1.. SPDX-License-Identifier: GPL-2.0
2
3===================================================
4The Kernel Test Anything Protocol (KTAP), version 1
5===================================================
6
7TAP, or the Test Anything Protocol is a format for specifying test results used
8by a number of projects. It's website and specification are found at this `link
9<https://testanything.org/>`_. The Linux Kernel largely uses TAP output for test
10results. However, Kernel testing frameworks have special needs for test results
11which don't align with the original TAP specification. Thus, a "Kernel TAP"
12(KTAP) format is specified to extend and alter TAP to support these use-cases.
13This specification describes the generally accepted format of KTAP as it is
14currently used in the kernel.
15
16KTAP test results describe a series of tests (which may be nested: i.e., test
17can have subtests), each of which can contain both diagnostic data -- e.g., log
18lines -- and a final result. The test structure and results are
19machine-readable, whereas the diagnostic data is unstructured and is there to
20aid human debugging.
21
22KTAP output is built from four different types of lines:
23- Version lines
24- Plan lines
25- Test case result lines
26- Diagnostic lines
27
28In general, valid KTAP output should also form valid TAP output, but some
29information, in particular nested test results, may be lost. Also note that
30there is a stagnant draft specification for TAP14, KTAP diverges from this in
31a couple of places (notably the "Subtest" header), which are described where
32relevant later in this document.
33
34Version lines
35-------------
36
37All KTAP-formatted results begin with a "version line" which specifies which
38version of the (K)TAP standard the result is compliant with.
39
40For example:
41- "KTAP version 1"
42- "TAP version 13"
43- "TAP version 14"
44
45Note that, in KTAP, subtests also begin with a version line, which denotes the
46start of the nested test results. This differs from TAP14, which uses a
47separate "Subtest" line.
48
49While, going forward, "KTAP version 1" should be used by compliant tests, it
50is expected that most parsers and other tooling will accept the other versions
51listed here for compatibility with existing tests and frameworks.
52
53Plan lines
54----------
55
56A test plan provides the number of tests (or subtests) in the KTAP output.
57
58Plan lines must follow the format of "1..N" where N is the number of tests or subtests.
59Plan lines follow version lines to indicate the number of nested tests.
60
61While there are cases where the number of tests is not known in advance -- in
62which case the test plan may be omitted -- it is strongly recommended one is
63present where possible.
64
65Test case result lines
66----------------------
67
68Test case result lines indicate the final status of a test.
69They are required and must have the format:
70
71.. code-block:: none
72
73	<result> <number> [<description>][ # [<directive>] [<diagnostic data>]]
74
75The result can be either "ok", which indicates the test case passed,
76or "not ok", which indicates that the test case failed.
77
78<number> represents the number of the test being performed. The first test must
79have the number 1 and the number then must increase by 1 for each additional
80subtest within the same test at the same nesting level.
81
82The description is a description of the test, generally the name of
83the test, and can be any string of words (can't include #). The
84description is optional, but recommended.
85
86The directive and any diagnostic data is optional. If either are present, they
87must follow a hash sign, "#".
88
89A directive is a keyword that indicates a different outcome for a test other
90than passed and failed. The directive is optional, and consists of a single
91keyword preceding the diagnostic data. In the event that a parser encounters
92a directive it doesn't support, it should fall back to the "ok" / "not ok"
93result.
94
95Currently accepted directives are:
96
97- "SKIP", which indicates a test was skipped (note the result of the test case
98  result line can be either "ok" or "not ok" if the SKIP directive is used)
99- "TODO", which indicates that a test is not expected to pass at the moment,
100  e.g. because the feature it is testing is known to be broken. While this
101  directive is inherited from TAP, its use in the kernel is discouraged.
102- "XFAIL", which indicates that a test is expected to fail. This is similar
103  to "TODO", above, and is used by some kselftest tests.
104- “TIMEOUT”, which indicates a test has timed out (note the result of the test
105  case result line should be “not ok” if the TIMEOUT directive is used)
106- “ERROR”, which indicates that the execution of a test has failed due to a
107  specific error that is included in the diagnostic data. (note the result of
108  the test case result line should be “not ok” if the ERROR directive is used)
109
110The diagnostic data is a plain-text field which contains any additional details
111about why this result was produced. This is typically an error message for ERROR
112or failed tests, or a description of missing dependencies for a SKIP result.
113
114The diagnostic data field is optional, and results which have neither a
115directive nor any diagnostic data do not need to include the "#" field
116separator.
117
118Example result lines include::
119
120	ok 1 test_case_name
121
122The test "test_case_name" passed.
123
124::
125
126	not ok 1 test_case_name
127
128The test "test_case_name" failed.
129
130::
131
132	ok 1 test # SKIP necessary dependency unavailable
133
134The test "test" was SKIPPED with the diagnostic message "necessary dependency
135unavailable".
136
137::
138
139	not ok 1 test # TIMEOUT 30 seconds
140
141The test "test" timed out, with diagnostic data "30 seconds".
142
143::
144
145	ok 5 check return code # rcode=0
146
147The test "check return code" passed, with additional diagnostic data “rcode=0”
148
149
150Diagnostic lines
151----------------
152
153If tests wish to output any further information, they should do so using
154"diagnostic lines". Diagnostic lines are optional, freeform text, and are
155often used to describe what is being tested and any intermediate results in
156more detail than the final result and diagnostic data line provides.
157
158Diagnostic lines are formatted as "# <diagnostic_description>", where the
159description can be any string.  Diagnostic lines can be anywhere in the test
160output. As a rule, diagnostic lines regarding a test are directly before the
161test result line for that test.
162
163Note that most tools will treat unknown lines (see below) as diagnostic lines,
164even if they do not start with a "#": this is to capture any other useful
165kernel output which may help debug the test. It is nevertheless recommended
166that tests always prefix any diagnostic output they have with a "#" character.
167
168Unknown lines
169-------------
170
171There may be lines within KTAP output that do not follow the format of one of
172the four formats for lines described above. This is allowed, however, they will
173not influence the status of the tests.
174
175This is an important difference from TAP.  Kernel tests may print messages
176to the system console or a log file.  Both of these destinations may contain
177messages either from unrelated kernel or userspace activity, or kernel
178messages from non-test code that is invoked by the test.  The kernel code
179invoked by the test likely is not aware that a test is in progress and
180thus can not print the message as a diagnostic message.
181
182Nested tests
183------------
184
185In KTAP, tests can be nested. This is done by having a test include within its
186output an entire set of KTAP-formatted results. This can be used to categorize
187and group related tests, or to split out different results from the same test.
188
189The "parent" test's result should consist of all of its subtests' results,
190starting with another KTAP version line and test plan, and end with the overall
191result. If one of the subtests fail, for example, the parent test should also
192fail.
193
194Additionally, all lines in a subtest should be indented. One level of
195indentation is two spaces: "  ". The indentation should begin at the version
196line and should end before the parent test's result line.
197
198"Unknown lines" are not considered to be lines in a subtest and thus are
199allowed to be either indented or not indented.
200
201An example of a test with two nested subtests:
202
203::
204
205	KTAP version 1
206	1..1
207	  KTAP version 1
208	  1..2
209	  ok 1 test_1
210	  not ok 2 test_2
211	# example failed
212	not ok 1 example
213
214An example format with multiple levels of nested testing:
215
216::
217
218	KTAP version 1
219	1..2
220	  KTAP version 1
221	  1..2
222	    KTAP version 1
223	    1..2
224	    not ok 1 test_1
225	    ok 2 test_2
226	  not ok 1 test_3
227	  ok 2 test_4 # SKIP
228	not ok 1 example_test_1
229	ok 2 example_test_2
230
231
232Major differences between TAP and KTAP
233--------------------------------------
234
235==================================================   =========  ===============
236Feature                                              TAP        KTAP
237==================================================   =========  ===============
238yaml and json in diagnosic message                   ok         not recommended
239TODO directive                                       ok         not recognized
240allows an arbitrary number of tests to be nested     no         yes
241"Unknown lines" are in category of "Anything else"   yes        no
242"Unknown lines" are                                  incorrect  allowed
243==================================================   =========  ===============
244
245The TAP14 specification does permit nested tests, but instead of using another
246nested version line, uses a line of the form
247"Subtest: <name>" where <name> is the name of the parent test.
248
249Example KTAP output
250--------------------
251::
252
253	KTAP version 1
254	1..1
255	  KTAP version 1
256	  1..3
257	    KTAP version 1
258	    1..1
259	    # test_1: initializing test_1
260	    ok 1 test_1
261	  ok 1 example_test_1
262	    KTAP version 1
263	    1..2
264	    ok 1 test_1 # SKIP test_1 skipped
265	    ok 2 test_2
266	  ok 2 example_test_2
267	    KTAP version 1
268	    1..3
269	    ok 1 test_1
270	    # test_2: FAIL
271	    not ok 2 test_2
272	    ok 3 test_3 # SKIP test_3 skipped
273	  not ok 3 example_test_3
274	not ok 1 main_test
275
276This output defines the following hierarchy:
277
278A single test called "main_test", which fails, and has three subtests:
279- "example_test_1", which passes, and has one subtest:
280
281   - "test_1", which passes, and outputs the diagnostic message "test_1: initializing test_1"
282
283- "example_test_2", which passes, and has two subtests:
284
285   - "test_1", which is skipped, with the explanation "test_1 skipped"
286   - "test_2", which passes
287
288- "example_test_3", which fails, and has three subtests
289
290   - "test_1", which passes
291   - "test_2", which outputs the diagnostic line "test_2: FAIL", and fails.
292   - "test_3", which is skipped with the explanation "test_3 skipped"
293
294Note that the individual subtests with the same names do not conflict, as they
295are found in different parent tests. This output also exhibits some sensible
296rules for "bubbling up" test results: a test fails if any of its subtests fail.
297Skipped tests do not affect the result of the parent test (though it often
298makes sense for a test to be marked skipped if _all_ of its subtests have been
299skipped).
300
301See also:
302---------
303
304- The TAP specification:
305  https://testanything.org/tap-version-13-specification.html
306- The (stagnant) TAP version 14 specification:
307  https://github.com/TestAnything/Specification/blob/tap-14-specification/specification.md
308- The kselftest documentation:
309  Documentation/dev-tools/kselftest.rst
310- The KUnit documentation:
311  Documentation/dev-tools/kunit/index.rst
312