1 README for libm-test math test suite
2 ====================================
4 The libm-test math test suite tests a number of function points of
5 math functions in the GNU C library. The following sections contain a
6 brief overview. Please note that the test drivers and the Python
7 script "gen-libm-test.py" have some options. A full list of options
8 is available with --help (for the test drivers) and -h for
14 The tests just evaluate the functions at specified points and compare
15 the results with precomputed values and the requirements of the ISO
18 Besides testing the special values mandated by IEEE 754 (infinities,
19 NaNs and minus zero), some more or less random values are tested.
21 Files that are part of libm-test
22 ================================
24 The main files are "libm-test-<func>.inc". They are independent of
25 the target platform and the specific real floating type and format and
26 contain placeholder test "templates" for math functions defined in
27 libm. These files, along with generated files named
28 "auto-libm-test-out-<func>", are preprocessed by the Python script
29 "gen-libm-test.py" to expand the templates and produce a set of test
30 cases for each math function that are specific to the target platform
31 but still independent of the real floating type. The results of the
32 processing are "libm-test-<func>.c" and a file "libm-test-ulps.h" with
33 platform specific deltas by which the actual math function results may
34 deviate from the expected results and still be considered correct.
36 The test drivers "test-double-<func>.c", "test-float-<func>.c", and
37 "test-ldouble-<func>.c", generated by the Makefile, test the normal
38 double, float and long double implementation of libm. Each driver
39 selects the desired real floating type to exercise the math functions
40 to test with (float, double, or long double) by defining a small set
41 of macros just before including the generic "libm-test.c" file. Each
42 driver is compiled into a single executable test program with the
45 As mentioned above, the "gen-libm-test.py" script looks for a file
46 named "libm-test-ulps" in the platform specific sysdep directory (or
47 its fpu or nofpu subdirectory) and for each variant (real floating
48 type and rounding mode) of every tested function reads from it the
49 maximum difference expressed as Units of Least Precision (ULP) the
50 actual result of the function may deviate from the expected result
51 before it's considered incorrect.
53 The "auto-libm-test-out-<func>" files contain sets of test cases to
54 exercise, the conditions under which to exercise each, and the
55 expected results. The files are generated by the
56 "gen-auto-libm-tests" program from the "auto-libm-test-in" file. See
57 the comments in gen-auto-libm-tests.c for details about the content
58 and format of the -in and -out files.
60 How can I generate "libm-test-ulps"?
61 ====================================
63 To automatically generate a new "libm-test-ulps" run "make regen-ulps".
64 This generates the file "math/NewUlps" in the build directory. The file
65 contains the sorted results of all the tests. You can use the "NewUlps"
66 file as the machine's updated "libm-test-ulps" file. Copy "NewUlps" to
67 "libm-test-ulps" in the appropriate machine sysdep directory. Verify
68 the changes, post your patch, and check it in after review.
70 To manually generate a new "libm-test-ulps" file, first remove "ULPs"
71 file in the current directory, then you can execute for example:
72 ./testrun.sh math/test-double -u --ignore-max-ulp=yes
73 This generates a file "ULPs" with all double ULPs in it, ignoring any
74 previously calculated ULPs, and running with the newly built dynamic
75 loader and math library (assumes you didn't install your build). Now
76 generate the ULPs for all other formats, the tests will be appending the
77 data to the "ULPs" file. As final step run "gen-libm-test.py" with the
78 file as input and ask to generate a pretty printed output in the file
80 gen-libm-test.py -u ULPs -n NewUlps
81 Copy "NewUlps" to "libm-test-ulps" in the appropriate machine sysdep
84 Note that the test drivers have an option "-u" to output an unsorted
85 list of all epsilons that the functions have. The output can be read
86 in directly but it's better to pretty print it first.
87 "gen-libm-test.py" has an option to generate a pretty-printed and
88 sorted new ULPs file from the output of the test drivers.
90 Contents of libm-test-ulps
91 ==========================
93 Since libm-test-ulps can be generated automatically, just a few notes.
94 The file contains lines for maximal errors of single functions, like:
99 The keywords are float, double, and ldouble.
101 Adding tests to libm-test-<func>.inc
102 ====================================
104 The tests are evaluated by a set of special test macros. The macros
105 start with "TEST_" followed by a specification the input values, an
106 underscore and a specification of the output values. As an example,
107 the test macro for a function with input of type FLOAT (FLOAT is
108 either float, double, long double) and output of type FLOAT is
109 "TEST_f_f". The macro's parameter are the name of the function, the
110 input parameter, output parameter and optionally one exception
113 The accepted parameter types are:
115 - "j" for long double.
116 - "a" for ARG_FLOAT, the argument type for narrowing functions.
117 - "b" for boolean - just tests if the output parameter evaluates to 0
118 or 1 (only for output).
119 - "c" for complex. This parameter needs two values, first the real,
120 then the imaginary part.
123 - "L" for long long int.
124 - "u" for unsigned int.
127 - "p" for an argument (described in the previous character) passed
128 through a pointer rather than directly.
129 - "F" for the address of a FLOAT (only as input parameter)
130 - "I" for the address of an int (only as input parameter)
131 - "1" for an additional output (either output through a pointer passed
132 as an argument, or to a global variable such as signgam).
134 How to read the test output
135 ===========================
137 Running each test on its own at the default level of verbosity will
138 print on stdout a line describing the implementation of math functions
139 exercised by the test (float, double, or long double). This is then
140 followed by the details of test failures (if any). The output concludes
141 by a summary listing the number of test cases exercised and the number
142 of test failures uncovered.
144 For each test failure (and for each test case at higher levels of
145 verbosity), the output contains the name of the function under test
146 and its arguments or conditions that triggered the failure. Note
147 that the name of the function in the output need not correspond
148 exactly to the name of the math function actually invoked. For example,
149 the output will refer to the "acos" function even if the actual function
150 under test is acosf (for the float version) or acosl (for the long
151 double version). Also note that the function arguments may be shown
152 in either the decimal or the hexadecimal floating point format which
153 may or may not correspond to the format used in the auto-libm-test-in
154 file. Besides the name of the function, for each test failure the
155 output contains the actual and expected results and the difference
156 between the two, printed in both the decimal and hexadecimal
157 floating point format, and the ULP and maximum ULP for the test