[Bug c++/58875] New: No float to int conversion warning in std::inner_product(x, x_end, y, 0)

[yangzhe1990 at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58875

            Bug ID: 58875
           Summary: No float to int conversion warning in
                    std::inner_product(x, x_end, y, 0)
           Product: gcc
           Version: 4.8.1
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: yangzhe1990 at gmail dot com

The correct code should be std::inner_product(x, x_end, y, 0.0), but with
-Wconversion enabled, there should be a float to int conversion warning.

Without the warning it's easy to write the wrong code and get the wrong answer.

For example, compiling

#include <algorithm>
#include <iostream>

int main() {
    float x[3] = {0.5, 0.5, 0.5};
    std::cout << std::inner_product(x, x + 3, x, 0) << std::endl;
    int y = 0;
    y += x[1];
    return 0;
}

generates the following warnings,

test.cpp: In function ‘int main()’:
test.cpp:8:4: warning: conversion to ‘float’ from ‘int’ may alter its value
[-Wconversion]
  y += x[1];
    ^
test.cpp:8:4: warning: conversion to ‘int’ from ‘float’ may alter its value
[-Wconversion]

But a warning to the inner_product line is expected.