my_err functions

[Look]

Hi,

In the following code, what is the difference between my_err1 and my_err2, if there is any.

Code: Select all

#include <stdlib.h>
#include <stdio.h>


void my_err1(const char* const message)
{
	perror(message);
	exit(EXIT_FAILURE);
}

void my_err2(const char* const message)
{
	fprintf(stderr,message);
	exit(EXIT_FAILURE);
}

int main(void)
{
    //my_err1("Oops\n");
    //my_err2("Oops\n");
	return EXIT_SUCCESS;
}

[brtzsnr]

man 3 perror

The routine perror() produces a message on the standard error output, describing the last error encountered during a call to a system or library function.

perror prints the errno when something like open, creat fails.

fprintf(stderr, "...") just prints to stderr.

[Look]

man 3 perror

The routine perror() produces a message on the standard error output, describing the last error encountered during a call to a system or library function.

perror prints the errno when something like open, creat fails.

fprintf(stderr, "...") just prints to stderr.

Why I get this nasty warning?

Code: Select all

#include <math.h>
#include <errno.h>
#include <stdio.h>
#include <stdlib.h>

int main(void)
{
    int x;
    int Y=0;
    errno = 0;
    x = sqrt(Y);
    if (errno) 
    {
	    perror("sqrt failed");
	    x = 0;
    }
    //x=0;
    return EXIT_SUCCESS;
}

Code: Select all

sh-4.3$ gcc -o main *.c -Wall -lm                                                                                                                                               
main.c: In function 'main':                                                                                                                                                     
main.c:8:9: warning: variable 'x' set but not used [-Wunused-but-set-variable]                                                                                                  
     int x;                                                                                                                                                                     
         ^    

[brtzsnr]

warning: variable 'x' set but not used [-Wunused-but-set-variable]

Because x is never used (read) only set (x = ... ).

[Look]

man 3 perror

The routine perror() produces a message on the standard error output, describing the last error encountered during a call to a system or library function.

perror prints the errno when something like open, creat fails.

fprintf(stderr, "...") just prints to stderr.

OK, now I understand it better:

Code: Select all

#include <math.h> 
#include <errno.h> 
#include <stdio.h> 
#include <stdlib.h> 

int main(void) 
{ 
    int i; 
    int Y=-1; 
    errno = 0; 
    i = sqrt(Y); 
    if (errno) 
    { 
       perror("sqrt failed"); 
       exit(EXIT_FAILURE);
       //i = 0; 
    } 
    //i++; 
    return EXIT_SUCCESS; 
}

http://www.tutorialspoint.com/compile_c_online.php
GCC output (without -Wall):

Code: Select all

sh-4.3$ gcc -o main *.c -lm                                                                                                                                                     
sh-4.3$ main                                                                                                                                                                    
sqrt failed: Numerical argument out of domain                                                                                                                                   
sh-4.3$  

[Look]

[...]

Code: Select all

#include <math.h> 
#include <errno.h> 
#include <stdio.h> 
#include <stdlib.h> 

int main(void) 
{ 
    int i; 
    int Y=-1; 
    errno = 0; 
    i = sqrt(Y); 
    if (errno) 
    { 
       perror("sqrt failed"); 
       exit(EXIT_FAILURE);
       //i = 0; 
    } 
    //i++; 
    return EXIT_SUCCESS; 
}

[...]

An improved version:

Code: Select all

#include <math.h> 
#include <errno.h> 
#include <stdio.h> 
#include <stdlib.h> 
#include <assert.h>

void my_err1(const char* const message) 
{ 
        assert(message!=NULL);
        perror(message); 
        exit(EXIT_FAILURE); 
} 

void my_err2(const char* const message) 
{ 
        assert(message!=NULL);
        fprintf(stderr,message);
        fprintf(stderr,"\n");
        exit(EXIT_FAILURE); 
} 

int main(void) 
{ 
    int i; 
    int Y=-1; 
    errno = 0; 
    i = sqrt(Y); 
    if (errno) 
    { 
        my_err1("sqrt failed");//my_err1(NULL); //my_err1(""); 
        //my_err2("sqrt failed");
    } 
    //i++; 
    return EXIT_SUCCESS; 
}

[Sven]

Code: Select all

void my_err2(const char* const message) 
{ 
        assert(message!=NULL);
        fprintf(stderr,message);
        fprintf(stderr,"\n");
        exit(EXIT_FAILURE); 
}

Even better, since printf-like functions should always get a format string as an argument (for safety reasons):

Code: Select all

void my_err2(const char* const message)
{
        assert(message!=NULL);
        fprintf(stderr, "%s\n", message);
        exit(EXIT_FAILURE);
}

[Look]

Code: Select all

void my_err2(const char* const message) 
{ 
        assert(message!=NULL);
        fprintf(stderr,message);
        fprintf(stderr,"\n");
        exit(EXIT_FAILURE); 
}

Even better, since printf-like functions should always get a format string as an argument (for safety reasons):

Code: Select all

void my_err2(const char* const message)
{
        assert(message!=NULL);
        fprintf(stderr, "%s\n", message);
        exit(EXIT_FAILURE);
}

Yes, I remembered about this rule. Why online compilers GCC (-Wall -Wextra -Wpedantic) or "Visual C++" (/W4) did not put a diagnostic for my questionable my_err2 function?