Path: utzoo!attcan!ncrcan!ziebmef!ross
From: ross@ziebmef.uucp (Ross Ridge)
Newsgroups: comp.sys.hp
Subject: Re: HP-UX type declaration problem with signal(2)
Message-ID: <1989Jul7.193846.3308@ziebmef.uucp>
Date: 7 Jul 89 23:38:45 GMT
References: <218@bnrunix.UUCP> <5570224@hpfcdc.HP.COM>
Reply-To: ross@ziebmef.UUCP (Ross Ridge)
Organization: Ziebmef Public Access Unix, Toronto, Ontario
Lines: 56

In article <5570224@hpfcdc.HP.COM> rer@hpfcdc.HP.COM (Rob Robason) writes:
>> upon looking at <signal.h> and comparing the 6.2 vs:  the 6.5, I find
>> this difference...
>
>> 6.2---
>> extern  (*signal())();
>> 6.5---
>> extern  void (*signal())();
>> 
>> Should not 6.5 have a statement more like 6.2?????
>
>This is a result of the support of ANSI C, which redefines the return
>value of signal to be a pointer to void, which is an ANSICism for a
>"generic" pointer.

This is not correct. ANSI C (and newer Sys V releases) defines signal
as returning a pointer to a function returning void (nothing). This
is not the "generic" pointer to void (void *). 

The old way to use signal was like this:

{
	int (*old_handler)();
	int int_handler();

	old_hanlder = signal(SIGINT, int_handler);
}

The new way is:

{
	void (*old_handler)(/* int */);
	void int_handler(/* int */);

	old_handler = signal(SIGINT, int_handler);
}

That's all that has changed. If you're worried about compability do something
like:

#ifdef NEW_SIGNAL

typedef signal_type void

#else

typdef signal_type int

#endif

								Ross Ridge
-- 
 l/                   attcan!tmsoft!ziebmef!ross			 //
[OO]                   					                [oo]
-()-                                                                    -()-
 db                     6502 assembly forever!                           //