[Mingw-users] MinGW missing _controlfp and related defines? Looks like not...

[Luis L. <lui...@gm...>]

Hello,

I'm compiling a C program that sets floating point control word:

<test.c>
#include <float.h>

int main(void)
{
	/* set FPU precision to 64-bit and round to nearest */
	_controlfp( _PC_64|_RC_NEAR, _MCW_PC|_MCW_RC );

	return 0;
}
</test.c>

Which results in the following:
V:\>gcc test.c -o test.exe
test.c: In function 'main':
test.c:6:14: error: '_PC_64' undeclared (first use in this function)
test.c:6:14: note: each undeclared identifier is reported only once
for each function it appears in
test.c:6:21: error: '_RC_NEAR' undeclared (first use in this function)
test.c:6:31: error: '_MCW_PC' undeclared (first use in this function)
test.c:6:39: error: '_MCW_RC' undeclared (first use in this function)

Here is defined in MSDN:
http://msdn.microsoft.com/en-us/library/e9b52ceh(v=vs.71).aspx

But peeking into float.h seems that those elements are only defined if
__STRICT_ANSI__ is not defined...

Why is __STRICT_ANSI__ defined even if I haven't requested it?

Looking at GCC documentation.

http://gcc.gnu.org/onlinedocs/gcc/C-Dialect-Options.html

Says that __STRICT_ANSI__ is defined when -ansi is used. Also mentions
that non-ISO compliant are not rejected gratuitously unless -pedantic
option is provided (which wasn't on this case).

GCC version in use is 4.6.2, mingwrt is 3.20 and w32api is 3.17-2.

Thanks in advance and pardon my ignorance.
-- 
Luis Lavena
AREA 17
-
Perfection in design is achieved not when there is nothing more to add,
but rather when there is nothing more to take away.
Antoine de Saint-Exupéry