Fix functions/values used for multiply x2 test

[jeanlucf22]

No description provided.

[jmohdyusof]

Should like 37 and 39 use the same format? I have no opinion on which is preferred (but happy to learn)

[jeanlucf22]

Should like 37 and 39 use the same format? I have no opinion on which is preferred (but happy to learn)

The 'f' is for float only, not double (if I understand it right)

[jmohdyusof]

So is 2e-6 interpreted as a double?

[jeanlucf22]

I believe so, that is why you do not see a warning for that one

[jmohdyusof]

This seems to be correct. I would prefer something like 2e-6f (but I am not sure it is allowed) just because I dislike having to count zeros

[nicolasbock]

With gcc-12.2.0 I get the following:

#include <stdlib.h>
#include <stdio.h>

int main() {
    printf("sizeof() = %lu\n", sizeof(1e-6f));
}

$ gcc test.c && ./a.out 
sizeof() = 4

So it is possible to write the literal in e notation (for @jmohdyusof 😄) and also force it to be float.

[nicolasbock]

I forgot to add that I would also prefer the 'e' notation. Counting zeros is error prone.

[jmohdyusof]

Assuming this does not fail with any compilers we need, I would advocate this change, and approve otherwise.

[jmohdyusof]

Approved, will test on Intel and Darwin