Consider the following two snippets

char abc[] = "abc";
char xyz[] = "xyz";
char *p = abc;

printf("%c abc = %s\n", ++*p, abc);
p = xyz;
printf("%c xyz = %s\n", *p++, xyz);

char *abc = "abc";
char *xyz = "xyz";
char *p = abc;

printf("%c abc = %s\n", ++*p, abc);
p = xyz;
printf("%c xyz = %s\n", *p++, xyz);

What’s the difference between them? One results in a segfault and the other doesn’t. A little explanation about why. The statement char abc[] = "abc" creates a character array which is automatically null-terminated ‘\0’, while char *abc = "abc" creates a string literal and any attempt to modify it results in an undefined behaviour (which is in this case a segfault). Weirdly enough neither gcc’s built-in static analyzer -fanalyzer nor scan-build detects this. Relevant stackexchange answer (the accepted answer is not good). .