README revision 1.2 
1$NetBSD: README,v 1.2 2009/09/14 17:15:32 apb Exp $
2
3NOTE: New tests should use the ATF framework; see atf(7)
4and the src/tests directory.
5
6------------------------------------------------------------------------
7This part of the source tree contains regression tests. There are special
8make targets and rules to follow. Most of these, however, are currently not
9enforced, and most tests available are not conforming.
10
11We hope to fix this someday. If you add new tests, please try to be conforming.
12
13What is a regression test?
14
15  A regression test is run by a makefile in a test directory (see below).
16  Each makefile may run multiple tests.
17
18What is a test directory?
19
20  A directory in this part of the tree is a regression test directory. It
21  contains a Makefile which implements the additional "regress" target,
22  and runs all it's regression tests during this target.
23
24May the make progress be stopped on failures?
25
26  No, the "make regress" target should succeed, unless some regression 
27  binaries could not be build, disk is full or other catastrophic failures
28  outside of the tested subsystem happen. A failing regression test should
29  log the failure (see below), but not make the target itself fail.
30
31What are the possible results of a regression test?
32
33  A test may either
34
35  - succeed, in which case it logs "PASSED" (see below for logging details)
36  - fail, in which case it logs "FAILED"
37  - not be able to run, in which case it logs "SKIPPED" and the reason
38    for the skip in the comment field (see below)
39
40  Typical reasons for tests to not being run are missing kernel options,
41  or missing privileges (test needs root, "make regress" is invoked by
42  mere mortal or vice versa). A test may not fail because of such
43  environmental issues, it must detect and properly log the problem.
44
45  If a test directory contains tests that may be skipped, it should have
46  a README file explaining the prerequisites (e.g. needed kernel options)
47
48  In future, we will mark affected makefiles and optimize run/skipped test
49  during repeated runs with differing privileges - but currently there is no
50  make framework in place to handle this.
51
52How and when does a test log results?
53
54  If the make/environment variable ${REGRESS_LOG} is defined, the final
55  results (and only those) should be logged to the file named by that
56  variable. We will, in the future, add make targets for this purpose.
57  The log format is line oriented, one line used per test. Each line
58  consists of the directory where the Makefile lives, followed by the
59  test name and the result (see above: PASSED, FAILED, SKIPPED).
60  Following this an optional comment may be added. For SKIPPED tests the
61  comment is not optional. Fields are separated by spaces.
62