Debian Codegraph @

clang-tools-extra/docs/clang-tidy/checks/bugprone/incorrect-roundings.rst

raw
1 .. title:: clang-tidy - bugprone-incorrect-roundings
2
3 bugprone-incorrect-roundings
4 ============================
5
6 Checks the usage of patterns known to produce incorrect rounding.
7 Programmers often use::
8
9 (int)(double_expression + 0.5)
10
11 to round the double expression to an integer. The problem with this:
12
13 1. It is unnecessarily slow.
14 2. It is incorrect. The number 0.499999975 (smallest representable float
15 number below 0.5) rounds to 1.0. Even worse behavior for negative
16 numbers where both -0.5f and -1.4f both round to 0.0.