as usual. The
4590 loggers with an attached LTTng-UST log appender can emit LTTng events.
4592 . Before exiting the application, remove the LTTng-UST log appender from
4593 the loggers attached to it and call its `close()` method:
4598 myLogger.removeAppender(lttngUstLogAppender);
4599 lttngUstLogAppender.close();
4603 This isn't strictly necessary, but it is recommended for a clean
4604 disposal of the resources of the appender.
4606 . Include the common and log4j-specific JAR
4607 files of the LTTng-UST Java agent, path:{lttng-ust-agent-common.jar} and
4608 path:{lttng-ust-agent-log4j.jar}, in the
4609 https://docs.oracle.com/javase/tutorial/essential/environment/paths.html[class
4610 path] when you build the Java application.
4612 The JAR files are typically located in dir:{/usr/share/java}.
4614 IMPORTANT: The LTTng-UST Java agent must be
4615 <<installing-lttng,installed>> for the logging framework your
4618 .Use the LTTng-UST Java agent for Apache log4j.
4623 import org.apache.log4j.Appender;
4624 import org.apache.log4j.Logger;
4625 import org.lttng.ust.agent.log4j.LttngLogAppender;
4629 private static final int answer = 42;
4631 public static void main(String[] argv) throws Exception
4634 Logger logger = Logger.getLogger("jello");
4636 // Create an LTTng-UST log appender
4637 Appender lttngUstLogAppender = new LttngLogAppender();
4639 // Add the LTTng-UST log appender to our logger
4640 logger.addAppender(lttngUstLogAppender);
4643 logger.info("some info");
4644 logger.warn("some warning");
4646 logger.debug("debug information; the answer is " + answer);
4648 logger.fatal("error!");
4650 // Not mandatory, but cleaner
4651 logger.removeAppender(lttngUstLogAppender);
4652 lttngUstLogAppender.close();
4658 Build this example (`$LOG4JPATH` is the path to the Apache log4j JAR
4663 $ javac -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-log4j.jar:$LOG4JPATH Test.java
4666 <<creating-destroying-tracing-sessions,Create a tracing session>>,
4667 <<enabling-disabling-events,create an event rule>> matching the
4668 `jello` log4j logger, and <<basic-tracing-session-control,start tracing>>:
4673 $ lttng enable-event --log4j jello
4677 Run the compiled class:
4681 $ java -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-log4j.jar:$LOG4JPATH:. Test
4684 <<basic-tracing-session-control,Stop tracing>> and inspect the
4694 In the resulting trace, an <<event,event record>> generated by a Java
4695 application using log4j is named `lttng_log4j:event` and
4696 has the following fields:
4705 Name of the class in which the log statement was executed.
4708 Name of the method in which the log statement was executed.
4711 Name of the file in which the executed log statement is located.
4714 Line number at which the log statement was executed.
4720 Log level integer value.
4723 Name of the Java thread in which the log statement was executed.
4725 Use the opt:lttng-enable-event(1):--loglevel or
4726 opt:lttng-enable-event(1):--loglevel-only option of the
4727 man:lttng-enable-event(1) command to target a range of Apache log4j
4728 log levels or a specific log4j log level.
4732 [[java-application-context]]
4733 ==== Provide application-specific context fields in a Java application
4735 A Java application-specific context field is a piece of state provided
4736 by the application which <<adding-context,you can add>>, using the
4737 man:lttng-add-context(1) command, to each <<event,event record>>
4738 produced by the log statements of this application.
4740 For example, a given object might have a current request ID variable.
4741 You can create a context information retriever for this object and
4742 assign a name to this current request ID. You can then, using the
4743 man:lttng-add-context(1) command, add this context field by name to
4744 the JUL or log4j <<channel,channel>>.
4746 To provide application-specific context fields in a Java application:
4748 . In the source code of the Java application, import the LTTng-UST
4749 Java agent context classes and interfaces:
4754 import org.lttng.ust.agent.context.ContextInfoManager;
4755 import org.lttng.ust.agent.context.IContextInfoRetriever;
4759 . Create a context information retriever class, that is, a class which
4760 implements the `IContextInfoRetriever` interface:
4765 class MyContextInfoRetriever implements IContextInfoRetriever
4768 public Object retrieveContextInfo(String key)
4770 if (key.equals("intCtx")) {
4772 } else if (key.equals("strContext")) {
4773 return "context value!";
4782 This `retrieveContextInfo()` method is the only member of the
4783 `IContextInfoRetriever` interface. Its role is to return the current
4784 value of a state by name to create a context field. The names of the
4785 context fields and which state variables they return depends on your
4788 All primitive types and objects are supported as context fields.
4789 When `retrieveContextInfo()` returns an object, the context field
4790 serializer calls its `toString()` method to add a string field to
4791 event records. The method can also return `null`, which means that
4792 no context field is available for the required name.
4794 . Register an instance of your context information retriever class to
4795 the context information manager singleton:
4800 IContextInfoRetriever cir = new MyContextInfoRetriever();
4801 ContextInfoManager cim = ContextInfoManager.getInstance();
4802 cim.registerContextInfoRetriever("retrieverName", cir);
4806 . Before exiting the application, remove your context information
4807 retriever from the context information manager singleton:
4812 ContextInfoManager cim = ContextInfoManager.getInstance();
4813 cim.unregisterContextInfoRetriever("retrieverName");
4817 This isn't strictly necessary, but it is recommended for a clean
4818 disposal of some resources of the manager.
4820 . Build your Java application with LTTng-UST Java agent support as
4821 usual, following the procedure for either the <<jul,JUL>> or
4822 <<log4j,Apache log4j>> framework.
4825 .Provide application-specific context fields in a Java application.
4830 import java.util.logging.Handler;
4831 import java.util.logging.Logger;
4832 import org.lttng.ust.agent.jul.LttngLogHandler;
4833 import org.lttng.ust.agent.context.ContextInfoManager;
4834 import org.lttng.ust.agent.context.IContextInfoRetriever;
4838 // Our context information retriever class
4839 private static class MyContextInfoRetriever
4840 implements IContextInfoRetriever
4843 public Object retrieveContextInfo(String key) {
4844 if (key.equals("intCtx")) {
4846 } else if (key.equals("strContext")) {
4847 return "context value!";
4854 private static final int answer = 42;
4856 public static void main(String args[]) throws Exception
4858 // Get the context information manager instance
4859 ContextInfoManager cim = ContextInfoManager.getInstance();
4861 // Create and register our context information retriever
4862 IContextInfoRetriever cir = new MyContextInfoRetriever();
4863 cim.registerContextInfoRetriever("myRetriever", cir);
4866 Logger logger = Logger.getLogger("jello");
4868 // Create an LTTng-UST log handler
4869 Handler lttngUstLogHandler = new LttngLogHandler();
4871 // Add the LTTng-UST log handler to our logger
4872 logger.addHandler(lttngUstLogHandler);
4875 logger.info("some info");
4876 logger.warning("some warning");
4878 logger.finer("finer information; the answer is " + answer);
4880 logger.severe("error!");
4882 // Not mandatory, but cleaner
4883 logger.removeHandler(lttngUstLogHandler);
4884 lttngUstLogHandler.close();
4885 cim.unregisterContextInfoRetriever("myRetriever");
4894 $ javac -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-jul.jar Test.java
4897 <<creating-destroying-tracing-sessions,Create a tracing session>>
4898 and <<enabling-disabling-events,create an event rule>> matching the
4904 $ lttng enable-event --jul jello
4907 <<adding-context,Add the application-specific context fields>> to the
4912 $ lttng add-context --jul --type='$app.myRetriever:intCtx'
4913 $ lttng add-context --jul --type='$app.myRetriever:strContext'
4916 <<basic-tracing-session-control,Start tracing>>:
4923 Run the compiled class:
4927 $ java -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-jul.jar:. Test
4930 <<basic-tracing-session-control,Stop tracing>> and inspect the
4942 [[python-application]]
4943 === User space Python agent
4945 You can instrument a Python{nbsp}2 or Python{nbsp}3 application which
4947 https://docs.python.org/3/library/logging.html[`logging`] package.
4949 Each log statement emits an LTTng event once the
4950 application module imports the
4951 <<lttng-ust-agents,LTTng-UST Python agent>> package.
4954 .A Python application importing the LTTng-UST Python agent.
4955 image::python-app.png[]
4957 To use the LTTng-UST Python agent:
4959 . In the source code of the Python application, import the LTTng-UST
4969 The LTTng-UST Python agent automatically adds its logging handler to the
4970 root logger at import time.
4972 Any log statement that the application executes before this import does
4973 not emit an LTTng event.
4975 IMPORTANT: The LTTng-UST Python agent must be
4976 <<installing-lttng,installed>>.
4978 . Use log statements and logging configuration as usual.
4979 Since the LTTng-UST Python agent adds a handler to the _root_
4980 logger, you can trace any log statement from any logger.
4982 .Use the LTTng-UST Python agent.
4993 logging.basicConfig()
4994 logger = logging.getLogger('my-logger')
4997 logger.debug('debug message')
4998 logger.info('info message')
4999 logger.warn('warn message')
5000 logger.error('error message')
5001 logger.critical('critical message')
5005 if __name__ == '__main__':
5009 NOTE: `logging.basicConfig()`, which adds to the root logger a basic
5010 logging handler which prints to the standard error stream, isn't
5011 strictly required for LTTng-UST tracing to work, but in versions of
5012 Python preceding{nbsp}3.2, you could see a warning message which indicates
5013 that no handler exists for the logger `my-logger`.
5015 <<creating-destroying-tracing-sessions,Create a tracing session>>,
5016 <<enabling-disabling-events,create an event rule>> matching the
5017 `my-logger` Python logger, and <<basic-tracing-session-control,start
5023 $ lttng enable-event --python my-logger
5027 Run the Python script:
5034 <<basic-tracing-session-control,Stop tracing>> and inspect the recorded
5044 In the resulting trace, an <<event,event record>> generated by a Python
5045 application is named `lttng_python:event` and has the following fields:
5048 Logging time (string).
5057 Name of the function in which the log statement was executed.
5060 Line number at which the log statement was executed.
5063 Log level integer value.
5066 ID of the Python thread in which the log statement was executed.
5069 Name of the Python thread in which the log statement was executed.
5071 Use the opt:lttng-enable-event(1):--loglevel or
5072 opt:lttng-enable-event(1):--loglevel-only option of the
5073 man:lttng-enable-event(1) command to target a range of Python log levels
5074 or a specific Python log level.
5076 When an application imports the LTTng-UST Python agent, the agent tries
5077 to register to a <<lttng-sessiond,session daemon>>. Note that you must
5078 <<start-sessiond,start the session daemon>> _before_ you run the Python
5079 application. If a session daemon is found, the agent tries to register
5080 to it during five seconds, after which the application continues
5081 without LTTng tracing support. Override this timeout value with
5082 the env:LTTNG_UST_PYTHON_REGISTER_TIMEOUT environment variable
5085 If the session daemon stops while a Python application with an imported
5086 LTTng-UST Python agent runs, the agent retries to connect and to
5087 register to a session daemon every three seconds. Override this
5088 delay with the env:LTTNG_UST_PYTHON_REGISTER_RETRY_DELAY environment
5093 [[proc-lttng-logger-abi]]
5096 The `lttng-tracer` Linux kernel module, part of
5097 <<lttng-modules,LTTng-modules>>, creates the special LTTng logger files
5098 path:{/proc/lttng-logger} and path:{/dev/lttng-logger} (since
5099 LTTng{nbsp}2.11) when it's loaded. Any application can write text data
5100 to any of those files to emit an LTTng event.
5103 .An application writes to the LTTng logger file to emit an LTTng event.
5104 image::lttng-logger.png[]
5106 The LTTng logger is the quickest method--not the most efficient,
5107 however--to add instrumentation to an application. It is designed
5108 mostly to instrument shell scripts:
5112 $ echo "Some message, some $variable" > /dev/lttng-logger
5115 Any event that the LTTng logger emits is named `lttng_logger` and
5116 belongs to the Linux kernel <<domain,tracing domain>>. However, unlike
5117 other instrumentation points in the kernel tracing domain, **any Unix
5118 user** can <<enabling-disabling-events,create an event rule>> which
5119 matches its event name, not only the root user or users in the
5120 <<tracing-group,tracing group>>.
5122 To use the LTTng logger:
5124 * From any application, write text data to the path:{/dev/lttng-logger}
5127 The `msg` field of `lttng_logger` event records contains the
5130 NOTE: The maximum message length of an LTTng logger event is
5131 1024{nbsp}bytes. Writing more than this makes the LTTng logger emit more
5132 than one event to contain the remaining data.
5134 You shouldn't use the LTTng logger to trace a user application which
5135 can be instrumented in a more efficient way, namely:
5137 * <<c-application,C and $$C++$$ applications>>.
5138 * <<java-application,Java applications>>.
5139 * <<python-application,Python applications>>.
5141 .Use the LTTng logger.
5146 echo 'Hello, World!' > /dev/lttng-logger
5148 df --human-readable --print-type / > /dev/lttng-logger
5151 <<creating-destroying-tracing-sessions,Create a tracing session>>,
5152 <<enabling-disabling-events,create an event rule>> matching the
5153 `lttng_logger` Linux kernel tracepoint, and
5154 <<basic-tracing-session-control,start tracing>>:
5159 $ lttng enable-event --kernel lttng_logger
5163 Run the Bash script:
5170 <<basic-tracing-session-control,Stop tracing>> and inspect the recorded
5181 [[instrumenting-linux-kernel]]
5182 === LTTng kernel tracepoints
5184 NOTE: This section shows how to _add_ instrumentation points to the
5185 Linux kernel. The subsystems of the kernel are already thoroughly
5186 instrumented at strategic places for LTTng when you
5187 <<installing-lttng,install>> the <<lttng-modules,LTTng-modules>>
5191 There are two methods to instrument the Linux kernel:
5193 . <<linux-add-lttng-layer,Add an LTTng layer>> over an existing ftrace
5194 tracepoint which uses the `TRACE_EVENT()` API.
5196 Choose this if you want to instrumentation a Linux kernel tree with an
5197 instrumentation point compatible with ftrace, perf, and SystemTap.
5199 . Use an <<linux-lttng-tracepoint-event,LTTng-only approach>> to
5200 instrument an out-of-tree kernel module.
5202 Choose this if you don't need ftrace, perf, or SystemTap support.
5206 [[linux-add-lttng-layer]]
5207 ==== [[instrumenting-linux-kernel-itself]][[mainline-trace-event]][[lttng-adaptation-layer]]Add an LTTng layer to an existing ftrace tracepoint
5209 This section shows how to add an LTTng layer to existing ftrace
5210 instrumentation using the `TRACE_EVENT()` API.
5212 This section doesn't document the `TRACE_EVENT()` macro. Read the
5213 following articles to learn more about this API:
5215 * http://lwn.net/Articles/379903/[Using the TRACE_EVENT() macro (Part{nbsp}1)]
5216 * http://lwn.net/Articles/381064/[Using the TRACE_EVENT() macro (Part{nbsp}2)]
5217 * http://lwn.net/Articles/383362/[Using the TRACE_EVENT() macro (Part{nbsp}3)]
5219 The following procedure assumes that your ftrace tracepoints are
5220 correctly defined in their own header and that they are created in
5221 one source file using the `CREATE_TRACE_POINTS` definition.
5223 To add an LTTng layer over an existing ftrace tracepoint:
5225 . Make sure the following kernel configuration options are
5231 * `CONFIG_HIGH_RES_TIMERS`
5232 * `CONFIG_TRACEPOINTS`
5235 . Build the Linux source tree with your custom ftrace tracepoints.
5236 . Boot the resulting Linux image on your target system.
5238 Confirm that the tracepoints exist by looking for their names in the
5239 dir:{/sys/kernel/debug/tracing/events/subsys} directory, where `subsys`
5240 is your subsystem name.
5242 . Get a copy of the latest LTTng-modules{nbsp}{revision}:
5247 $ cd $(mktemp -d) &&
5248 wget http://lttng.org/files/lttng-modules/lttng-modules-latest-2.12.tar.bz2 &&
5249 tar -xf lttng-modules-latest-2.12.tar.bz2 &&
5250 cd lttng-modules-2.12.*
5254 . In dir:{instrumentation/events/lttng-module}, relative to the root
5255 of the LTTng-modules source tree, create a header file named
5256 +__subsys__.h+ for your custom subsystem +__subsys__+ and write your
5257 LTTng-modules tracepoint definitions using the LTTng-modules
5260 Start with this template:
5264 .path:{instrumentation/events/lttng-module/my_subsys.h}
5267 #define TRACE_SYSTEM my_subsys
5269 #if !defined(_LTTNG_MY_SUBSYS_H) || defined(TRACE_HEADER_MULTI_READ)
5270 #define _LTTNG_MY_SUBSYS_H
5272 #include "../../../probes/lttng-tracepoint-event.h"
5273 #include <linux/tracepoint.h>
5275 LTTNG_TRACEPOINT_EVENT(
5277 * Format is identical to the TRACE_EVENT() version for the three
5278 * following macro parameters:
5281 TP_PROTO(int my_int, const char *my_string),
5282 TP_ARGS(my_int, my_string),
5284 /* LTTng-modules specific macros */
5286 ctf_integer(int, my_int_field, my_int)
5287 ctf_string(my_bar_field, my_bar)
5291 #endif /* !defined(_LTTNG_MY_SUBSYS_H) || defined(TRACE_HEADER_MULTI_READ) */
5293 #include "../../../probes/define_trace.h"
5297 The entries in the `TP_FIELDS()` section are the list of fields for the
5298 LTTng tracepoint. This is similar to the `TP_STRUCT__entry()` part of
5299 the `TRACE_EVENT()` ftrace macro.
5301 See <<lttng-modules-tp-fields,Tracepoint fields macros>> for a
5302 complete description of the available `ctf_*()` macros.
5304 . Create the kernel module C{nbsp}source file of the LTTng-modules
5305 probe, +probes/lttng-probe-__subsys__.c+, where +__subsys__+ is your
5310 .path:{probes/lttng-probe-my-subsys.c}
5312 #include <linux/module.h>
5313 #include "../lttng-tracer.h"
5316 * Build-time verification of mismatch between mainline
5317 * TRACE_EVENT() arguments and the LTTng-modules adaptation
5318 * layer LTTNG_TRACEPOINT_EVENT() arguments.
5320 #include <trace/events/my_subsys.h>
5322 /* Create LTTng tracepoint probes */
5323 #define LTTNG_PACKAGE_BUILD
5324 #define CREATE_TRACE_POINTS
5325 #define TRACE_INCLUDE_PATH ../instrumentation/events/lttng-module
5327 #include "../instrumentation/events/lttng-module/my_subsys.h"
5329 MODULE_LICENSE("GPL and additional rights");
5330 MODULE_AUTHOR("Your name <your-email>");
5331 MODULE_DESCRIPTION("LTTng my_subsys probes");
5332 MODULE_VERSION(__stringify(LTTNG_MODULES_MAJOR_VERSION) "."
5333 __stringify(LTTNG_MODULES_MINOR_VERSION) "."
5334 __stringify(LTTNG_MODULES_PATCHLEVEL_VERSION)
5335 LTTNG_MODULES_EXTRAVERSION);
5339 . Edit path:{probes/KBuild} and add your new kernel module object
5340 next to the existing ones:
5344 .path:{probes/KBuild}
5348 obj-m += lttng-probe-module.o
5349 obj-m += lttng-probe-power.o
5351 obj-m += lttng-probe-my-subsys.o
5357 . Build and install the LTTng kernel modules:
5362 $ make KERNELDIR=/path/to/linux
5363 # make modules_install && depmod -a
5367 Replace `/path/to/linux` with the path to the Linux source tree where
5368 you defined and used tracepoints with the `TRACE_EVENT()` ftrace macro.
5370 Note that you can also use the
5371 <<lttng-tracepoint-event-code,`LTTNG_TRACEPOINT_EVENT_CODE()` macro>>
5372 instead of `LTTNG_TRACEPOINT_EVENT()` to use custom local variables and
5373 C code that need to be executed before the event fields are recorded.
5375 The best way to learn how to use the previous LTTng-modules macros is to
5376 inspect the existing LTTng-modules tracepoint definitions in the
5377 dir:{instrumentation/events/lttng-module} header files. Compare them
5378 with the Linux kernel mainline versions in the
5379 dir:{include/trace/events} directory of the Linux source tree.
5383 [[lttng-tracepoint-event-code]]
5384 ===== Use custom C code to access the data for tracepoint fields
5386 Although we recommended to always use the
5387 <<lttng-adaptation-layer,`LTTNG_TRACEPOINT_EVENT()`>> macro to describe
5388 the arguments and fields of an LTTng-modules tracepoint when possible,
5389 sometimes you need a more complex process to access the data that the
5390 tracer records as event record fields. In other words, you need local
5391 variables and multiple C{nbsp}statements instead of simple
5392 argument-based expressions that you pass to the
5393 <<lttng-modules-tp-fields,`ctf_*()` macros of `TP_FIELDS()`>>.
5395 Use the `LTTNG_TRACEPOINT_EVENT_CODE()` macro instead of
5396 `LTTNG_TRACEPOINT_EVENT()` to declare custom local variables and define
5397 a block of C{nbsp}code to be executed before LTTng records the fields.
5398 The structure of this macro is:
5401 .`LTTNG_TRACEPOINT_EVENT_CODE()` macro syntax.
5403 LTTNG_TRACEPOINT_EVENT_CODE(
5405 * Format identical to the LTTNG_TRACEPOINT_EVENT()
5406 * version for the following three macro parameters:
5409 TP_PROTO(int my_int, const char *my_string),
5410 TP_ARGS(my_int, my_string),
5412 /* Declarations of custom local variables */
5415 unsigned long b = 0;
5416 const char *name = "(undefined)";
5417 struct my_struct *my_struct;
5421 * Custom code which uses both tracepoint arguments
5422 * (in TP_ARGS()) and local variables (in TP_locvar()).
5424 * Local variables are actually members of a structure pointed
5425 * to by the special variable tp_locvar.
5429 tp_locvar->a = my_int + 17;
5430 tp_locvar->my_struct = get_my_struct_at(tp_locvar->a);
5431 tp_locvar->b = my_struct_compute_b(tp_locvar->my_struct);
5432 tp_locvar->name = my_struct_get_name(tp_locvar->my_struct);
5433 put_my_struct(tp_locvar->my_struct);
5442 * Format identical to the LTTNG_TRACEPOINT_EVENT()
5443 * version for this, except that tp_locvar members can be
5444 * used in the argument expression parameters of
5445 * the ctf_*() macros.
5448 ctf_integer(unsigned long, my_struct_b, tp_locvar->b)
5449 ctf_integer(int, my_struct_a, tp_locvar->a)
5450 ctf_string(my_string_field, my_string)
5451 ctf_string(my_struct_name, tp_locvar->name)
5456 IMPORTANT: The C code defined in `TP_code()` must not have any side
5457 effects when executed. In particular, the code must not allocate
5458 memory or get resources without deallocating this memory or putting
5459 those resources afterwards.
5462 [[instrumenting-linux-kernel-tracing]]
5463 ==== Load and unload a custom probe kernel module
5465 You must load a <<lttng-adaptation-layer,created LTTng-modules probe
5466 kernel module>> in the kernel before it can emit LTTng events.
5468 To load the default probe kernel modules and a custom probe kernel
5471 * Use the opt:lttng-sessiond(8):--extra-kmod-probes option to give extra
5472 probe modules to load when starting a root <<lttng-sessiond,session
5476 .Load the `my_subsys`, `usb`, and the default probe modules.
5480 # lttng-sessiond --extra-kmod-probes=my_subsys,usb
5485 You only need to pass the subsystem name, not the whole kernel module
5488 To load _only_ a given custom probe kernel module:
5490 * Use the opt:lttng-sessiond(8):--kmod-probes option to give the probe
5491 modules to load when starting a root session daemon:
5494 .Load only the `my_subsys` and `usb` probe modules.
5498 # lttng-sessiond --kmod-probes=my_subsys,usb
5503 To confirm that a probe module is loaded:
5510 $ lsmod | grep lttng_probe_usb
5514 To unload the loaded probe modules:
5516 * Kill the session daemon with `SIGTERM`:
5521 # pkill lttng-sessiond
5525 You can also use the man:modprobe(8) `--remove` option if the session
5526 daemon terminates abnormally.
5529 [[controlling-tracing]]
5532 Once an application or a Linux kernel is
5533 <<instrumenting,instrumented>> for LTTng tracing,
5536 This section is divided in topics on how to use the various
5537 <<plumbing,components of LTTng>>, in particular the <<lttng-cli,cmd:lttng
5538 command-line tool>>, to _control_ the LTTng daemons and tracers.
5540 NOTE: In the following subsections, we refer to an man:lttng(1) command
5541 using its man page name. For example, instead of _Run the `create`
5542 command to..._, we use _Run the man:lttng-create(1) command to..._.
5546 === Start a session daemon
5548 In some situations, you need to run a <<lttng-sessiond,session daemon>>
5549 (man:lttng-sessiond(8)) _before_ you can use the man:lttng(1)
5552 You will see the following error when you run a command while no session
5556 Error: No session daemon is available
5559 The only command that automatically runs a session daemon is
5560 man:lttng-create(1), which you use to
5561 <<creating-destroying-tracing-sessions,create a tracing session>>. While
5562 this is most of the time the first operation that you do, sometimes it's
5563 not. Some examples are:
5565 * <<list-instrumentation-points,List the available instrumentation points>>.
5566 * <<saving-loading-tracing-session,Load a tracing session configuration>>.
5568 [[tracing-group]] Each Unix user must have its own running session
5569 daemon to trace user applications. The session daemon that the root user
5570 starts is the only one allowed to control the LTTng kernel tracer. Users
5571 that are part of the _tracing group_ can control the root session
5572 daemon. The default tracing group name is `tracing`; set it to something
5573 else with the opt:lttng-sessiond(8):--group option when you start the
5574 root session daemon.
5576 To start a user session daemon:
5578 * Run man:lttng-sessiond(8):
5583 $ lttng-sessiond --daemonize
5587 To start the root session daemon:
5589 * Run man:lttng-sessiond(8) as the root user:
5594 # lttng-sessiond --daemonize
5598 In both cases, remove the opt:lttng-sessiond(8):--daemonize option to
5599 start the session daemon in foreground.
5601 To stop a session daemon, use man:kill(1) on its process ID (standard
5604 Note that some Linux distributions could manage the LTTng session daemon
5605 as a service. In this case, you should use the service manager to
5606 start, restart, and stop session daemons.
5609 [[creating-destroying-tracing-sessions]]
5610 === Create and destroy a tracing session
5612 Almost all the LTTng control operations happen in the scope of
5613 a <<tracing-session,tracing session>>, which is the dialogue between the
5614 <<lttng-sessiond,session daemon>> and you.
5616 To create a tracing session with a generated name:
5618 * Use the man:lttng-create(1) command:
5627 The name of the created tracing session is `auto` followed by the
5630 To create a tracing session with a specific name:
5632 * Use the optional argument of the man:lttng-create(1) command:
5637 $ lttng create my-session
5641 Replace `my-session` with the specific tracing session name.
5643 LTTng appends the creation date to the name of the created tracing
5646 LTTng writes the traces of a tracing session in
5647 +$LTTNG_HOME/lttng-trace/__name__+ by default, where +__name__+ is the
5648 name of the tracing session. Note that the env:LTTNG_HOME environment
5649 variable defaults to `$HOME` if not set.
5651 To output LTTng traces to a non-default location:
5653 * Use the opt:lttng-create(1):--output option of the man:lttng-create(1) command:
5658 $ lttng create my-session --output=/tmp/some-directory
5662 You may create as many tracing sessions as you wish.
5664 To list all the existing tracing sessions for your Unix user:
5666 * Use the man:lttng-list(1) command:
5675 [[cur-tracing-session]]When you create a tracing session, it is set as
5676 the _current tracing session_. The following man:lttng(1) commands
5677 operate on the current tracing session when you don't specify one:
5679 [role="list-3-cols"]
5680 * man:lttng-add-context(1)
5681 * man:lttng-clear(1)
5682 * man:lttng-destroy(1)
5683 * man:lttng-disable-channel(1)
5684 * man:lttng-disable-event(1)
5685 * man:lttng-disable-rotation(1)
5686 * man:lttng-enable-channel(1)
5687 * man:lttng-enable-event(1)
5688 * man:lttng-enable-rotation(1)
5690 * man:lttng-regenerate(1)
5691 * man:lttng-rotate(1)
5693 * man:lttng-snapshot(1)
5694 * man:lttng-start(1)
5695 * man:lttng-status(1)
5697 * man:lttng-track(1)
5698 * man:lttng-untrack(1)
5701 To change the current tracing session:
5703 * Use the man:lttng-set-session(1) command:
5708 $ lttng set-session new-session
5712 Replace `new-session` by the name of the new current tracing session.
5714 When you're done tracing in a given tracing session, destroy it. This
5715 operation frees the resources taken by the tracing session to destroy;
5716 it doesn't destroy the trace data that LTTng wrote for this tracing
5717 session (see <<clear,Clear a tracing session>> for one way to do this).
5719 To destroy the current tracing session:
5721 * Use the man:lttng-destroy(1) command:
5730 The man:lttng-destroy(1) command also runs the man:lttng-stop(1)
5731 command implicitly (see <<basic-tracing-session-control,Start and stop a
5732 tracing session>>). You need to stop tracing to make LTTng flush the
5733 remaining trace data and make the trace readable.
5736 [[list-instrumentation-points]]
5737 === List the available instrumentation points
5739 The <<lttng-sessiond,session daemon>> can query the running instrumented
5740 user applications and the Linux kernel to get a list of available
5741 instrumentation points. For the Linux kernel <<domain,tracing domain>>,
5742 they are tracepoints and system calls. For the user space tracing
5743 domain, they are tracepoints. For the other tracing domains, they are
5746 To list the available instrumentation points:
5748 * Use the man:lttng-list(1) command with the option of the requested
5749 tracing domain amongst:
5752 opt:lttng-list(1):--kernel::
5753 Linux kernel tracepoints (your Unix user must be a root user, or it
5754 must be a member of the <<tracing-group,tracing group>>).
5756 opt:lttng-list(1):--kernel with opt:lttng-list(1):--syscall::
5757 Linux kernel system calls (your Unix user must be a root user, or it
5758 must be a member of the tracing group).
5760 opt:lttng-list(1):--userspace::
5761 User space tracepoints.
5763 opt:lttng-list(1):--jul::
5764 `java.util.logging` loggers.
5766 opt:lttng-list(1):--log4j::
5767 Apache log4j loggers.
5769 opt:lttng-list(1):--python::
5773 .List the available user space tracepoints.
5777 $ lttng list --userspace
5781 .List the available Linux kernel system call tracepoints.
5785 $ lttng list --kernel --syscall
5790 [[enabling-disabling-events]]
5791 === Create and enable an event rule
5793 Once you <<creating-destroying-tracing-sessions,create a tracing
5794 session>>, you can create <<event,event rules>> with the
5795 man:lttng-enable-event(1) command.
5797 You specify each condition with a command-line option. The available
5798 condition arguments are shown in the following table.
5800 [role="growable",cols="asciidoc,asciidoc,default"]
5801 .Condition command-line arguments for the man:lttng-enable-event(1) command.
5803 |Argument |Description |Applicable tracing domains
5809 . +--probe=__ADDR__+
5810 . +--function=__ADDR__+
5811 . +--userspace-probe=__PATH__:__SYMBOL__+
5812 . +--userspace-probe=sdt:__PATH__:__PROVIDER__:__NAME__+
5815 Instead of using the default _tracepoint_ instrumentation type, use:
5817 . A Linux system call (entry and exit).
5818 . A Linux https://lwn.net/Articles/132196/[kprobe] (symbol or address).
5819 . The entry and return points of a Linux function (symbol or address).
5820 . The entry point of a user application or library function (path to
5821 application/library and symbol).
5822 . A https://www.sourceware.org/systemtap/wiki/AddingUserSpaceProbingToApps[SystemTap
5823 Statically Defined Tracing] (USDT) probe (path to application/library,
5824 provider and probe names).
5828 |First positional argument.
5831 Tracepoint or system call name.
5833 With the opt:lttng-enable-event(1):--probe,
5834 opt:lttng-enable-event(1):--function, and
5835 opt:lttng-enable-event(1):--userspace-probe options, this is a custom
5836 name given to the event rule. With the JUL, log4j, and Python domains,
5837 this is a logger name.
5839 With a tracepoint, logger, or system call name, use the special
5840 `*` globbing character to match anything (for example, `sched_*`,
5848 . +--loglevel=__LEVEL__+
5849 . +--loglevel-only=__LEVEL__+
5852 . Match only tracepoints or log statements with a logging level at
5853 least as severe as +__LEVEL__+.
5854 . Match only tracepoints or log statements with a logging level
5855 equal to +__LEVEL__+.
5857 See man:lttng-enable-event(1) for the list of available logging level
5860 |User space, JUL, log4j, and Python.
5862 |+--exclude=__EXCLUSIONS__+
5865 When you use a `*` character at the end of the tracepoint or logger
5866 name (first positional argument), exclude the specific names in the
5867 comma-delimited list +__EXCLUSIONS__+.
5870 User space, JUL, log4j, and Python.
5872 |+--filter=__EXPR__+
5875 Match only events which satisfy the expression +__EXPR__+.
5877 See man:lttng-enable-event(1) to learn more about the syntax of a
5884 You attach an event rule to a <<channel,channel>> on creation. If you do
5885 not specify the channel with the opt:lttng-enable-event(1):--channel
5886 option, and if the event rule to create is the first in its
5887 <<domain,tracing domain>> for a given tracing session, then LTTng
5888 creates a _default channel_ for you. This default channel is reused in
5889 subsequent invocations of the man:lttng-enable-event(1) command for the
5890 same tracing domain.
5892 An event rule is always enabled at creation time.
5894 The following examples show how to combine the previous
5895 command-line options to create simple to more complex event rules.
5897 .Create an event rule targetting a Linux kernel tracepoint (default channel).
5901 $ lttng enable-event --kernel sched_switch
5905 .Create an event rule matching four Linux kernel system calls (default channel).
5909 $ lttng enable-event --kernel --syscall open,write,read,close
5913 .Create event rules matching tracepoints with filter expressions (default channel).
5917 $ lttng enable-event --kernel sched_switch --filter='prev_comm == "bash"'
5922 $ lttng enable-event --kernel --all \
5923 --filter='$ctx.tid == 1988 || $ctx.tid == 1534'
5928 $ lttng enable-event --jul my_logger \
5929 --filter='$app.retriever:cur_msg_id > 3'
5932 IMPORTANT: Make sure to always quote the filter string when you
5933 use man:lttng(1) from a shell.
5935 See also <<pid-tracking,Track process attributes>> which offers another,
5936 more efficient filtering mechanism for process ID, user ID, and group
5940 .Create an event rule matching any user space tracepoint of a given tracepoint provider with a log level range (default channel).
5944 $ lttng enable-event --userspace my_app:'*' --loglevel=TRACE_INFO
5947 IMPORTANT: Make sure to always quote the wildcard character when you
5948 use man:lttng(1) from a shell.
5951 .Create an event rule matching multiple Python loggers with a wildcard and with exclusions (default channel).
5955 $ lttng enable-event --python my-app.'*' \
5956 --exclude='my-app.module,my-app.hello'
5960 .Create an event rule matching any Apache log4j logger with a specific log level (default channel).
5964 $ lttng enable-event --log4j --all --loglevel-only=LOG4J_WARN
5968 .Create an event rule attached to a specific channel matching a specific user space tracepoint provider and tracepoint.
5972 $ lttng enable-event --userspace my_app:my_tracepoint --channel=my-channel
5976 .Create an event rule matching the `malloc` function entry in path:{/usr/lib/libc.so.6}:
5980 $ lttng enable-event --kernel --userspace-probe=/usr/lib/libc.so.6:malloc \
5985 .Create an event rule matching the `server`/`accept_request` https://www.sourceware.org/systemtap/wiki/AddingUserSpaceProbingToApps[USDT probe] in path:{/usr/bin/serv}:
5989 $ lttng enable-event --kernel --userspace-probe=sdt:serv:server:accept_request \
5990 server_accept_request
5994 The event rules of a given channel form a whitelist: as soon as an
5995 emitted event passes one of them, LTTng can record the event. For
5996 example, an event named `my_app:my_tracepoint` emitted from a user space
5997 tracepoint with a `TRACE_ERROR` log level passes both of the following
6002 $ lttng enable-event --userspace my_app:my_tracepoint
6003 $ lttng enable-event --userspace my_app:my_tracepoint \
6004 --loglevel=TRACE_INFO
6007 The second event rule is redundant: the first one includes
6011 [[disable-event-rule]]
6012 === Disable an event rule
6014 To disable an event rule that you <<enabling-disabling-events,created>>
6015 previously, use the man:lttng-disable-event(1) command. This command
6016 disables _all_ the event rules (of a given tracing domain and channel)
6017 which match an instrumentation point. The other conditions aren't
6018 supported as of LTTng{nbsp}{revision}.
6020 The LTTng tracer doesn't record an emitted event which passes
6021 a _disabled_ event rule.
6023 .Disable an event rule matching a Python logger (default channel).
6027 $ lttng disable-event --python my-logger
6031 .Disable an event rule matching all `java.util.logging` loggers (default channel).
6035 $ lttng disable-event --jul '*'
6039 .Disable _all_ the event rules of the default channel.
6041 The opt:lttng-disable-event(1):--all-events option isn't, like the
6042 opt:lttng-enable-event(1):--all option of man:lttng-enable-event(1), the
6043 equivalent of the event name `*` (wildcard): it disables _all_ the event
6044 rules of a given channel.
6048 $ lttng disable-event --jul --all-events
6052 NOTE: You can't delete an event rule once you create it.
6056 === Get the status of a tracing session
6058 To get the status of the <<cur-tracing-session,current tracing
6059 session>>, that is, its parameters, its channels, event rules, and their
6062 * Use the man:lttng-status(1) command:
6071 To get the status of any tracing session:
6073 * Use the man:lttng-list(1) command with the name of the tracing
6079 $ lttng list my-session
6083 Replace `my-session` with the desired tracing session name.
6086 [[basic-tracing-session-control]]
6087 === Start and stop a tracing session
6089 Once you <<creating-destroying-tracing-sessions,create a tracing
6091 <<enabling-disabling-events,create one or more event rules>>,
6092 you can start and stop the tracers for this tracing session.
6094 To start tracing in the <<cur-tracing-session,current tracing session>>:
6096 * Use the man:lttng-start(1) command:
6105 LTTng is very flexible: you can launch user applications before
6106 or after the you start the tracers. The tracers only record the events
6107 if they pass enabled event rules and if they occur while the tracers are
6110 To stop tracing in the current tracing session:
6112 * Use the man:lttng-stop(1) command:
6121 If there were <<channel-overwrite-mode-vs-discard-mode,lost event
6122 records>> or lost sub-buffers since the last time you ran
6123 man:lttng-start(1), warnings are printed when you run the
6124 man:lttng-stop(1) command.
6126 IMPORTANT: You need to stop tracing to make LTTng flush the remaining
6127 trace data and make the trace readable. Note that the
6128 man:lttng-destroy(1) command (see
6129 <<creating-destroying-tracing-sessions,Create and destroy a tracing
6130 session>>) also runs the man:lttng-stop(1) command implicitly.
6134 === Clear a tracing session
6136 You might need to remove all the current tracing data of one or more
6137 <<tracing-session,tracing sessions>> between multiple attempts to
6138 reproduce a problem without interrupting the LTTng tracing activity.
6140 To clear the tracing data of the
6141 <<cur-tracing-session,current tracing session>>:
6143 * Use the man:lttng-clear(1) command:
6152 To clear the tracing data of all the tracing sessions:
6154 * Use the `lttng clear` command with the opt:lttng-clear(1):--all
6165 [[enabling-disabling-channels]]
6166 === Create a channel
6168 Once you create a tracing session, you can create a <<channel,channel>>
6169 with the man:lttng-enable-channel(1) command.
6171 Note that LTTng automatically creates a default channel when, for a
6172 given <<domain,tracing domain>>, no channels exist and you
6173 <<enabling-disabling-events,create>> the first event rule. This default
6174 channel is named `channel0` and its attributes are set to reasonable
6175 values. Therefore, you only need to create a channel when you need
6176 non-default attributes.
6178 You specify each non-default channel attribute with a command-line
6179 option when you use the man:lttng-enable-channel(1) command. The
6180 available command-line options are:
6182 [role="growable",cols="asciidoc,asciidoc"]
6183 .Command-line options for the man:lttng-enable-channel(1) command.
6185 |Option |Description
6191 <<channel-overwrite-mode-vs-discard-mode,event record loss mode>> instead
6192 of the default _discard_ mode.
6194 |`--buffers-pid` (user space tracing domain only)
6197 Use the per-process <<channel-buffering-schemes,buffering scheme>>
6198 instead of the default per-user buffering scheme.
6200 |+--subbuf-size=__SIZE__+
6203 Allocate sub-buffers of +__SIZE__+ bytes (power of two), for each CPU,
6204 either for each Unix user (default), or for each instrumented process.
6206 See <<channel-subbuf-size-vs-subbuf-count,Sub-buffer count and size>>.
6208 |+--num-subbuf=__COUNT__+
6211 Allocate +__COUNT__+ sub-buffers (power of two), for each CPU, either
6212 for each Unix user (default), or for each instrumented process.
6214 See <<channel-subbuf-size-vs-subbuf-count,Sub-buffer count and size>>.
6216 |+--tracefile-size=__SIZE__+
6219 Set the maximum size of each trace file that this channel writes within
6220 a stream to +__SIZE__+ bytes instead of no maximum.
6222 See <<tracefile-rotation,Trace file count and size>>.
6224 |+--tracefile-count=__COUNT__+
6227 Limit the number of trace files that this channel creates to
6228 +__COUNT__+ channels instead of no limit.
6230 See <<tracefile-rotation,Trace file count and size>>.
6232 |+--switch-timer=__PERIODUS__+
6235 Set the <<channel-switch-timer,switch timer period>>
6236 to +__PERIODUS__+{nbsp}µs.
6238 |+--read-timer=__PERIODUS__+
6241 Set the <<channel-read-timer,read timer period>>
6242 to +__PERIODUS__+{nbsp}µs.
6244 |[[opt-blocking-timeout]]+--blocking-timeout=__TIMEOUTUS__+
6247 Set the timeout of user space applications which load LTTng-UST
6248 in blocking mode to +__TIMEOUTUS__+:
6251 Never block (non-blocking mode).
6254 Block forever until space is available in a sub-buffer to record
6257 __n__, a positive value::
6258 Wait for at most __n__ µs when trying to write into a sub-buffer.
6260 Note that, for this option to have any effect on an instrumented
6261 user space application, you need to run the application with a set
6262 env:LTTNG_UST_ALLOW_BLOCKING environment variable.
6264 |+--output=__TYPE__+ (Linux kernel tracing domain only)
6267 Set the output type of the channel to +__TYPE__+, either `mmap` or
6272 You can only create a channel in the Linux kernel and user space
6273 <<domain,tracing domains>>: other tracing domains have their own channel
6274 created on the fly when <<enabling-disabling-events,creating event
6279 Because of a current LTTng limitation, you must create all channels
6280 _before_ you <<basic-tracing-session-control,start tracing>> in a given
6281 tracing session, that is, before the first time you run
6284 Since LTTng automatically creates a default channel when you use the
6285 man:lttng-enable-event(1) command with a specific tracing domain, you
6286 can't, for example, create a Linux kernel event rule, start tracing,
6287 and then create a user space event rule, because no user space channel
6288 exists yet and it's too late to create one.
6290 For this reason, make sure to configure your channels properly
6291 before starting the tracers for the first time!
6294 The following examples show how to combine the previous
6295 command-line options to create simple to more complex channels.
6297 .Create a Linux kernel channel with default attributes.
6301 $ lttng enable-channel --kernel my-channel
6305 .Create a user space channel with four sub-buffers or 1{nbsp}MiB each, per CPU, per instrumented process.
6309 $ lttng enable-channel --userspace --num-subbuf=4 --subbuf-size=1M \
6310 --buffers-pid my-channel
6314 .[[blocking-timeout-example]]Create a default user space channel with an infinite blocking timeout.
6316 <<creating-destroying-tracing-sessions,Create a tracing-session>>,
6317 create the channel, <<enabling-disabling-events,create an event rule>>,
6318 and <<basic-tracing-session-control,start tracing>>:
6323 $ lttng enable-channel --userspace --blocking-timeout=inf blocking-channel
6324 $ lttng enable-event --userspace --channel=blocking-channel --all
6328 Run an application instrumented with LTTng-UST and allow it to block:
6332 $ LTTNG_UST_ALLOW_BLOCKING=1 my-app
6336 .Create a Linux kernel channel which rotates eight trace files of 4{nbsp}MiB each for each stream
6340 $ lttng enable-channel --kernel --tracefile-count=8 \
6341 --tracefile-size=4194304 my-channel
6345 .Create a user space channel in overwrite (or _flight recorder_) mode.
6349 $ lttng enable-channel --userspace --overwrite my-channel
6353 <<enabling-disabling-events,Create>> the same event rule in
6354 two different channels:
6358 $ lttng enable-event --userspace --channel=my-channel app:tp
6359 $ lttng enable-event --userspace --channel=other-channel app:tp
6362 If both channels are enabled, when a tracepoint named `app:tp` is
6363 reached, LTTng records two events, one for each channel.
6367 === Disable a channel
6369 To disable a specific channel that you <<enabling-disabling-channels,created>>
6370 previously, use the man:lttng-disable-channel(1) command.
6372 .Disable a specific Linux kernel channel.
6376 $ lttng disable-channel --kernel my-channel
6380 The state of a channel precedes the individual states of event rules
6381 attached to it: event rules which belong to a disabled channel, even if
6382 they are enabled, are also considered disabled.
6386 === Add context fields to a channel
6388 Event record fields in trace files provide important information about
6389 events that occured previously, but sometimes some external context may
6390 help you solve a problem faster.
6392 Examples of context fields are:
6394 * The **process ID**, **thread ID**, **process name**, and
6395 **process priority** of the thread in which the event occurs.
6396 * The **hostname** of the system on which the event occurs.
6397 * The Linux kernel and user call stacks (since
6399 * The current values of many possible **performance counters** using
6401 ** CPU cycles, stalled cycles, idle cycles, and the other cycle types.
6403 ** Branch instructions, misses, and loads.
6405 * Any context defined at the application level (supported for the
6406 JUL and log4j <<domain,tracing domains>>).
6408 To get the full list of available context fields, see
6409 `lttng add-context --list`. Some context fields are reserved for a
6410 specific <<domain,tracing domain>> (Linux kernel or user space).
6412 You add context fields to <<channel,channels>>. All the events
6413 that a channel with added context fields records contain those fields.
6415 To add context fields to one or all the channels of a given tracing
6418 * Use the man:lttng-add-context(1) command.
6420 .Add context fields to all the channels of the current tracing session.
6422 The following command line adds the virtual process identifier and
6423 the per-thread CPU cycles count fields to all the user space channels
6425 <<cur-tracing-session,current tracing session>>.
6429 $ lttng add-context --userspace --type=vpid --type=perf:thread:cpu-cycles
6433 .Add performance counter context fields by raw ID
6435 See man:lttng-add-context(1) for the exact format of the context field
6436 type, which is partly compatible with the format used in
6441 $ lttng add-context --userspace --type=perf:thread:raw:r0110:test
6442 $ lttng add-context --kernel --type=perf:cpu:raw:r0013c:x86unhalted
6446 .Add context fields to a specific channel.
6448 The following command line adds the thread identifier and user call
6449 stack context fields to the Linux kernel channel named `my-channel` in
6450 the current tracing session.
6454 $ lttng add-context --kernel --channel=my-channel \
6455 --type=tid --type=callstack-user
6459 .Add an application-specific context field to a specific channel.
6461 The following command line adds the `cur_msg_id` context field of the
6462 `retriever` context retriever for all the instrumented
6463 <<java-application,Java applications>> recording <<event,event records>>
6464 in the channel named `my-channel`:
6468 $ lttng add-context --kernel --channel=my-channel \
6469 --type='$app:retriever:cur_msg_id'
6472 IMPORTANT: Make sure to always quote the `$` character when you
6473 use man:lttng-add-context(1) from a shell.
6476 NOTE: You can't remove context fields from a channel once you add it.
6481 === Track process attributes
6483 It's often useful to only allow processes with specific attributes to
6484 emit events. For example, you may wish to record all the system calls
6485 which a given process makes (à la
6486 http://linux.die.net/man/1/strace[strace]).
6488 The man:lttng-track(1) and man:lttng-untrack(1) commands serve this
6489 purpose. Both commands operate on _inclusion sets_ of process attribute
6490 values. The available process attribute types are:
6492 Linux kernel <<domain,tracing domain>> only::
6496 * Virtual process ID (VPID).
6498 This is the PID as seen by the application.
6500 * Unix user ID (UID) (since LTTng{nbsp}2.12).
6502 * Virtual Unix user ID (VUID) (since LTTng{nbsp}2.12).
6504 This is the UID as seen by the application.
6506 * Unix group ID (GID) (since LTTng{nbsp}2.12).
6508 * Virtual Unix group ID (VGID) (since LTTng{nbsp}2.12).
6510 This is the GID as seen by the application.
6513 User space tracing domain::
6516 * VUID (since LTTng{nbsp}2.12).
6517 * VGID (since LTTng{nbsp}2.12).
6519 Each tracing domain has one inclusion set per process attribute type:
6520 the Linux kernel tracing domain has six while the user space tracing
6523 For a given event which passes an enabled <<event,event rule>> to be
6524 recorded, _all_ the attributes of its executing process must be part of
6525 the inclusion sets of the tracing domain of the event rule.
6527 Add entries to an inclusion set with the man:lttng-track(1) command and
6528 remove entries with the man:lttng-untrack(1) command. A process
6529 attribute is _tracked_ when it's part of an inclusion set and
6530 _untracked_ otherwise.
6534 The process attribute values are _numeric_.
6536 Should a process with a given tracked process ID, for example, exit, and
6537 then a new process be given this ID, then the latter would also be
6538 allowed to emit events.
6540 With the `lttng track` command, you can add Unix user and group _names_
6541 to the user and group inclusion sets: the <<lttng-sessiond,session
6542 daemon>> finds the corresponding UID, VUID, GID, or VGID once on
6543 _addition_ to the inclusion set. This means that if you rename the user
6544 or group after you run `lttng track`, its user/group ID remains tracked.
6547 .Track and untrack virtual process IDs.
6549 For the sake of the following example, assume the target system has
6550 16{nbsp}possible VPIDs.
6553 <<creating-destroying-tracing-sessions,create a tracing session>>,
6554 the user space VPID inclusion set contains _all_ the possible VPIDs:
6557 .All VPIDs are tracked.
6558 image::track-all.png[]
6560 When the inclusion set is full and you use the man:lttng-track(1)
6561 command to specify some VPIDs to track, LTTng first clears the inclusion
6562 set, and then it adds the specific VPIDs to track. After:
6566 $ lttng track --userspace --vpid=3,4,7,10,13
6569 the VPID inclusion set is:
6572 .VPIDs 3, 4, 7, 10, and 13 are tracked.
6573 image::track-3-4-7-10-13.png[]
6575 Add more VPIDs to the inclusion set afterwards:
6579 $ lttng track --userspace --vpid=1,15,16
6585 .VPIDs 1, 15, and 16 are added to the inclusion set.
6586 image::track-1-3-4-7-10-13-15-16.png[]
6588 The man:lttng-untrack(1) command removes entries from process attribute
6589 inclusion sets. Given the previous example, the following command:
6593 $ lttng untrack --userspace --vpid=3,7,10,13
6596 leads to this VPID inclusion set:
6599 .VPIDs 3, 7, 10, and 13 are removed from the inclusion set.
6600 image::track-1-4-15-16.png[]
6602 LTTng can track all the possible VPIDs again using the
6603 opt:lttng-track(1):--all option:
6607 $ lttng track --userspace --vpid --all
6610 The result is, again:
6613 .All VPIDs are tracked.
6614 image::track-all.png[]
6617 .Track only specific process attributes.
6619 A typical use case with process attribute tracking is to start with an
6620 empty inclusion set, then <<basic-tracing-session-control,start the
6621 tracers>>, and then add entries manually while the tracers are active.
6623 Use the opt:lttng-untrack(1):--all option of the
6624 man:lttng-untrack(1) command to clear the inclusion set after you
6625 <<creating-destroying-tracing-sessions,create a tracing session>>, for
6626 example (with UIDs):
6630 $ lttng untrack --kernel --uid --all
6636 .No UIDs are tracked.
6637 image::untrack-all.png[]
6639 If you trace with this inclusion set configuration, the LTTng kernel
6640 tracer records no events within the <<cur-tracing-session,current
6641 tracing session>> because it doesn't track any UID. Use the
6642 man:lttng-track(1) command as usual to track specific UIDs when you need
6647 $ lttng track --kernel --uid=http,11
6653 .UIDs 6 (`http`) and 11 are tracked.
6654 image::track-6-11.png[]
6659 [[saving-loading-tracing-session]]
6660 === Save and load tracing session configurations
6662 Configuring a <<tracing-session,tracing session>> can be long. Some of
6663 the tasks involved are:
6665 * <<enabling-disabling-channels,Create channels>> with
6666 specific attributes.
6667 * <<adding-context,Add context fields>> to specific channels.
6668 * <<enabling-disabling-events,Create event rules>> with specific log
6669 level and filter conditions.
6671 If you use LTTng to solve real world problems, chances are you have to
6672 record events using the same tracing session setup over and over,
6673 modifying a few variables each time in your instrumented program
6674 or environment. To avoid constant tracing session reconfiguration,
6675 the man:lttng(1) command-line tool can save and load tracing session
6676 configurations to/from XML files.
6678 To save a given tracing session configuration:
6680 * Use the man:lttng-save(1) command:
6685 $ lttng save my-session
6689 Replace `my-session` with the name of the tracing session to save.
6691 LTTng saves tracing session configurations to
6692 dir:{$LTTNG_HOME/.lttng/sessions} by default. Note that the
6693 env:LTTNG_HOME environment variable defaults to `$HOME` if not set. Use
6694 the opt:lttng-save(1):--output-path option to change this destination
6697 LTTng saves all configuration parameters, for example:
6699 * The tracing session name.
6700 * The trace data output path.
6701 * The channels with their state and all their attributes.
6702 * The context fields you added to channels.
6703 * The event rules with their state, log level and filter conditions.
6705 To load a tracing session:
6707 * Use the man:lttng-load(1) command:
6712 $ lttng load my-session
6716 Replace `my-session` with the name of the tracing session to load.
6718 When LTTng loads a configuration, it restores your saved tracing session
6719 as if you just configured it manually.
6721 See man:lttng-load(1) for the complete list of command-line options. You
6722 can also save and load many sessions at a time, and decide in which
6723 directory to output the XML files.
6726 [[sending-trace-data-over-the-network]]
6727 === Send trace data over the network
6729 LTTng can send the recorded trace data to a remote system over the
6730 network instead of writing it to the local file system.
6732 To send the trace data over the network:
6734 . On the _remote_ system (which can also be the target system),
6735 start an LTTng <<lttng-relayd,relay daemon>> (man:lttng-relayd(8)):
6744 . On the _target_ system, create a tracing session configured to
6745 send trace data over the network:
6750 $ lttng create my-session --set-url=net://remote-system
6754 Replace `remote-system` by the host name or IP address of the
6755 remote system. See man:lttng-create(1) for the exact URL format.
6757 . On the target system, use the man:lttng(1) command-line tool as usual.
6758 When tracing is active, the consumer daemon of the target sends
6759 sub-buffers to the relay daemon running on the remote system instead
6760 of flushing them to the local file system. The relay daemon writes the
6761 received packets to the local file system.
6763 The relay daemon writes trace files to
6764 +$LTTNG_HOME/lttng-traces/__hostname__/__session__+ by default, where
6765 +__hostname__+ is the host name of the target system and +__session__+
6766 is the tracing session name. Note that the env:LTTNG_HOME environment
6767 variable defaults to `$HOME` if not set. Use the
6768 opt:lttng-relayd(8):--output option of man:lttng-relayd(8) to write
6769 trace files to another base directory.
6774 === View events as LTTng emits them (noch:{LTTng} live)
6776 LTTng live is a network protocol implemented by the <<lttng-relayd,relay
6777 daemon>> (man:lttng-relayd(8)) to allow compatible trace viewers to
6778 display events as LTTng emits them on the target system while tracing is
6781 The relay daemon creates a _tee_: it forwards the trace data to both
6782 the local file system and to connected live viewers:
6785 .The relay daemon creates a _tee_, forwarding the trace data to both trace files and a connected live viewer.
6790 . On the _target system_, create a <<tracing-session,tracing session>>
6796 $ lttng create my-session --live
6800 This spawns a local relay daemon.
6802 . Start the live viewer and configure it to connect to the relay
6803 daemon. For example, with
6804 https://babeltrace.org/docs/v2.0/man1/babeltrace2.1/[cmd:babeltrace2]:
6809 $ babeltrace2 net://localhost/host/hostname/my-session
6816 * `hostname` with the host name of the target system.
6817 * `my-session` with the name of the tracing session to view.
6820 . Configure the tracing session as usual with the man:lttng(1)
6821 command-line tool, and <<basic-tracing-session-control,start tracing>>.
6823 List the available live tracing sessions with Babeltrace{nbsp}2:
6827 $ babeltrace2 net://localhost
6830 You can start the relay daemon on another system. In this case, you need
6831 to specify the URL of the relay daemon when you create the tracing
6832 session with the opt:lttng-create(1):--set-url option. You also need to
6833 replace `localhost` in the procedure above with the host name of the
6834 system on which the relay daemon is running.
6836 See man:lttng-create(1) and man:lttng-relayd(8) for the complete list of
6837 command-line options.
6841 [[taking-a-snapshot]]
6842 === Take a snapshot of the current sub-buffers of a tracing session
6844 The normal behavior of LTTng is to append full sub-buffers to growing
6845 trace data files. This is ideal to keep a full history of the events
6846 that occurred on the target system, but it can
6847 represent too much data in some situations. For example, you may wish
6848 to trace your application continuously until some critical situation
6849 happens, in which case you only need the latest few recorded
6850 events to perform the desired analysis, not multi-gigabyte trace files.
6852 With the man:lttng-snapshot(1) command, you can take a snapshot of the
6853 current sub-buffers of a given <<tracing-session,tracing session>>.
6854 LTTng can write the snapshot to the local file system or send it over
6858 .A snapshot is a copy of the current sub-buffers, which aren't cleared after the operation.
6859 image::snapshot.png[]
6861 If you wish to create unmanaged, self-contained, non-overlapping
6862 trace chunk archives instead of a simple copy of the current
6863 sub-buffers, see the <<session-rotation,tracing session rotation>>
6864 feature (available since LTTng{nbsp}2.11).
6868 . Create a tracing session in _snapshot mode_:
6873 $ lttng create my-session --snapshot
6877 The <<channel-overwrite-mode-vs-discard-mode,event record loss mode>> of
6878 <<channel,channels>> created in this mode is automatically set to
6879 _overwrite_ (flight recorder mode).
6881 . Configure the tracing session as usual with the man:lttng(1)
6882 command-line tool, and <<basic-tracing-session-control,start tracing>>.
6884 . **Optional**: When you need to take a snapshot,
6885 <<basic-tracing-session-control,stop tracing>>.
6887 You can take a snapshot when the tracers are active, but if you stop
6888 them first, you're sure that the data in the sub-buffers doesn't
6889 change before you actually take the snapshot.
6896 $ lttng snapshot record --name=my-first-snapshot
6900 LTTng writes the current sub-buffers of all the channels of the
6901 <<cur-tracing-session,current tracing session>> to
6902 trace files on the local file system. Those trace files have
6903 `my-first-snapshot` in their name.
6905 There is no difference between the format of a normal trace file and the
6906 format of a snapshot: viewers of LTTng traces also support LTTng
6909 By default, LTTng writes snapshot files to the path shown by
6910 `lttng snapshot list-output`. You can change this path or decide to send
6911 snapshots over the network using either:
6913 . An output path or URL that you specify when you
6914 <<creating-destroying-tracing-sessions,create the tracing session>>.
6915 . A snapshot output path or URL that you add using
6916 `lttng snapshot add-output`.
6917 . An output path or URL that you provide directly to the
6918 `lttng snapshot record` command.
6920 Method{nbsp}3 overrides method{nbsp}2, which overrides method 1. When
6921 you specify a URL, a relay daemon must listen on a remote system (see
6922 <<sending-trace-data-over-the-network,Send trace data over the
6927 [[session-rotation]]
6928 === Archive the current trace chunk (rotate a tracing session)
6930 The <<taking-a-snapshot,snapshot user guide>> shows how to dump the
6931 current sub-buffers of a tracing session to the file system or send them
6932 over the network. When you take a snapshot, LTTng doesn't clear the ring
6933 buffers of the tracing session: if you take another snapshot immediately
6934 after, both snapshots could contain overlapping trace data.
6936 Inspired by https://en.wikipedia.org/wiki/Log_rotation[log rotation],
6937 _tracing session rotation_ is a feature which appends the content of the
6938 ring buffers to what's already on the file system or sent over the
6939 network since the creation of the tracing session or since the last
6940 rotation, and then clears those ring buffers to avoid trace data
6943 What LTTng is about to write when performing a tracing session rotation
6944 is called the _current trace chunk_. When this current trace chunk is
6945 written to the file system or sent over the network, it becomes a _trace
6946 chunk archive_. Therefore, a tracing session rotation _archives_ the
6947 current trace chunk.
6950 .A tracing session rotation operation _archives_ the current trace chunk.
6951 image::rotation.png[]
6953 A trace chunk archive is a self-contained LTTng trace which LTTng
6954 doesn't manage anymore: you can read it, modify it, move it, or remove
6957 There are two methods to perform a tracing session rotation: immediately
6958 or with a rotation schedule.
6960 To perform an immediate tracing session rotation:
6962 . <<creating-destroying-tracing-sessions,Create a tracing session>>
6963 in _normal mode_ or _network streaming mode_
6964 (only those two creation modes support tracing session rotation):
6969 $ lttng create my-session
6973 . <<enabling-disabling-events,Create one or more event rules>>
6974 and <<basic-tracing-session-control,start tracing>>:
6979 $ lttng enable-event --kernel sched_'*'
6984 . When needed, immediately rotate the
6985 <<cur-tracing-session,current tracing session>>:
6994 The cmd:lttng-rotate command prints the path to the created trace
6995 chunk archive. See man:lttng-rotate(1) to learn about the format
6996 of trace chunk archive directory names.
6998 Perform other immediate rotations while the tracing session is
6999 active. It is guaranteed that all the trace chunk archives don't
7000 contain overlapping trace data. You can also perform an immediate
7001 rotation once you have <<basic-tracing-session-control,stopped>> the
7004 . When you're done tracing,
7005 <<creating-destroying-tracing-sessions,destroy the current tracing
7015 The tracing session destruction operation creates one last trace
7016 chunk archive from the current trace chunk.
7018 A tracing session rotation schedule is a planned rotation which LTTng
7019 performs automatically based on one of the following conditions:
7021 * A timer with a configured period times out.
7023 * The total size of the flushed part of the current trace chunk
7024 becomes greater than or equal to a configured value.
7026 To schedule a tracing session rotation, set a _rotation schedule_:
7028 . <<creating-destroying-tracing-sessions,Create a tracing session>>
7029 in _normal mode_ or _network streaming mode_
7030 (only those two creation modes support tracing session rotation):
7035 $ lttng create my-session
7039 . <<enabling-disabling-events,Create one or more event rules>>:
7044 $ lttng enable-event --kernel sched_'*'
7048 . Set a tracing session rotation schedule:
7053 $ lttng enable-rotation --timer=10s
7057 In this example, we set a rotation schedule so that LTTng performs a
7058 tracing session rotation every ten seconds.
7060 See man:lttng-enable-rotation(1) to learn more about other ways to set a
7063 . <<basic-tracing-session-control,Start tracing>>:
7072 LTTng performs tracing session rotations automatically while the tracing
7073 session is active thanks to the rotation schedule.
7075 . When you're done tracing,
7076 <<creating-destroying-tracing-sessions,destroy the current tracing
7086 The tracing session destruction operation creates one last trace chunk
7087 archive from the current trace chunk.
7089 Use man:lttng-disable-rotation(1) to unset a tracing session
7092 NOTE: man:lttng-rotate(1) and man:lttng-enable-rotation(1) list
7093 limitations regarding those two commands.
7098 === Use the machine interface
7100 With any command of the man:lttng(1) command-line tool, set the
7101 opt:lttng(1):--mi option to `xml` (before the command name) to get an
7102 XML machine interface output, for example:
7106 $ lttng --mi=xml enable-event --kernel --syscall open
7109 A schema definition (XSD) is
7110 https://github.com/lttng/lttng-tools/blob/stable-{revision}/src/common/src/common/mi-lttng-4.0.xsd[available]
7111 to ease the integration with external tools as much as possible.
7115 [[metadata-regenerate]]
7116 === Regenerate the metadata of an LTTng trace
7118 An LTTng trace, which is a http://diamon.org/ctf[CTF] trace, has both
7119 data stream files and a metadata file. This metadata file contains,
7120 amongst other things, information about the offset of the clock sources
7121 used to timestamp <<event,event records>> when tracing.
7123 If, once a <<tracing-session,tracing session>> is
7124 <<basic-tracing-session-control,started>>, a major
7125 https://en.wikipedia.org/wiki/Network_Time_Protocol[NTP] correction
7126 happens, the clock offset of the trace also needs to be updated. Use
7127 the `metadata` item of the man:lttng-regenerate(1) command to do so.
7129 The main use case of this command is to allow a system to boot with
7130 an incorrect wall time and trace it with LTTng before its wall time
7131 is corrected. Once the system is known to be in a state where its
7132 wall time is correct, it can run `lttng regenerate metadata`.
7134 To regenerate the metadata of an LTTng trace:
7136 * Use the `metadata` item of the man:lttng-regenerate(1) command:
7141 $ lttng regenerate metadata
7147 `lttng regenerate metadata` has the following limitations:
7149 * Tracing session <<creating-destroying-tracing-sessions,created>>
7151 * User space <<channel,channels>>, if any, are using
7152 <<channel-buffering-schemes,per-user buffering>>.
7157 [[regenerate-statedump]]
7158 === Regenerate the state dump of a tracing session
7160 The LTTng kernel and user space tracers generate state dump
7161 <<event,event records>> when the application starts or when you
7162 <<basic-tracing-session-control,start a tracing session>>. An analysis
7163 can use the state dump event records to set an initial state before it
7164 builds the rest of the state from the following event records.
7165 http://tracecompass.org/[Trace Compass] is a notable example of an
7166 application which uses the state dump of an LTTng trace.
7168 When you <<taking-a-snapshot,take a snapshot>>, it's possible that the
7169 state dump event records aren't included in the snapshot because they
7170 were recorded to a sub-buffer that has been consumed or overwritten
7173 Use the `lttng regenerate statedump` command to emit the state
7174 dump event records again.
7176 To regenerate the state dump of the current tracing session, provided
7177 create it in snapshot mode, before you take a snapshot:
7179 . Use the `statedump` item of the man:lttng-regenerate(1) command:
7184 $ lttng regenerate statedump
7188 . <<basic-tracing-session-control,Stop the tracing session>>:
7197 . <<taking-a-snapshot,Take a snapshot>>:
7202 $ lttng snapshot record --name=my-snapshot
7206 Depending on the event throughput, you should run steps 1 and 2
7207 as closely as possible.
7209 NOTE: To record the state dump events, you need to
7210 <<enabling-disabling-events,create event rules>> which enable them.
7211 LTTng-UST state dump tracepoints start with `lttng_ust_statedump:`.
7212 LTTng-modules state dump tracepoints start with `lttng_statedump_`.
7216 [[persistent-memory-file-systems]]
7217 === Record trace data on persistent memory file systems
7219 https://en.wikipedia.org/wiki/Non-volatile_random-access_memory[Non-volatile random-access memory]
7220 (NVRAM) is random-access memory that retains its information when power
7221 is turned off (non-volatile). Systems with such memory can store data
7222 structures in RAM and retrieve them after a reboot, without flushing
7223 to typical _storage_.
7225 Linux supports NVRAM file systems thanks to either
7226 https://www.kernel.org/doc/Documentation/filesystems/dax.txt[DAX]{nbsp}+{nbsp}http://lkml.iu.edu/hypermail/linux/kernel/1504.1/03463.html[pmem]
7227 (requires Linux{nbsp}4.1+) or http://pramfs.sourceforge.net/[PRAMFS] (requires Linux{nbsp}<{nbsp}4).
7229 This section doesn't describe how to operate such file systems;
7230 we assume that you have a working persistent memory file system.
7232 When you create a <<tracing-session,tracing session>>, you can specify
7233 the path of the shared memory holding the sub-buffers. If you specify a
7234 location on an NVRAM file system, then you can retrieve the latest
7235 recorded trace data when the system reboots after a crash.
7237 To record trace data on a persistent memory file system and retrieve the
7238 trace data after a system crash:
7240 . Create a tracing session with a sub-buffer shared memory path located
7241 on an NVRAM file system:
7246 $ lttng create my-session --shm-path=/path/to/shm
7250 . Configure the tracing session as usual with the man:lttng(1)
7251 command-line tool, and <<basic-tracing-session-control,start tracing>>.
7253 . After a system crash, use the man:lttng-crash(1) command-line tool to
7254 view the trace data recorded on the NVRAM file system:
7259 $ lttng-crash /path/to/shm
7263 The binary layout of the ring buffer files isn't exactly the same as
7264 the trace files layout. This is why you need to use man:lttng-crash(1)
7265 instead of your preferred trace viewer directly.
7267 To convert the ring buffer files to LTTng trace files:
7269 * Use the opt:lttng-crash(1):--extract option of man:lttng-crash(1):
7274 $ lttng-crash --extract=/path/to/trace /path/to/shm
7280 [[notif-trigger-api]]
7281 === Get notified when the buffer usage of a channel is too high or too low
7283 With the $$C/C++$$ notification and trigger API of LTTng, your user
7284 application can get notified when the buffer usage of one or more
7285 <<channel,channels>> becomes too low or too high. Use this API
7286 and enable or disable <<event,event rules>> during tracing to avoid
7287 <<channel-overwrite-mode-vs-discard-mode,discarded event records>>.
7289 .Have a user application get notified when the buffer usage of an LTTng channel is too high.
7291 In this example, we create and build an application which gets notified
7292 when the buffer usage of a specific LTTng channel is higher than
7293 75{nbsp}%. We only print that it is the case in the example, but we
7294 could as well use the API of <<liblttng-ctl-lttng,`liblttng-ctl`>> to
7295 disable event rules when this happens.
7297 . Create the C{nbsp}source file of application:
7305 #include <lttng/domain.h>
7306 #include <lttng/action/action.h>
7307 #include <lttng/action/notify.h>
7308 #include <lttng/condition/condition.h>
7309 #include <lttng/condition/buffer-usage.h>
7310 #include <lttng/condition/evaluation.h>
7311 #include <lttng/notification/channel.h>
7312 #include <lttng/notification/notification.h>
7313 #include <lttng/trigger/trigger.h>
7314 #include <lttng/endpoint.h>
7316 int main(int argc, char *argv[])
7318 int exit_status = 0;
7319 struct lttng_notification_channel *notification_channel;
7320 struct lttng_condition *condition;
7321 struct lttng_action *action;
7322 struct lttng_trigger *trigger;
7323 const char *tracing_session_name;
7324 const char *channel_name;
7327 tracing_session_name = argv[1];
7328 channel_name = argv[2];
7331 * Create a notification channel. A notification channel
7332 * connects the user application to the LTTng session daemon.
7333 * This notification channel can be used to listen to various
7334 * types of notifications.
7336 notification_channel = lttng_notification_channel_create(
7337 lttng_session_daemon_notification_endpoint);
7340 * Create a "high buffer usage" condition. In this case, the
7341 * condition is reached when the buffer usage is greater than or
7342 * equal to 75 %. We create the condition for a specific tracing
7343 * session name, channel name, and for the user space tracing
7346 * The "low buffer usage" condition type also exists.
7348 condition = lttng_condition_buffer_usage_high_create();
7349 lttng_condition_buffer_usage_set_threshold_ratio(condition, .75);
7350 lttng_condition_buffer_usage_set_session_name(
7351 condition, tracing_session_name);
7352 lttng_condition_buffer_usage_set_channel_name(condition,
7354 lttng_condition_buffer_usage_set_domain_type(condition,
7358 * Create an action (get a notification) to take when the
7359 * condition created above is reached.
7361 action = lttng_action_notify_create();
7364 * Create a trigger. A trigger associates a condition to an
7365 * action: the action is executed when the condition is reached.
7367 trigger = lttng_trigger_create(condition, action);
7369 /* Register the trigger to LTTng. */
7370 lttng_register_trigger(trigger);
7373 * Now that we have registered a trigger, a notification will be
7374 * emitted everytime its condition is met. To receive this
7375 * notification, we must subscribe to notifications that match
7376 * the same condition.
7378 lttng_notification_channel_subscribe(notification_channel,
7382 * Notification loop. Put this in a dedicated thread to avoid
7383 * blocking the main thread.
7386 struct lttng_notification *notification;
7387 enum lttng_notification_channel_status status;
7388 const struct lttng_evaluation *notification_evaluation;
7389 const struct lttng_condition *notification_condition;
7390 double buffer_usage;
7392 /* Receive the next notification. */
7393 status = lttng_notification_channel_get_next_notification(
7394 notification_channel, ¬ification);
7397 case LTTNG_NOTIFICATION_CHANNEL_STATUS_OK:
7399 case LTTNG_NOTIFICATION_CHANNEL_STATUS_NOTIFICATIONS_DROPPED:
7401 * The session daemon can drop notifications if a monitoring
7402 * application isn't consuming the notifications fast
7406 case LTTNG_NOTIFICATION_CHANNEL_STATUS_CLOSED:
7408 * The notification channel has been closed by the
7409 * session daemon. This is typically caused by a session
7410 * daemon shutting down.
7414 /* Unhandled conditions or errors. */
7420 * A notification provides, amongst other things:
7422 * * The condition that caused this notification to be
7424 * * The condition evaluation, which provides more
7425 * specific information on the evaluation of the
7428 * The condition evaluation provides the buffer usage
7429 * value at the moment the condition was reached.
7431 notification_condition = lttng_notification_get_condition(
7433 notification_evaluation = lttng_notification_get_evaluation(
7436 /* We're subscribed to only one condition. */
7437 assert(lttng_condition_get_type(notification_condition) ==
7438 LTTNG_CONDITION_TYPE_BUFFER_USAGE_HIGH);
7441 * Get the exact sampled buffer usage from the
7442 * condition evaluation.
7444 lttng_evaluation_buffer_usage_get_usage_ratio(
7445 notification_evaluation, &buffer_usage);
7448 * At this point, instead of printing a message, we
7449 * could do something to reduce the buffer usage of the channel,
7450 * like disable specific events.
7452 printf("Buffer usage is %f %% in tracing session \"%s\", "
7453 "user space channel \"%s\".\n", buffer_usage * 100,
7454 tracing_session_name, channel_name);
7455 lttng_notification_destroy(notification);
7459 lttng_action_destroy(action);
7460 lttng_condition_destroy(condition);
7461 lttng_trigger_destroy(trigger);
7462 lttng_notification_channel_destroy(notification_channel);
7468 . Build the `notif-app` application, linking it to `liblttng-ctl`:
7473 $ gcc -o notif-app notif-app.c -llttng-ctl
7477 . <<creating-destroying-tracing-sessions,Create a tracing session>>,
7478 <<enabling-disabling-events,create an event rule>> matching all the
7479 user space tracepoints, and
7480 <<basic-tracing-session-control,start tracing>>:
7485 $ lttng create my-session
7486 $ lttng enable-event --userspace --all
7491 If you create the channel manually with the man:lttng-enable-channel(1)
7492 command, control how frequently LTTng samples the current values of the
7493 channel properties to evaluate user conditions with the
7494 opt:lttng-enable-channel(1):--monitor-timer option.
7496 . Run the `notif-app` application. This program accepts the
7497 <<tracing-session,tracing session>> name and the user space channel
7498 name as its two first arguments. The channel which LTTng automatically
7499 creates with the man:lttng-enable-event(1) command above is named
7505 $ ./notif-app my-session channel0
7509 . In another terminal, run an application with a very high event
7510 throughput so that the 75{nbsp}% buffer usage condition is reached.
7512 In the first terminal, the application should print lines like this:
7515 Buffer usage is 81.45197 % in tracing session "my-session", user space
7519 If you don't see anything, try modifying the condition in
7520 path:{notif-app.c} to a lower value (0.1, for example), rebuilding it
7521 (step{nbsp}2) and running it again (step{nbsp}4).
7528 [[lttng-modules-ref]]
7529 === noch:{LTTng-modules}
7533 [[lttng-tracepoint-enum]]
7534 ==== `LTTNG_TRACEPOINT_ENUM()` usage
7536 Use the `LTTNG_TRACEPOINT_ENUM()` macro to define an enumeration:
7540 LTTNG_TRACEPOINT_ENUM(name, TP_ENUM_VALUES(entries))
7545 * `name` with the name of the enumeration (C identifier, unique
7546 amongst all the defined enumerations).
7547 * `entries` with a list of enumeration entries.
7549 The available enumeration entry macros are:
7551 +ctf_enum_value(__name__, __value__)+::
7552 Entry named +__name__+ mapped to the integral value +__value__+.
7554 +ctf_enum_range(__name__, __begin__, __end__)+::
7555 Entry named +__name__+ mapped to the range of integral values between
7556 +__begin__+ (included) and +__end__+ (included).
7558 +ctf_enum_auto(__name__)+::
7559 Entry named +__name__+ mapped to the integral value following the
7562 The last value of a `ctf_enum_value()` entry is its +__value__+
7565 The last value of a `ctf_enum_range()` entry is its +__end__+ parameter.
7567 If `ctf_enum_auto()` is the first entry in the list, its integral
7570 Use the `ctf_enum()` <<lttng-modules-tp-fields,field definition macro>>
7571 to use a defined enumeration as a tracepoint field.
7573 .Define an enumeration with `LTTNG_TRACEPOINT_ENUM()`.
7577 LTTNG_TRACEPOINT_ENUM(
7580 ctf_enum_auto("AUTO: EXPECT 0")
7581 ctf_enum_value("VALUE: 23", 23)
7582 ctf_enum_value("VALUE: 27", 27)
7583 ctf_enum_auto("AUTO: EXPECT 28")
7584 ctf_enum_range("RANGE: 101 TO 303", 101, 303)
7585 ctf_enum_auto("AUTO: EXPECT 304")
7593 [[lttng-modules-tp-fields]]
7594 ==== Tracepoint fields macros (for `TP_FIELDS()`)
7596 [[tp-fast-assign]][[tp-struct-entry]]The available macros to define
7597 tracepoint fields, which must be listed within `TP_FIELDS()` in
7598 `LTTNG_TRACEPOINT_EVENT()`, are:
7600 [role="func-desc growable",cols="asciidoc,asciidoc"]
7601 .Available macros to define LTTng-modules tracepoint fields
7603 |Macro |Description and parameters
7606 +ctf_integer(__t__, __n__, __e__)+
7608 +ctf_integer_nowrite(__t__, __n__, __e__)+
7610 +ctf_user_integer(__t__, __n__, __e__)+
7612 +ctf_user_integer_nowrite(__t__, __n__, __e__)+
7614 Standard integer, displayed in base{nbsp}10.
7617 Integer C type (`int`, `long`, `size_t`, ...).
7623 Argument expression.
7626 +ctf_integer_hex(__t__, __n__, __e__)+
7628 +ctf_user_integer_hex(__t__, __n__, __e__)+
7630 Standard integer, displayed in base{nbsp}16.
7639 Argument expression.
7641 |+ctf_integer_oct(__t__, __n__, __e__)+
7643 Standard integer, displayed in base{nbsp}8.
7652 Argument expression.
7655 +ctf_integer_network(__t__, __n__, __e__)+
7657 +ctf_user_integer_network(__t__, __n__, __e__)+
7659 Integer in network byte order (big-endian), displayed in base{nbsp}10.
7668 Argument expression.
7671 +ctf_integer_network_hex(__t__, __n__, __e__)+
7673 +ctf_user_integer_network_hex(__t__, __n__, __e__)+
7675 Integer in network byte order, displayed in base{nbsp}16.
7684 Argument expression.
7687 +ctf_enum(__N__, __t__, __n__, __e__)+
7689 +ctf_enum_nowrite(__N__, __t__, __n__, __e__)+
7691 +ctf_user_enum(__N__, __t__, __n__, __e__)+
7693 +ctf_user_enum_nowrite(__N__, __t__, __n__, __e__)+
7698 Name of a <<lttng-tracepoint-enum,previously defined enumeration>>.
7701 Integer C type (`int`, `long`, `size_t`, ...).
7707 Argument expression.
7710 +ctf_string(__n__, __e__)+
7712 +ctf_string_nowrite(__n__, __e__)+
7714 +ctf_user_string(__n__, __e__)+
7716 +ctf_user_string_nowrite(__n__, __e__)+
7718 Null-terminated string; undefined behavior if +__e__+ is `NULL`.
7724 Argument expression.
7727 +ctf_array(__t__, __n__, __e__, __s__)+
7729 +ctf_array_nowrite(__t__, __n__, __e__, __s__)+
7731 +ctf_user_array(__t__, __n__, __e__, __s__)+
7733 +ctf_user_array_nowrite(__t__, __n__, __e__, __s__)+
7735 Statically-sized array of integers.
7738 Array element C type.
7744 Argument expression.
7750 +ctf_array_bitfield(__t__, __n__, __e__, __s__)+
7752 +ctf_array_bitfield_nowrite(__t__, __n__, __e__, __s__)+
7754 +ctf_user_array_bitfield(__t__, __n__, __e__, __s__)+
7756 +ctf_user_array_bitfield_nowrite(__t__, __n__, __e__, __s__)+
7758 Statically-sized array of bits.
7760 The type of +__e__+ must be an integer type. +__s__+ is the number
7761 of elements of such type in +__e__+, not the number of bits.
7764 Array element C type.
7770 Argument expression.
7776 +ctf_array_text(__t__, __n__, __e__, __s__)+
7778 +ctf_array_text_nowrite(__t__, __n__, __e__, __s__)+
7780 +ctf_user_array_text(__t__, __n__, __e__, __s__)+
7782 +ctf_user_array_text_nowrite(__t__, __n__, __e__, __s__)+
7784 Statically-sized array, printed as text.
7786 The string doesn't need to be null-terminated.
7789 Array element C type (always `char`).
7795 Argument expression.
7801 +ctf_sequence(__t__, __n__, __e__, __T__, __E__)+
7803 +ctf_sequence_nowrite(__t__, __n__, __e__, __T__, __E__)+
7805 +ctf_user_sequence(__t__, __n__, __e__, __T__, __E__)+
7807 +ctf_user_sequence_nowrite(__t__, __n__, __e__, __T__, __E__)+
7809 Dynamically-sized array of integers.
7811 The type of +__E__+ must be unsigned.
7814 Array element C type.
7820 Argument expression.
7823 Length expression C type.
7829 +ctf_sequence_hex(__t__, __n__, __e__, __T__, __E__)+
7831 +ctf_user_sequence_hex(__t__, __n__, __e__, __T__, __E__)+
7833 Dynamically-sized array of integers, displayed in base{nbsp}16.
7835 The type of +__E__+ must be unsigned.
7838 Array element C type.
7844 Argument expression.
7847 Length expression C type.
7852 |+ctf_sequence_network(__t__, __n__, __e__, __T__, __E__)+
7854 Dynamically-sized array of integers in network byte order (big-endian),
7855 displayed in base{nbsp}10.
7857 The type of +__E__+ must be unsigned.
7860 Array element C type.
7866 Argument expression.
7869 Length expression C type.
7875 +ctf_sequence_bitfield(__t__, __n__, __e__, __T__, __E__)+
7877 +ctf_sequence_bitfield_nowrite(__t__, __n__, __e__, __T__, __E__)+
7879 +ctf_user_sequence_bitfield(__t__, __n__, __e__, __T__, __E__)+
7881 +ctf_user_sequence_bitfield_nowrite(__t__, __n__, __e__, __T__, __E__)+
7883 Dynamically-sized array of bits.
7885 The type of +__e__+ must be an integer type. +__s__+ is the number
7886 of elements of such type in +__e__+, not the number of bits.
7888 The type of +__E__+ must be unsigned.
7891 Array element C type.
7897 Argument expression.
7900 Length expression C type.
7906 +ctf_sequence_text(__t__, __n__, __e__, __T__, __E__)+
7908 +ctf_sequence_text_nowrite(__t__, __n__, __e__, __T__, __E__)+
7910 +ctf_user_sequence_text(__t__, __n__, __e__, __T__, __E__)+
7912 +ctf_user_sequence_text_nowrite(__t__, __n__, __e__, __T__, __E__)+
7914 Dynamically-sized array, displayed as text.
7916 The string doesn't need to be null-terminated.
7918 The type of +__E__+ must be unsigned.
7920 The behaviour is undefined if +__e__+ is `NULL`.
7923 Sequence element C type (always `char`).
7929 Argument expression.
7932 Length expression C type.
7938 Use the `_user` versions when the argument expression, `e`, is
7939 a user space address. In the cases of `ctf_user_integer*()` and
7940 `ctf_user_float*()`, `&e` must be a user space address, thus `e` must
7943 The `_nowrite` versions omit themselves from the session trace, but are
7944 otherwise identical. This means the `_nowrite` fields won't be written
7945 in the recorded trace. Their primary purpose is to make some
7946 of the event context available to the
7947 <<enabling-disabling-events,event filters>> without having to
7948 commit the data to sub-buffers.
7954 Terms related to LTTng and to tracing in general:
7957 The http://diamon.org/babeltrace[Babeltrace] project, which includes:
7960 https://babeltrace.org/docs/v2.0/man1/babeltrace2.1/[cmd:babeltrace2]
7961 command-line interface.
7962 * The libbabeltrace2 library which offers a
7963 https://babeltrace.org/docs/v2.0/libbabeltrace2/[C API].
7964 * https://babeltrace.org/docs/v2.0/python/bt2/[Python{nbsp}3 bindings].
7967 [[def-buffering-scheme]]<<channel-buffering-schemes,buffering scheme>>::
7968 A layout of <<def-sub-buffer,sub-buffers>> applied to a given channel.
7970 [[def-channel]]<<channel,channel>>::
7971 An entity which is responsible for a set of
7972 <<def-ring-buffer,ring buffers>>.
7974 <<def-event-rule,Event rules>> are always attached to a specific
7978 A source of time for a <<def-tracer,tracer>>.
7980 [[def-consumer-daemon]]<<lttng-consumerd,consumer daemon>>::
7981 A process which is responsible for consuming the full
7982 <<def-sub-buffer,sub-buffers>> and write them to a file system or
7983 send them over the network.
7985 [[def-current-trace-chunk]]current trace chunk::
7986 A <<def-trace-chunk,trace chunk>> which includes the current content
7987 of all the <<def-sub-buffer,sub-buffers>> of the
7988 <<def-tracing-session-rotation,tracing session>> and the stream files
7989 produced since the latest event amongst:
7991 * The creation of the <<def-tracing-session,tracing session>>.
7992 * The last tracing session rotation, if any.
7994 <<channel-overwrite-mode-vs-discard-mode,discard mode>>::
7995 The <<def-event-record-loss-mode,event record loss mode>> in which
7996 the <<def-tracer,tracer>> _discards_ new event records when there's no
7997 <<def-sub-buffer,sub-buffer>> space left to store them.
7999 [[def-event]]event::
8000 The consequence of the execution of an
8001 <<def-instrumentation-point,instrumentation point>>, like a
8002 <<def-tracepoint,tracepoint>> that you manually place in some source
8003 code, or a Linux kernel kprobe.
8005 An event is said to _occur_ at a specific time. <<def-lttng,LTTng>> can
8006 take various actions upon the occurrence of an event, like record its
8007 payload to a <<def-sub-buffer,sub-buffer>>.
8009 [[def-event-name]]event name::
8010 The name of an <<def-event,event>>, which is also the name of the
8011 <<def-event-record,event record>>.
8013 This is also called the _instrumentation point name_.
8015 [[def-event-record]]event record::
8016 A record, in a <<def-trace,trace>>, of the payload of an
8017 <<def-event,event>> which occured.
8019 [[def-event-record-loss-mode]]<<channel-overwrite-mode-vs-discard-mode,event record loss mode>>::
8020 The mechanism by which event records of a given
8021 <<def-channel,channel>> are lost (not recorded) when there is no
8022 <<def-sub-buffer,sub-buffer>> space left to store them.
8024 [[def-event-rule]]<<event,event rule>>::
8025 Set of conditions which must be satisfied for one or more occuring
8026 <<def-event,events>> to be recorded.
8028 [[def-incl-set]]inclusion set::
8029 In the <<pid-tracking,process attribute tracking>> context: a
8030 set of <<def-proc-attr,process attributes>> of a given type.
8032 <<instrumenting,instrumentation>>::
8033 The use of <<def-lttng,LTTng>> probes to make a piece of software
8036 [[def-instrumentation-point]]instrumentation point::
8037 A point in the execution path of a piece of software that, when
8038 reached by this execution, can emit an <<def-event,event>>.
8040 instrumentation point name::
8041 See _<<def-event-name,event name>>_.
8043 `java.util.logging`::
8045 https://docs.oracle.com/javase/7/docs/api/java/util/logging/package-summary.html[core logging facilities]
8046 of the Java platform.
8049 A http://logging.apache.org/log4j/1.2/[logging library] for Java
8050 developed by the Apache Software Foundation.
8053 Level of severity of a log statement or user space
8054 <<def-instrumentation-point,instrumentation point>>.
8056 [[def-lttng]]LTTng::
8057 The _Linux Trace Toolkit: next generation_ project.
8059 <<lttng-cli,cmd:lttng>>::
8060 A command-line tool provided by the <<def-lttng-tools,LTTng-tools>>
8061 project which you can use to send and receive control messages to and
8062 from a <<def-session-daemon,session daemon>>.
8065 The https://github.com/lttng/lttng-analyses[LTTng analyses] project,
8066 which is a set of analyzing programs that you can use to obtain a
8067 higher level view of an <<def-lttng,LTTng>> <<def-trace,trace>>.
8069 cmd:lttng-consumerd::
8070 The name of the <<def-consumer-daemon,consumer daemon>> program.
8073 A utility provided by the <<def-lttng-tools,LTTng-tools>> project
8074 which can convert <<def-ring-buffer,ring buffer>> files (usually
8075 <<persistent-memory-file-systems,saved on a persistent memory file
8076 system>>) to <<def-trace,trace>> files.
8078 See man:lttng-crash(1).
8080 LTTng Documentation::
8083 <<lttng-live,LTTng live>>::
8084 A communication protocol between the <<lttng-relayd,relay daemon>> and
8085 live viewers which makes it possible to see <<def-event-record,event
8086 records>> ``live'', as they are received by the
8087 <<def-relay-daemon,relay daemon>>.
8089 <<lttng-modules,LTTng-modules>>::
8090 The https://github.com/lttng/lttng-modules[LTTng-modules] project,
8091 which contains the Linux kernel modules to make the Linux kernel
8092 <<def-instrumentation-point,instrumentation points>> available for
8093 <<def-lttng,LTTng>> tracing.
8096 The name of the <<def-relay-daemon,relay daemon>> program.
8098 cmd:lttng-sessiond::
8099 The name of the <<def-session-daemon,session daemon>> program.
8101 [[def-lttng-tools]]LTTng-tools::
8102 The https://github.com/lttng/lttng-tools[LTTng-tools] project, which
8103 contains the various programs and libraries used to
8104 <<controlling-tracing,control tracing>>.
8106 [[def-lttng-ust]]<<lttng-ust,LTTng-UST>>::
8107 The https://github.com/lttng/lttng-ust[LTTng-UST] project, which
8108 contains libraries to instrument
8109 <<def-user-application,user applications>>.
8111 <<lttng-ust-agents,LTTng-UST Java agent>>::
8112 A Java package provided by the <<def-lttng-ust,LTTng-UST>> project to
8113 allow the LTTng instrumentation of `java.util.logging` and Apache
8114 log4j{nbsp}1.2 logging statements.
8116 <<lttng-ust-agents,LTTng-UST Python agent>>::
8117 A Python package provided by the <<def-lttng-ust,LTTng-UST>> project
8118 to allow the <<def-lttng,LTTng>> instrumentation of Python logging
8121 <<channel-overwrite-mode-vs-discard-mode,overwrite mode>>::
8122 The <<def-event-record-loss-mode,event record loss mode>> in which new
8123 <<def-event-record,event records>> _overwrite_ older event records
8124 when there's no <<def-sub-buffer,sub-buffer>> space left to store
8127 <<channel-buffering-schemes,per-process buffering>>::
8128 A <<def-buffering-scheme,buffering scheme>> in which each instrumented
8129 process has its own <<def-sub-buffer,sub-buffers>> for a given user
8130 space <<def-channel,channel>>.
8132 <<channel-buffering-schemes,per-user buffering>>::
8133 A <<def-buffering-scheme,buffering scheme>> in which all the processes
8134 of a Unix user share the same <<def-sub-buffer,sub-buffers>> for a
8135 given user space <<def-channel,channel>>.
8137 [[def-proc-attr]]process attribute::
8138 In the <<pid-tracking,process attribute tracking>> context:
8141 * A virtual process ID.
8143 * A virtual Unix user ID.
8145 * A virtual Unix group ID.
8147 [[def-relay-daemon]]<<lttng-relayd,relay daemon>>::
8148 A process which is responsible for receiving the <<def-trace,trace>>
8149 data which a distant <<def-consumer-daemon,consumer daemon>> sends.
8151 [[def-ring-buffer]]ring buffer::
8152 A set of <<def-sub-buffer,sub-buffers>>.
8155 See _<<def-tracing-session-rotation,tracing session rotation>>_.
8157 [[def-session-daemon]]<<lttng-sessiond,session daemon>>::
8158 A process which receives control commands from you and orchestrates
8159 the <<def-tracer,tracers>> and various <<def-lttng,LTTng>> daemons.
8161 <<taking-a-snapshot,snapshot>>::
8162 A copy of the current data of all the <<def-sub-buffer,sub-buffers>>
8163 of a given <<def-tracing-session,tracing session>>, saved as
8164 <<def-trace,trace>> files.
8166 [[def-sub-buffer]]sub-buffer::
8167 One part of an <<def-lttng,LTTng>> <<def-ring-buffer,ring buffer>>
8168 which contains <<def-event-record,event records>>.
8171 The time information attached to an <<def-event,event>> when it is
8174 [[def-trace]]trace (_noun_)::
8177 * One http://diamon.org/ctf/[CTF] metadata stream file.
8178 * One or more CTF data stream files which are the concatenations of one
8179 or more flushed <<def-sub-buffer,sub-buffers>>.
8181 [[def-trace-verb]]trace (_verb_)::
8182 The action of recording the <<def-event,events>> emitted by an
8183 application or by a system, or to initiate such recording by
8184 controlling a <<def-tracer,tracer>>.
8186 [[def-trace-chunk]]trace chunk::
8187 A self-contained <<def-trace,trace>> which is part of a
8188 <<def-tracing-session,tracing session>>. Each
8189 <<def-tracing-session-rotation, tracing session rotation>> produces a
8190 <<def-trace-chunk-archive,trace chunk archive>>.
8192 [[def-trace-chunk-archive]]trace chunk archive::
8193 The result of a <<def-tracing-session-rotation, tracing session rotation>>.
8195 <<def-lttng,LTTng>> doesn't manage any trace chunk archive, even if its
8196 containing <<def-tracing-session,tracing session>> is still active: you
8197 are free to read it, modify it, move it, or remove it.
8200 The http://tracecompass.org[Trace Compass] project and application.
8202 [[def-tracepoint]]tracepoint::
8203 An instrumentation point using the tracepoint mechanism of the Linux
8204 kernel or of <<def-lttng-ust,LTTng-UST>>.
8206 tracepoint definition::
8207 The definition of a single <<def-tracepoint,tracepoint>>.
8210 The name of a <<def-tracepoint,tracepoint>>.
8212 [[def-tracepoint-provider]]tracepoint provider::
8213 A set of functions providing <<def-tracepoint,tracepoints>> to an
8214 instrumented <<def-user-application,user application>>.
8216 Not to be confused with a <<def-tracepoint-provider-package,tracepoint
8217 provider package>>: many tracepoint providers can exist within a
8218 tracepoint provider package.
8220 [[def-tracepoint-provider-package]]tracepoint provider package::
8221 One or more <<def-tracepoint-provider,tracepoint providers>> compiled
8222 as an https://en.wikipedia.org/wiki/Object_file[object file] or as a
8223 link:https://en.wikipedia.org/wiki/Library_(computing)#Shared_libraries[shared
8226 [[def-tracer]]tracer::
8227 A software which records emitted <<def-event,events>>.
8229 <<domain,tracing domain>>::
8230 A namespace for <<def-event,event>> sources.
8232 <<tracing-group,tracing group>>::
8233 The Unix group in which a Unix user can be to be allowed to
8234 <<def-trace-verb,trace>> the Linux kernel.
8236 [[def-tracing-session]]<<tracing-session,tracing session>>::
8237 A stateful dialogue between you and a <<lttng-sessiond,session daemon>>.
8239 [[def-tracing-session-rotation]]<<session-rotation,tracing session rotation>>::
8240 The action of archiving the
8241 <<def-current-trace-chunk,current trace chunk>> of a
8242 <<def-tracing-session,tracing session>>.
8244 tracked <<def-proc-attr,process attribute>>::
8245 A process attribute which is part of an <<def-incl-set,inclusion
8248 untracked process attribute::
8249 A process attribute which isn't part of an <<def-incl-set,inclusion
8252 [[def-user-application]]user application::
8253 An application running in user space, as opposed to a Linux kernel
8254 module, for example.