1 Things to do Automated Testing Framework
2 ===========================================================================
5 Last revised: November 30th, 2010
8 This document includes the list of things that need to be done in ATF that
9 are most requested by the users. This information used to be available in
10 an ad-hoc bug tracker but that proved to be a bad idea. I have collected
11 all worthy comments in here.
13 Please note that most work these days is going into Kyua (see
14 http://code.google.com/p/kyua/). The ideas listed here apply to the
15 components of ATF that have *not* been migrated to the new codebase yet.
16 For bug reports or ideas that apply to the components that already have
17 been migrated, please use the bug tracker in the URL above. Similarly,
18 whenever a component is migrated, the ideas in this file should be revised
19 and migrated to the new bug tracker where appropriate.
22 ---------------------------------------------------------------------------
23 Add build-time checks to atf-sh
25 The 0.7 release introduced build-time tests to atf-c and atf-c++, but not
26 to atf-sh. Expose the functionality to the shell interface.
28 This will probably require writing an atf-build utility that exposes the C
29 code and can be called from the shell.
31 ---------------------------------------------------------------------------
32 Revisit what to do when an Atffile lists a non-existent file
34 ---------------------------------------------------------------------------
35 Add ATF_CHECK* versions to atf-c++ to support non-fatal tests
37 ---------------------------------------------------------------------------
38 Implement race-condition tests
42 I would think that stress/negative tests would be of more value than race
43 condition tests (they're similar, but not exactly the same in my mind).
47 1. Feed through as much data as possible to determine where reporting
49 2. Feed through data quickly and terminate ASAP. The data should be
50 captured. Terminate child applications with unexpected exit codes and
51 signals (in particular, SIGCHLD, SIGPIPE, exit codes that terminate,
53 3. Open up a file descriptor in the test application, don't close the file
55 4. fork(2) a process; don't wait(2) for the application to complete.
57 There are other scenarios that could be exercised, but these are the ones
58 I could think of off the topic of my head.
64 1. The thing is: how do you express any of this in a portable/abstract
65 interface? How do you express that a test case "receives data"? What
66 does that exactly mean? I don't think the framework should care about
67 this: each test should be free to decide where its data is and how to
72 3. Not sure I understand your request, but testing for "unexpected exit
73 codes" is already supported. See wiki:DesignXFail for the feature
76 4. What's the problem with this case? The test case exits right away after
77 terminating the execution of its body; any open file descriptors,
78 leaked memory, etc. die with it.
80 5. forking and not waiting for a subprocess was a problem already
83 I kinda have an idea of what Antti means with "race condition tests", but
84 every time I have tried to describe my understanding of matters I seem to
85 be wrong. Would be nice to have a clear description of what this involves;
86 in particular, what are the expectations from the framework and how should
87 the feature be exposed.
89 As of now, what I understand by "race condition test" is: a test case that
90 exercises a race condition. The test case may finish without triggering
91 the race, in which case it just exists with a successful status.
92 Otherwise, if the race triggers, the test case gets stuck and times out.
93 The result should be reported as an "expected failure" different from
100 Yup. Plus some atf-wide mechanism for the operator to supply some kind of
101 guideline if the test should try to trigger the race for a second or for
108 Alright. While mocking up some code for this, I think that your two
109 requests are complementary.
111 On the one hand, when you are talking about a "race condition" test you
112 really mean an "expected race condition" test. Correct? If so, we need to
113 extend the xfail mechanism to add one more case, which is to report any
114 failures as a race condition error and, if there is no failure, report the
117 On the other hand, the atf-wide mechanism to support how long the test
118 should run for can be thought as a "stress test" mechanism. I.e. run this
119 test for X time / iterations and report its results regularly without
120 involving xfail at all.
122 So, with this in mind:
124 * For a test that triggers an unfixed race condition, you set xfail to
125 race mode and define the test as a stress test. Any failures are
126 reported as expected failures.
128 * For a test that verifies a supposedly-fixed race condition, you do *not*
129 set xfail to race mode, and only set the test to stress test. Any
130 failures are reported as real failures.
132 These stress test cases implement a single iteration of the test and
133 atf-run is in charge of running the test several times, stopping on the
136 Does that make sense?
138 ---------------------------------------------------------------------------
139 Implement ATF_REQUIRE_ERRNO
143 Most of the lines in tests against system functionality are:
145 if (syscall(args) == -1)
146 atf_tc_fail_errno("flop")
148 Some shorthand would be helpful, like ATF_REQUIRE_ERRNO(syscall(args))
149 Also, a variant which allows arbitrary return value checks (e.g. "!= 0" or
150 "< 124" or "!= size") would be nice.
156 There's a problem with this request; not all functions fail in the same
157 way ... in particular compare the pthread family of functions (which
158 return errno) vs many native syscalls. Furthermore, compare some
159 fcntl-like syscalls vs other syscalls. One size fits all solutions may not
160 be a wise idea in this case, so I think that the problem statement needs
161 to be better defined, because the above request is too loose.
163 FWIW, there's also a TEST macro in LTP, which tests for non-zero status,
164 and sets an appropriate set of global variables for errnos and return
165 codes, respectively. It was a good idea, but has been mostly abandoned
166 because it's too difficult to define a success and failure in a universal
167 manner, so I think that we need to be careful with what's implemented in
168 ATF to not repeat the mistakes that others have made.
174 I think you've got a good point.
176 This was mostly intended to simplify the handling of the stupid errno
177 global variable. I think this is valuable to have, but maybe the
178 macro/function name should be different because _ERRNO can be confusing.
179 Probably something like an ATF_CHECK_LIBC / ATF_CHECK_PTHREAD approach
180 would be more flexible and simple.
183 ===========================================================================
184 vim: filetype=text:textwidth=75:expandtab:shiftwidth=2:softtabstop=2