[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: [vincent@vinc17.org: Re: mpfr 2.0.3: bugfix for test suite]



On Wed, 2004-10-27 at 20:17, Michael Fedrowitz wrote:
> On Wed, 2004-10-27 at 12:56 -0400, Steve M. Robbins wrote:
> > The M68k buildd machine fails the test suite (see below for log
> > URL) and it appears that the underlying problem is that the machine
> > doesn't support IEEE-754 floating point.
> 
> This might be the old excess precision problem.
> 
> > Is there a mechanism (compiler args?) to enable IEEE-754?  Or is
> > it not possible on this platform?
> 
> Try -ffloat-store.

And/or set the precision yourself: there are definitions in
<fpu_control.h> to change the precision (the same definitions exist on
the x86):

/* precision control */
#define _FPU_EXTENDED 0x00   /* RECOMMENDED */
#define _FPU_DOUBLE   0x80
#define _FPU_SINGLE   0x40     /* DO NOT USE */

/* Type of the control word.  */
typedef unsigned int fpu_control_t __attribute__ ((__mode__ (__SI__)));
 
/* Macros for accessing the hardware control word.  */
#define _FPU_GETCW(cw) __asm__ ("fmove%.l %!, %0" : "=dm" (cw))
#define _FPU_SETCW(cw) __asm__ volatile ("fmove%.l %0, %!" : : "dm" (cw))
 
/* Default control word set at startup.  */
extern fpu_control_t __fpu_control;


Kind regards,

Kars.




Reply to: