🚀 go-pugleaf

I believe I have found a syntax checking bug in gcc 2.2.2.
That is, if my understanding about prototypes is right...

This bug is an ANSI-ism and appears to be the converse of the "char
prototype" point that you make in the Bugs section of gcc.info.

My understanding is that if you declare a function in the "old" style,
but also specify prototypes, you must widen the parameter declarations
in the prototypes to "int" from char or short.

However, gcc -Wall, says the following prototype doesn't match the
declaration in the following program:

---- start
typedef unsigned short  Dimension;

int foo(unsigned int);

int foo(x)
	Dimension x;
{
	return x + 1;
}
---- end

If I remove the word "unsigned" from the foo prototype, gcc compiles
this program happily, however, my understanding is that "unsigned"
can't just widen to a signed type.

Just to be specific, here's the output of gcc -v:

Reading specs from /usr/local/lib/gcc/mips-dec-ultrix4.2/2.2.2/specs
gcc version 2.2.2

dave cohrs