Diagnose decimal integer constants too large for intmax_t

[Jakub Jelinek]

On Sat, Apr 25, 2009 at 06:46:20PM +0000, Joseph S. Myers wrote:
> --- libcpp/expr.c	(revision 146747)
> +++ libcpp/expr.c	(working copy)
> @@ -538,8 +538,27 @@ cpp_interpret_integer (cpp_reader *pfile
>  		    && pfile->state.in_directive)
>  	       && !num_positive (result, precision))
>  	{
> +	  /* This is for constants within the range of uintmax_t but
> +	     not that or intmax_t.  For such decimal constants, a

s/or/of/ ?

	Jakub