Newbie Question about Long Double in Mingw

[Fringe]

I'm using Mingw 3.2 (Windows version of GCC) which supposedly supports 10bit
floating point numbers using Long Double. However, Long Doubles don't print
correctly even in a simple program like the following:

int main()
{
long double P = 1.123456789012345678901234L;
printf("%Lf\n", P);


system("PAUSE");
return 0;
}

Output: -0.000000

Do I need to write my own library functions in order to do operations on
Long Doubles in Mingw? Thanks

 

[Ben Pfaff]

I'm using Mingw 3.2 (Windows version of GCC) which supposedly supports 10bit
floating point numbers using Long Double. However, Long Doubles don't print
correctly even in a simple program like the following:

Questions about compilers are best asked in newsgroups that
discuss those compilers.

 

[CBFalconer]

I'm using Mingw 3.2 (Windows version of GCC) which supposedly supports 10bit
floating point numbers using Long Double. However, Long Doubles don't print
correctly even in a simple program like the following:

10 bit floating point seems pretty useless to me.

int main()
{
long double P = 1.123456789012345678901234L;
printf("%Lf\n", P);

system("PAUSE");
return 0;
}

Output: -0.000000

This might work better if it was a legal C program. #include
<stdio.h>, unload that silly system call, and specify the decimals
wanted in the printf call.

c:\c\junk>cat junk.c
#include <stdio.h>
int main(void)
{
long double P = 1.123456789012345678901234L;

printf("%.24Lf\n", P);
return 0;
}

c:\c\junk>.\a
1.123456789012345678887769

 

[jacob navia]

Use lcc-win32:

int main()
{
long double P = 1.123456789012345678901234L;
printf("%18.17Lf\n", P);
system("PAUSE");
return 0;
}

D:\lcc\mc52\test>tld
1.12345678901234568

http://www.cs.virginia.edu/~lcc-win32

P.S. lcc-win32 offers standard 350 bit "qfloats"

 

[jacob navia]

Mingw doesn't support long doubles since it uses CRTDLL.DLL.
That library doesn't support long doubles.