程序的非正常跳转

 

我们一般在程序中执行的跳转可以通过goto、while、for、if等来实现。但是这些跳转都局限在函数内部，也就是说是一种“non-local goto”，不能在执行环境中进行任意的跳转。幸运的是绝大多数编译器还支持另外两个函数：setjmp和longjmp。原型如下：

 

int setjmp( jmp_bufenv );

void longjmp( jmp_bufenv, int value );

setjmp保存当前执行环境下的栈。根据具体的情况，函数的返回值有两种情况。第一种情况：函数保存当前执行环境下的栈时，其返回值为0。第二种情况：函数的返回值由longjmp的第二个参数决定。然而，如果longjmp的第二个参数为0时，函数的返回值是1。

longjmp用setjmp保存下来的栈，恢复程序在执行到setjmp时的状态，然后程序的执行路径直接跳到setjmp之后。注意：setjmp任然被执行，但是返回值将由longjmp的第二个参数来决定。

下面的代码是摘录自MSDN的一个例子程序，简要的说明了这两个函数的用法。

/* FPRESET.C: This program uses signal to setup 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>

 

jmp_buf mark;              /* Address for long jump to jumpto */

int    fperr;             /* Global errornumber */

 

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

void fpcheck( void );

 

void main( void )

{

       doublen1, n2, r;

       intjmpret;

 

       /*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-pointerror) 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, 0 );

}

 

 

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 );

}