Re: [PATCH] MIPS: Transform old-style macros to newer "__noreturn" standard.

["Robert P. J. Day" <rpjday@mindspring.com>]

On Fri, 25 May 2007, H. Peter Anvin wrote:

> Robert P. J. Day wrote:
> > Convert old/obsolete NORET_TYPE and ATTRIB_NORET macros to use the
> > newer standard of "__noreturn" as defined in compiler-gcc.h.
> >
> > Signed-off-by: Robert P. J. Day <rpjday@mindspring.com>
>
> > 1) in a function declaration, the "__noreturn" will go at the end of
> > the declaration.
> >
> > 2) in a definition, "__noreturn" will go between the return type and
> > the function name
> >
> > 3) in a function typedef, "__noreturn" will go immediately after the
> > return type, just like with definitions.
> >
> > 4) if a function definition already includes "__noreturn", there's no
> > point in having any external references to it also say the same thing.
> > (right?)
>
> This is dumb, though.
>
> "void __noreturn" is redundant.  It would be much cleaner to have a
> macro which amounts to "void __attribute__((noreturn))" and use it
> instead of giving a return type.
>
> Even "void" as the return type is bogus -- the function never
> returns so it doesn't *have* a return type...

i agree, and i just did a quick test and noticed that, if you try to
declare a function thusly:

  f() __attribute__((noreturn)) ;

you get:

  warning: data definition has no type or storage class

but gcc doesn't complain if you declare it thusly:

  __attribute__((noreturn)) f() ;

that strikes me as a flaw in gcc, no?

rday
-- 
========================================================================
Robert P. J. Day
Linux Consulting, Training and Annoying Kernel Pedantry
Waterloo, Ontario, CANADA

http://fsdev.net/wiki/index.php?title=Main_Page
========================================================================