Re: stop using P_, __P in header files

[Dan Nicolaescu <dann@gnu.org>]

Juanma Barranquero <lekktu@gmail.com> writes:

> On Sun, Jul 4, 2010 at 10:08, Dan Nicolaescu <dann@gnu.org> wrote:
>
>> I've done most of this.
>
> Thanks. I've fixed a few prototypes that break compilation on Windows
> or generate warnings.

Thanks!

> A few questions:
>
> 1) This one and similar ones on unexelf.c seem errors:
>
> -round_up (x, y)
> -     ElfW(Addr) x, y;
> +round_up (Elf32_Addr x, Elf32_Addr y)
>
> The prototypes should presumably still use ElfW(Addr), ElfW(Ehdr),
> etc., shouldn't they?

Yes, they should.  I did not convert the code by hand, but by using
protoize (which runs the preprocessor) so this is why such prototypes got changed.
I caught a few similar ones and worked around them, but it looks like
some similar things got away.


> 2) In strftme.c you've converted just some functions, not all, and
> removed the LOCALE_PARAM* macros.
>
>  static CHAR_T *
> -memcpy_lowcase (dest, src, len LOCALE_PARAM)
> -     CHAR_T *dest;
> -     const CHAR_T *src;
> -     size_t len;
> -     LOCALE_PARAM_DECL
> +memcpy_lowcase (char *dest, const char *src, size_t len)
>
> Apparently, there haven't been any changes from gnulib merged back
> into strftime.c in the past seven years (since 2003-06-24). Should the
> LOCALE_PARAM* macros still be used (and so, re-added to
> memcpy_(low|upp)case), or would it be better to just get rid of that
> cruft?

I'd say we should resync with gnulib.
protoize made this change, so it was not intentional...

> 3) After this change in print.c
>
>  static void
> -strout (ptr, size, size_byte, printcharfun, multibyte)
> -     char *ptr;
> -     int size, size_byte;
> -     Lisp_Object printcharfun;
> -     int multibyte;
> +strout (char *ptr, int size, int size_byte, Lisp_Object printcharfun,
> int multibyte)
>  {
>
> compilation throws this warning:
>
> print.c: In function 'print_object':
> print.c:1973: warning: passing argument 1 of 'strout' discards
> qualifiers from pointer target type
> print.c:353: note: expected 'char *' but argument is of type 'const char *'
>
> becase of the call
>
> 	  strout (XSUBR (obj)->symbol_name, -1, -1, printcharfun, 0);
>
> AFAICS, the object pointed to by PTR is never modified, so it seems
> safe to use const char *, and I've done so.

That seems TRTD.