V632. Consider inspecting the NN argument of the 'Foo' function. It is odd that the argument is of the 'T' type.
The analyzer has detected a potential error: an odd argument is passed into a function. An argument having the format of a floating-point number has been passed into a function, although it was awaiting an integer type. It is incorrect because the argument value will be cast to an integer type.
Consider the following sample:
double buf[N]; ... memset(buf, 1.0, sizeof(buf));
The programmer intended to fill the array with values '1.0'. But this code will fill the array with garbage.
The second argument of the 'memset' function has an integer type. This argument defines the value to fill each byte of the array with.
Value '1.0' will be cast to the integer value '1'. The 'buf' data array will be filled byte-by-byte with "one" values. This result is different from what we get when filling each array item with value '1.0'.
This is the fixed code:
double buf[N]; ... for (size_t i = 0; i != N; ++i) buf[i] = 1.0;
This diagnostic is classified as: