Re: PATCH: PR binutils/3186: symbol `.debug_abbrev' required but not present

[Nick Clifton <nickc at redhat dot com>]

Hi Pedro,

> Why not have a standard macro for this?
>
> #define STR_STRLEN_PAIR(STR) \
>     STR, (sizeof (STR) - 1)
>
> Then these cases become either:
>
>         { STR_STRLEN_PAIR("debug")  },    /* 'd' */
>         { NULL,         0  },    /* 'e' */
>         { NULL,         0  },    /* 'f' */
>       { STR_STRLEN_PAIR("gnu.linkonce.wi.") },    /* 'g' */
>        { NULL,         0  },    /* 'h' */
>         { NULL,         0  },    /* 'i' */
>         { NULL,         0  },    /* 'j' */
>
> If you want to keep looking at the string for NULLness,
> or ...:
>
>         { STR_STRLEN_PAIR("debug") },    /* 'd' */
>         { STR_STRLEN_PAIR("") },    /* 'd' */
>         { STR_STRLEN_PAIR("") },    /* 'e' */
>         { STR_STRLEN_PAIR("")  },    /* 'f' */
>       { STR_STRLEN_PAIR("gnu.linkonce.wi.") },    /* 'g' */
>         { STR_STRLEN_PAIR("")  },    /* 'h' */
>         { STR_STRLEN_PAIR("")  },    /* 'i' */
>         { STR_STRLEN_PAIR("")  },    /* 'j' */
>
> ... if you just look at the len.
>
> Just a thought, I am not suggesting to go all over
> the code base to adapt to something like this, but we
> could make new code use something like this.

Code improvements that remove the potential for bugs are always welcome. 
 In this particular case you could always make the macro slightly more 
intelligent, so that it could handle the NULL pointer, eg:

  #define STR_STRLEN_PAIR(STR) \
   (STR), (STR) ? sizeof (STR) - 1 : 0

Cheers
  Nick