From owner-freebsd-current  Thu Feb 20 21:12:52 1997
Return-Path: <owner-current>
Received: (from root@localhost)
          by freefall.freebsd.org (8.8.5/8.8.5) id VAA02735
          for current-outgoing; Thu, 20 Feb 1997 21:12:52 -0800 (PST)
Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19])
          by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id VAA02730
          for <freebsd-current@freebsd.org>; Thu, 20 Feb 1997 21:12:50 -0800 (PST)
Received: (from bde@localhost) by godzilla.zeta.org.au (8.8.3/8.6.9) id QAA28404; Fri, 21 Feb 1997 16:06:31 +1100
Date: Fri, 21 Feb 1997 16:06:31 +1100
From: Bruce Evans <bde@zeta.org.au>
Message-Id: <199702210506.QAA28404@godzilla.zeta.org.au>
To: freebsd-current@freebsd.org, jmg@nike.efn.org
Subject: Re: bzero and FD_ZERO
Sender: owner-current@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

>today I was writing a program that uses FD_ZERO...  but I wasn't using
>string.h... if you compile it with -Wall you will get a warning about
>implicet declaration of bzero...
>
>should this happen?  and what is the proper fix for it?  (do we really
>require including string.h?  should I document that as a header that is
>required? )

The library should use a private or reserved function, e.g., __bzero()
or memset().  <string.h> shouldn't be included or a prerequisite.

NetBSD uses memset() except in the kernel.  The Gnu library uses
__FD_ZERO(), which on the i386 expands to i586-pessimized inline asm
code involving stosl.

There are many more bugs like this if you compile with all warning enabled:

	-Wall -ansi -pedantic \
	-Wbad-function-cast \
	-Wcast-align \
	-Wcast-qual \
	-Wchar-subscripts \
	-Wconversion \
	-Winline \
	-Wmissing-prototypes \
	-Wnested-externs \
	-Wpointer-arith \
	-Wredundant-decls \
	-Wshadow \
	-Wstrict-prototypes \
	-Wtraditional \
	-Wwrite-strings

Bruce