3 Michael Stapelberg <michael@i3wm.org>
6 This document explains how the i3 testsuite works, how to use it and extend it.
7 It is targeted at developers who not necessarily have been doing testing before
8 or have not been testing in Perl before. In general, the testsuite is not of
9 interest for end users.
14 The i3 testsuite is a collection of files which contain testcases for various
15 i3 features. Some of them test if a certain workflow works correctly (moving
16 windows, focus behaviour, …). Others are regression tests and contain code
17 which previously made i3 crash or lead to unexpected behaviour. They then check
18 if i3 still runs (meaning it did not crash) and if it handled everything
21 The goal of having these tests is to automatically find problems and to
22 automatically get a feel for whether a change in the source code breaks any
23 existing feature. After every modification of the i3 sourcecode, the developer
24 should run the full testsuite. If one of the tests fails, the corresponding
25 problem should be fixed (or, in some cases, the testcase has to be modified).
26 For every bugreport, a testcase should be written to test the correct
27 behaviour. Initially, it will fail, but after fixing the bug, it will pass.
28 This ensures (or increases the chance) that bugs which have been fixed once
29 will never be found again.
31 Also, when implementing a new feature, a testcase might be a good way to be
32 able to easily test if the feature is working correctly. Many developers will
33 test manually if everything works. Having a testcase not only helps you with
34 that, but it will also be useful for every future change.
36 == Relevant documentation
38 Apart from this document, you should also have a look at:
40 1. The "Modern Perl" book, which can be found at
41 http://onyxneon.com/books/modern_perl/modern_perl_a4.pdf
42 2. The latest Perl documentation of the "i3test" (general testcase setup) and
43 "i3test::Test" (additional test instructions) modules:
44 http://build.i3wm.org/docs/lib-i3test.html respectively
45 http://build.i3wm.org/docs/lib-i3test-test.html
46 3. The latest documentation on i3’s IPC interface:
47 http://build.i3wm.org/docs/ipc.html
51 For several reasons, the i3 testsuite has been implemented in Perl:
53 1. Perl has a long tradition of testing. Every popular/bigger Perl module which
54 you can find on CPAN will not only come with documentation, but also with
55 tests. Therefore, the available infrastructure for tests is comprehensive.
56 See for example the excellent http://search.cpan.org/perldoc?Test::More
57 and the referenced http://search.cpan.org/perldoc?Test::Tutorial.
59 2. Perl is widely available and has a well-working package infrastructure.
60 3. The author is familiar with Perl :).
61 4. It is a good idea to use a different language for the tests than the
62 implementation itself.
64 Please do not start programming language flamewars at this point.
66 === Installing the dependencies
68 As usual with Perl programs, the testsuite ships with a +Makefile.PL+.
69 This file specifies which Perl modules the testsuite depends on and can be used
70 to install all of them.
72 Perl modules are distributed via CPAN, and there is the official, standard CPAN
73 client, simply called +cpan+. It comes with every Perl installation and can be
74 used to install the testsuite. Many users prefer to use the more modern
75 +cpanminus+ instead, though (because it asks no questions and just works):
77 The tests additionally require +Xephyr(1)+ to run a nested X server. Install
78 +xserver-xephyr+ on Debian or +xorg-xserver-xephyr+ on Arch Linux.
80 .Installing testsuite dependencies using cpanminus (preferred)
81 --------------------------------------------------------------------------------
83 $ sudo apt-get install cpanminus
85 --------------------------------------------------------------------------------
87 If you don’t want to use cpanminus for some reason, the same works with cpan:
89 .Installing testsuite dependencies using cpan
90 --------------------------------------------------------------------------------
93 --------------------------------------------------------------------------------
95 In case you don’t have root permissions, you can also install into your home
96 directory, see http://michael.stapelberg.de/cpan/
100 ==== Script: complete-run
102 The testcases are run by a script called +complete-run.pl+. It runs all
103 testcases by default, but you can be more specific and let it only run one or
104 more testcases. Also, it takes care of starting up a separate instance of i3
105 with an appropriate configuration file and creates a folder for each run
106 containing the appropriate i3 logfile for each testcase. The latest folder can
107 always be found under the symlink +latest/+. Unless told differently, it will
108 run the tests on a separate X server instance (using Xephyr).
110 Xephyr will open a window where you can inspect the running test. You can run
111 the tests without an X session with Xvfb, such as with +xvfb-run
112 ./complete-run+. This will also speed up the tests signficantly especially on
113 machines without a powerful video card.
115 .Example invocation of complete-run.pl+
116 ---------------------------------------
120 # output omitted because it is very long
121 All tests successful.
122 Files=78, Tests=734, 27 wallclock secs ( 0.38 usr 0.48 sys + 17.65 cusr 3.21 csys = 21.72 CPU)
125 $ ./complete-run.pl t/04-floating.t
126 [:3] i3 startup: took 0.07s, status = 1
127 [:3] Running t/04-floating.t with logfile testsuite-2011-09-24-16-06-04-4.0.2-226-g1eb011a/i3-log-for-04-floating.t
128 [:3] t/04-floating.t finished
130 output for t/04-floating.t:
131 ok 1 - use X11::XCB::Window;
132 ok 2 - The object isa X11::XCB::Window
133 ok 3 - Window is mapped
134 ok 4 - i3 raised the width to 75
135 ok 5 - i3 raised the height to 50
136 ok 6 - i3 did not map it to (0x0)
137 ok 7 - The object isa X11::XCB::Window
138 ok 8 - i3 let the width at 80
139 ok 9 - i3 let the height at 90
140 ok 10 - i3 mapped it to x=1
141 ok 11 - i3 mapped it to y=18
142 ok 12 - The object isa X11::XCB::Window
143 ok 13 - i3 let the width at 80
144 ok 14 - i3 let the height at 90
147 All tests successful.
148 Files=1, Tests=14, 0 wallclock secs ( 0.01 usr 0.00 sys + 0.19 cusr 0.03 csys = 0.23 CPU)
151 $ less latest/i3-log-for-04-floating.t
152 ----------------------------------------
154 If your attempt to run the tests with a bare call to ./complete-run.pl fails, try this:
156 ---------------------------------------------------
157 $ ./complete-run.pl --parallel=1 --keep-xserver-output
158 ---------------------------------------------------
160 This will show the output of Xephyr, which is the X server implementation we
163 ==== Coverage testing
165 Coverage testing is possible with +lcov+, the front-end for GCC's coverage
166 testing tool +gcov+. The testcases can generate a nice html report that tells
167 you which functions and lines were covered during a run of the tests. You can
168 use this tool to judge how effective your tests are.
170 To use test coverage tools, first compile with coverage enabled.
172 ---------------------------------------------------
174 ---------------------------------------------------
176 Then run the tests with the +--coverage-testing+ flag.
178 ---------------------------------------------------
179 ./complete-run.pl --coverage-testing
180 ---------------------------------------------------
182 Then open +latest/i3-coverage/index.html+ in your web browser.
186 The testsuite makes extensive use of the IPC (Inter-Process Communication)
187 interface which i3 provides. It is used for the startup process of i3, for
188 terminating it cleanly and (most importantly) for modifying and getting the
189 current state (layout tree).
191 See [http://i3wm.org/docs/ipc.html] for documentation on the IPC interface.
195 In order to open new windows, change attributes, get events, etc., the
196 testsuite uses X11::XCB, a new (and quite specific to i3 at the moment) Perl
197 module which uses the XCB protocol description to generate Perl bindings to
198 X11. They work in a very similar way to libxcb (which i3 uses) and provide
199 relatively high-level interfaces (objects such as +X11::XCB::Window+) aswell as
200 access to the low-level interface, which is very useful when testing a window
203 === Filesystem structure
205 In the git root of i3, the testcases live in the folder +testcases+. This
206 folder contains the +complete-run.pl+ and a base configuration file which will
207 be used for the tests. The different testcases (their file extension is .t, not
208 .pl) themselves can be found in the conventionally named subfolder +t+:
210 .Filesystem structure
211 --------------------------------------------
213 │ ├── complete-run.pl
217 │ │ ├── SocketActivation.pm
218 │ │ └── StartXDummy.pm
222 │ │ ├── 02-fullscreen.t
224 │ │ ├── omitted for brevity
226 │ │ └── 74-regress-focus-toggle.t
227 --------------------------------------------
229 == Anatomy of a testcase
231 Learning by example is definitely a good strategy when you are wondering how to
232 write a testcase. Let's take +t/11-goto.t+ as an easy example and go through it
235 .t/11-goto.t: Boilerplate
236 ----------------------
238 # vim:ts=4:sw=4:expandtab
243 my $x = X11::XCB::Connection->new;
244 -----------------------
246 This is what we call boilerplate. It exists at the top of every test file (to
247 some extent). The first line is the shebang, which specifies that this file is
248 a Perl script. The second line contains VIM specific settings on how to
249 edit/format this file (use spaces instead of tabs, indent using 4 spaces).
250 Afterwards, the +i3test+ module is used. This module contains i3 testsuite
251 specific functions which you are strongly encouraged to use. They make writing
252 testcases a lot easier and will make it easier for other people to read your
255 The next line uses the +File::Temp+ module. This is specific to this testcase,
256 because it needs to generate a temporary name during the test. Many testcases
257 use only the +i3test+ module.
259 The last line opens a connection to X11. You might or might not need this in
260 your testcase, depending on whether you are going to open windows (etc.) or
261 only use i3 commands.
264 ----------------------
265 my $tmp = fresh_workspace;
268 ----------------------
270 The first line calls i3test's +fresh_workspace+ function which looks for a
271 currently unused workspace, switches to it, and returns its name. The variable
272 +$tmp+ will end up having a value such as +"/tmp/87kBVcHbA9"+. Note that this
273 is not (necessarily) a valid path, it's just a random workspace name.
275 So, now that we are on a new workspace, we ensure that the workspace uses
276 horizontal orientation by issuing the +split h+ command (see the i3 User's
277 Guide for a list of commands). This is not strictly necessary, but good style.
278 In general, the +cmd+ function executes the specified i3 command by using the
279 IPC interface and returns once i3 acknowledged the command.
282 ----------------------
283 #####################################################################
284 # Create two windows and make sure focus switching works
285 #####################################################################
287 my $top = open_window($x);
288 my $mid = open_window($x);
289 my $bottom = open_window($x);
290 ----------------------
292 In every major section of a testcase, you should put a comment like the one
293 above. This makes it immediately clear how the file is structured.
295 The +open_window+ function opens a standard window, which will then be put into
296 tiling mode by i3. If you want a floating window, use the
297 +open_floating_window+ function. These functions accept the same parameters as
298 +X11::XCB::Window->new+, see the i3test documentation at TODO.
300 .t/11-goto.t: Helper function
301 ----------------------
303 # Returns the input focus after sending the given command to i3 via IPC
304 # and syncing with i3
311 return $x->input_focus;
313 ----------------------
315 This section defines a helper function which will be used over and over in this
316 testcase. If you have code which gets executed more than once or twice
317 (depending on the length of your test, use your best judgement), please put it
318 in a function. Tests should be short, concise and clear.
320 The +focus_after+ function executes a command and returns the X11 focus after
321 the command was executed. The +sync_with_i3+ command makes sure that i3 could
322 push its state to X11. See <<i3_sync>> to learn how this works exactly.
324 .t/11-goto.t: Test assumptions
325 ----------------------
326 $focus = $x->input_focus;
327 is($focus, $bottom->id, "Latest window focused");
329 $focus = focus_after('focus left');
330 is($focus, $mid->id, "Middle window focused");
331 ----------------------
333 Now, we run the first two real tests. They use +Test::More+'s +is+ function,
334 which compares two values and prints the differences if they are not the same.
335 After the arguments, we supply a short comment to indicate what we are testing
336 here. This makes it vastly more easy for the developer to spot which testcase
337 is the problem in case one fails.
339 The first test checks that the most recently opened window is focused.
340 Afterwards, the command +focus left+ is issued and it is verified that the
341 middle window now has focus.
343 Note that this is not a comprehensive test of the +focus+ command -- we would
344 have to test wrapping, focus when using a more complex layout, focusing the
345 parent/child containers, etc. But that is not the point of this testcase.
346 Instead, we just want to know if +$x->input_focus+ corresponds with what we are
347 expecting. If not, something is completely wrong with the test environment and
348 this trivial test will fail.
350 .t/11-goto.t: Test that the feature does not work (yet)
351 ----------------------
352 #####################################################################
353 # Now goto a mark which does not exist
354 #####################################################################
356 my $random_mark = mktemp('mark.XXXXXX');
358 $focus = focus_after(qq|[con_mark="$random_mark"] focus|);
359 is($focus, $mid->id, "focus unchanged");
360 ----------------------
362 Syntax hint: The qq keyword is the interpolating quote operator. It lets you
363 chose a quote character (in this case the +|+ character, a pipe). This makes
364 having double quotes in our string easy.
366 In this new major section, a random mark (mark is an identifier for a window,
367 see "VIM-like marks" in the i3 User’s Guide) will be generated. Afterwards, we
368 test that trying to focus that mark will not do anything. This is important: Do
369 not only test that using a feature has the expected outcome, but also test that
370 using it without properly initializing it does no harm. This command could for
371 example have changed focus anyways (a bug) or crash i3 (obviously a bug).
373 .t/11-goto.t: Test that the feature does work
374 ----------------------
375 cmd "mark $random_mark";
377 $focus = focus_after('focus left');
378 is($focus, $top->id, "Top window focused");
380 $focus = focus_after(qq|[con_mark="$random_mark"] focus|);
381 is($focus, $mid->id, "goto worked");
382 ----------------------
384 Remember: Focus was on the middle window (we verified that earlier in "Test
385 assumptions"). We now mark the middle window with our randomly generated mark.
386 Afterwards, we switch focus away from the middle window to be able to tell if
387 focusing it via its mark will work. If the test works, the goto command seems
390 .t/11-goto.t: Test corner case
391 ----------------------
392 # check that we can specify multiple criteria
394 $focus = focus_after('focus left');
395 is($focus, $top->id, "Top window focused");
397 $focus = focus_after(qq|[con_mark="$random_mark" con_mark="$random_mark"] focus|);
398 is($focus, $mid->id, "goto worked");
399 ----------------------
401 Now we test the same feature, but specifying the mark twice in the command.
402 This should have no effect, but let’s be sure: test it and see if things go
405 .t/11-goto.t: Test second code path
406 ----------------------
407 #####################################################################
408 # Check whether the focus command will switch to a different
409 # workspace if necessary
410 #####################################################################
412 my $tmp2 = fresh_workspace;
414 is(focused_ws(), $tmp2, 'tmp2 now focused');
416 cmd qq|[con_mark="$random_mark"] focus|;
418 is(focused_ws(), $tmp, 'tmp now focused');
419 ----------------------
421 This part of the test checks that focusing windows by mark works across
422 workspaces. It uses i3test's +focused_ws+ function to get the current
425 .t/11-goto.t: Test second code path
426 ----------------------
428 ----------------------
430 The end of every testcase has to contain the +done_testing+ line. This tells
431 +complete-run.pl+ that the test was finished successfully. If it does not
432 occur, the test might have crashed during execution -- some of the reasons why
433 that could happen are bugs in the used modules, bugs in the testcase itself or
434 an i3 crash resulting in the testcase being unable to communicate with i3 via
438 == Appendix A: The i3 sync protocol
440 Consider the following situation: You open two windows in your testcase, then
441 you use +focus left+ and want to verify that the X11 focus has been updated
442 properly. Sounds simple, right? Let’s assume you use this straight-forward
445 .Racey focus testcase
447 my $left = open_window($x);
448 my $right = open_window($x);
450 is($x->input_focus, $left->id, 'left window focused');
453 However, the test fails. Sometimes. Apparantly, there is a race condition in
454 your test. If you think about it, this is because you are using two different
455 pieces of software: You tell i3 to update focus, i3 confirms that, and then you
456 ask X11 to give you the current focus. There is a certain time i3 needs to
457 update the X11 state. If the testcase gets CPU time before X11 processed i3's
458 requests, the test will fail.
460 image::i3-sync.png["Diagram of the race condition", title="Diagram of the race condition"]
462 One way to "solve" this would be to add +sleep 0.5;+ after the +cmd+ call.
463 After 0.5 seconds it should be safe to assume that focus has been updated,
466 In practice, this usually works. However, it has several problems:
468 1. This is obviously not a clean solution, but a workaround. Ugly.
469 2. On very slow machines, this might not work. Unlikely, but in different
470 situations (a delay to wait for i3 to startup) the necessary time is much
471 harder to guess, even for fast machines.
472 3. This *wastes a lot of time*. Usually, your computer is much faster than 0.5s
473 to update the status. However, sometimes, it might take 0.4s, so we can’t
476 To illustrate how grave the problem with wasting time actually is: Before
477 removing all sleeps from the testsuite, a typical run using 4 separate X
478 servers took around 50 seconds on my machine. After removing all the sleeps,
479 we achieved times of about 25 seconds. This is very significant and influences
480 the way you think about tests -- the faster they are, the more likely you are
481 to check whether everything still works quite often (which you should).
483 What I am trying to say is: Delays adds up quickly and make the test suite
486 The real solution for this problem is a mechanism which I call "the i3 sync
487 protocol". The idea is to send a request (which does not modify state) via X11
488 to i3 which will then be answered. Due to the request's position in the event
489 queue (*after* all previous events), you can be sure that by the time you
490 receive the reply, all other events have been dealt with by i3 (and, more
493 image::i3-sync-working.png["Diagram of the i3 sync solution", title="Diagram of the i3 sync solution"]
495 === Implementation details
497 The client which wants to sync with i3 initiates the protocol by sending a
498 ClientMessage to the X11 root window:
502 # Generate a ClientMessage, see xcb_client_message_t
503 my $msg = pack "CCSLLLLLLL",
504 CLIENT_MESSAGE, # response_type
507 $root, # destination window
508 $x->atom(name => 'I3_SYNC')->id,
510 $_sync_window->id, # data[0]: our own window id
511 $myrnd, # data[1]: a random value to identify the request
516 # Send it to the root window -- since i3 uses the SubstructureRedirect
517 # event mask, it will get the ClientMessage.
518 $x->send_event(0, $root, EVENT_MASK_SUBSTRUCTURE_REDIRECT, $msg);
521 i3 will then reply with the same ClientMessage, sent to the window specified in
522 +data[0]+. In the reply, +data[0]+ and +data[1]+ are exactly the same as in the
523 request. You should use a random value in +data[1]+ and check that you received
524 the same one when getting the reply.
526 == Appendix B: Socket activation
528 Socket activation is a mechanism which was made popular by systemd, an init
529 replacement. It basically describes creating a listening socket before starting
530 a program. systemd will invoke the program only when an actual connection to
531 the socket is made, hence the term socket activation.
533 The interesting part of this (in the i3 context) is that you can very precisely
534 detect when the program is ready (finished its initialization).
536 === Preparing the listening socket
538 +complete-run.pl+ will create a listening UNIX socket which it will then pass
539 to i3. This socket will be used by i3 as an additional IPC socket, just like
540 the one it will create on its own. Passing the socket happens implicitly
541 because children will inherit the parent’s sockets when fork()ing and sockets
542 will continue to exist after an exec() call (unless CLOEXEC is set of course).
544 The only explicit things +complete-run.pl+ has to do is setting the +LISTEN_FDS+
545 environment variable to the number of sockets which exist (1 in our case) and
546 setting the +LISTEN_PID+ environment variable to the current process ID. Both
547 variables are necessary so that the program (i3) knows how many sockets it
548 should use and if the environment variable is actually intended for it. i3 will
549 then start looking for sockets at file descriptor 3 (since 0, 1 and 2 are used
550 for stdin, stdout and stderr, respectively).
552 The actual Perl code which sets up the socket, fork()s, makes sure the socket
553 has file descriptor 3 and sets up the environment variables follows (shortened
557 .Setup socket and environment
558 -----------------------------
559 my $socket = IO::Socket::UNIX->new(
561 Local => $args{unix_socket_path},
566 $ENV{LISTEN_PID} = $$;
567 $ENV{LISTEN_FDS} = 1;
569 # Only pass file descriptors 0 (stdin), 1 (stdout),
570 # 2 (stderr) and 3 (socket) to the child.
573 # If the socket does not use file descriptor 3 by chance
574 # already, we close fd 3 and dup2() the socket to 3.
575 if (fileno($socket) != 3) {
577 POSIX::dup2(fileno($socket), 3);
582 -----------------------------
584 === Waiting for a reply
586 In the parent process, we want to know when i3 is ready to answer our IPC
587 requests and handle our windows. Therefore, after forking, we immediately close
588 the listening socket (i3 will handle this side of the socket) and connect to it
589 (remember, we are talking about a named UNIX socket) as a client. This connect
590 call will immediately succeed because the kernel buffers it. Then, we send a
591 request (of type GET_TREE, but that is not really relevant). Writing data to
592 the socket will also succeed immediately because, again, the kernel buffers it
593 (only up to a certain amount of data of course).
595 Afterwards, we just blockingly wait until we get an answer. In the child
596 process, i3 will setup the listening socket in its event loop. Immediately
597 after actually starting the event loop, it will notice a new client connecting
598 (the parent process) and handle its request. Since all initialization has been
599 completed successfully by the time the event loop is entered, we can now assume
602 === Timing and conclusion
604 A beautiful feature of this mechanism is that it does not depend on timing. It
605 does not matter when the child process gets CPU time or when the parent process
606 gets CPU time. On heavily loaded machines (or machines with multiple CPUs,
607 cores or unreliable schedulers), this makes waiting for i3 much more robust.
609 Before using socket activation, we typically used a +sleep(1)+ and hoped that
610 i3 was initialized by that time. Of course, this breaks on some (slow)
611 computers and wastes a lot of time on faster computers. By using socket
612 activation, we decreased the total amount of time necessary to run all tests
613 (72 files at the time of writing) from > 100 seconds to 16 seconds. This makes
614 it significantly more attractive to run the test suite more often (or at all)
617 An alternative approach to using socket activation is polling for the existance
618 of the IPC socket and connecting to it. While this might be slightly easier to
619 implement, it wastes CPU time and is considerably uglier than this solution
620 :). After all, +lib/SocketActivation.pm+ contains only 54 SLOC.