Re: __builtin_string_constant_p

["Joseph S. Myers" <jsm at polyomino dot org dot uk>]

On Wed, 11 Aug 2004, Florian Weimer wrote:

> For some routines that should only accept string literals, it would be
> nice to have a __builtin_string_constant_p built-in.  For example,

Meaning what exactly, for such cases as

foo(complicated_expression ? "bar" : non_string)

where complicated_expression might or might not be optimised to 1?

Or indeed

inline int foo(const char *p)
{
  return __builtin_string_constant_p(p);
}

inline int bar(const char *p)
{
  return foo(p);
}

- is bar("baz") required to be 1?  What of bar(((("baz"))))?

(I've just written 1000 lines of formal models of constant expression 
rules for C90, C99 and GNU C, and the specification of 
__builtin_constant_p has the unsatisfactory property of depending on 
whether the compiler folds it at parse time (as e.g. in static 
initializers) or waits to see if it can be optimized later (e.g. constant 
parameters to inline functions).  In general a language specification that 
depends on what the compiler of the day can optimize is not ideal, and in 
most regards the formal models avoid such dependency.)

-- 
Joseph S. Myers               http://www.srcf.ucam.org/~jsm28/gcc/
    jsm@polyomino.org.uk (personal mail)
    jsm28@gcc.gnu.org (Bugzilla assignments and CCs)