Date: Tue, 21 Apr 1998 12:07:27 +1000 From: Bruce Evans <bde@zeta.org.au> To: bde@zeta.org.au, brian@Awfulhak.org Cc: brian@FreeBSD.ORG, cvs-committers@FreeBSD.ORG Subject: Re: cvs commit: src/usr.sbin/ppp bundle.c fsm.c id.c ipcp.c lcp.c log.h lqr.c modem.c prompt.h sig.c slcompress.c timer.c Message-ID: <199804210207.MAA31545@godzilla.zeta.org.au>
next in thread | raw e-mail | index | archive | help
>> Ifdefing the use of __attribute__ is unnecessary because <sys/cdefs.h> >> already defines away __attribute__ (#if __GNUC__ < 2). >... >So should <sys/cdefs.h> be included from my .h files ? It seems >prone to error if I expect the .c to include it before the .h with >__attribute__. See style(9), although its points about cdefs.h are out of date. Almost all CSRG sources depend on it defining __P(), but not all of them include one of the 2 headers that are documented to include it (<sys/types.h> and (indirectly) <sys/param.h>). In practice, it is guaranteed to be included by almost every header in /usr/include, because application headers that declare a prototype should include it (for C++ support if not for __P()). However, all of this is unportable. I now like your original version better :-). Bruce To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199804210207.MAA31545>