Re: [Sdcc-user] -mpic14 bug, test, patch

[Steve T. <te...@te...>]

On Fri, 4 Apr 2003, UNIX savvy Brian Witt wrote:

> > I've just submitted another one, this time through sourceforge (#715085).
> > This patch+regression gets a lot of function-pointers working:
> 
> I have never seen syntax like that to call through a function pointer.
> Here are the two forms I have seen....
> 
> 
> 	pfunc();		// Compiler figures out what to do.
> 	(*pfunc)();		// Explicit about the ptr deref.
 
Quite right; the last is the most accepted form.
(Note to self: always reread K&Rv2 before criticizing a C compiler's
language acceptance...)

All three forms are accepted correctly by SDCC for
pointer-to-function-returning-void. But only the last form is accepted for
pointer-to-function-returning-char, as in:
	char c;
 	char (*pfunc)();
 	char call0(void) { }
 	...
 	pfunc = call0;
 	c = (pfunc)();  /* error */
 	c = pfunc();  /* error */
 	c = (*pfunc)();  /* works */

Gcc accepts 'em all.  But that's far from a good  test of language
conformance.

> If there is a regression test, please make sure you include these
> two other forms, please.
 
Since I botched attaching the the patch last night, I get to tweak the
regression and regenerate it.  It now includes all six forms, but the two
causing an error are commented.

Steve


--
Steve Tell  te...@te...