_fpreset

Article
09/15/2006

Resets the floating-point package.

void_fpreset(void);

Function	Required Header	Compatibility
_fpreset	<float.h>	Win 95, Win NT

For additional compatibility information, see Compatibility in the Introduction.

Libraries

LIBC.LIB	Single thread static library, retail version
LIBCMT.LIB	Multithread static library, retail version
MSVCRT.LIB	Import library for MSVCRT.DLL, retail version

Return Value

None

Remarks

The _fpreset function reinitializes the floating-point math package. _fpreset is usually used with signal, system, or the _exec or _spawn functions. If a program traps floating-point error signals (SIGFPE) with signal, it can safely recover from floating-point errors by invoking _fpreset and using longjmp.

Example

/* FPRESET.C: This program uses signal to set up a
 * routine for handling floating-point errors.
 */

#include <stdio.h>
#include <signal.h>
#include <setjmp.h>
#include <stdlib.h>
#include <float.h>
#include <math.h>
#include <string.h>

#pragma warning(disable : 4113) /* C4113 warning expected */

jmp_buf mark;              /* Address for long jump to jump to */
int     fperr;             /* Global error number */

void __cdecl fphandler( int sig, int num );   /* Prototypes */
void fpcheck( void );

void main( void )
{
   double n1, n2, r;
   int jmpret;
   /* Unmask all floating-point exceptions. */
    _control87( 0, _MCW_EM );
   /* Set up floating-point error handler. The compiler
    * will generate a warning because it expects
    * signal-handling functions to take only one argument.
    */
    if( signal( SIGFPE, fphandler ) == SIG_ERR )

   {
      fprintf( stderr, "Couldn't set SIGFPE\n" );
      abort();   }

   /* Save stack environment for return in case of error. First
    * time through, jmpret is 0, so true conditional is executed.
    * If an error occurs, jmpret will be set to -1 and false
    * conditional will be executed.
    */
   jmpret = setjmp( mark );
   if( jmpret == 0 )
   {
      printf( "Test for invalid operation - " );
      printf( "enter two numbers: " );
      scanf( "%lf %lf", &n1, &n2 );
      r = n1 / n2;
      /* This won't be reached if error occurs. */
      printf( "\n\n%4.3g / %4.3g = %4.3g\n", n1, n2, r );

      r = n1 * n2;
      /* This won't be reached if error occurs. */
      printf( "\n\n%4.3g * %4.3g = %4.3g\n", n1, n2, r );
   }
   else
      fpcheck();
}
/* fphandler handles SIGFPE (floating-point error) interrupt. Note
 * that this prototype accepts two arguments and that the
 * prototype for signal in the run-time library expects a signal
 * handler to have only one argument.
 * The second argument in this signal handler allows processing of
 * _FPE_INVALID, _FPE_OVERFLOW, _FPE_UNDERFLOW, and
 * _FPE_ZERODIVIDE, all of which are Microsoft-specific symbols
 * that augment the information provided by SIGFPE. The compiler
 * will generate a warning, which is harmless and expected.

 */
void fphandler( int sig, int num )
{
   /* Set global for outside check since we don't want
    * to do I/O in the handler.
    */
   fperr = num;
   /* Initialize floating-point package. */
   _fpreset();
   /* Restore calling environment and jump back to setjmp. Return
    * -1 so that setjmp will return false for conditional test.
    */
   longjmp( mark, -1 );
}
void fpcheck( void )
{
   char fpstr[30];
   switch( fperr )
   {
   case _FPE_INVALID:
       strcpy( fpstr, "Invalid number" );
       break;
   case _FPE_OVERFLOW:
       strcpy( fpstr, "Overflow" );

       break;
   case _FPE_UNDERFLOW:
       strcpy( fpstr, "Underflow" );
       break;
   case _FPE_ZERODIVIDE:
       strcpy( fpstr, "Divide by zero" );
       break;
   default:
       strcpy( fpstr, "Other floating point error" );
       break;
   }
   printf( "Error %d: %s\n", fperr, fpstr );
}

Output

Test for invalid operation - enter two numbers: 5 0
Error 131: Divide by zero

Floating-Point Support Routines

_fpreset

Additional resources