Date: Wed, 26 Aug 1998 23:02:31 -0700 From: John Polstra <jdp@polstra.com> To: dmm125@bellatlantic.net Cc: bugs@FreeBSD.ORG Subject: Re: error in libcompat (2.2.7) Message-ID: <199808270602.XAA08550@austin.polstra.com> In-Reply-To: <Pine.NEB.3.96.980826235640.21704A-100000@myname.my.domain> References: <Pine.NEB.3.96.980826235640.21704A-100000@myname.my.domain>
next in thread | previous in thread | raw e-mail | index | archive | help
In article <Pine.NEB.3.96.980826235640.21704A-100000@myname.my.domain>,
Donn Miller <dmm125@bellatlantic.net> wrote:
> Could someone running the stock 2.2.7 check out /usr/src/lib/libcompat?
>
> If you cd /usr/src/lib/libcompat, and make, see if SysV gets made, which
> includes ftok.c. e.g. do make |& tee make.log and then grep ftok make.log
In 2.2.7, the SysV subdirectory of libcompat doesn't even exist. I
eliminated it in revision 1.1.6.1 of ftok.c. At the same time, I
moved that file into libc. If you still have a SysV directory, then
you probably forgot the "-P" flag when doing a cvs update.
> Also, in SysV/ftok.c, ftok is defined as
>
> key_t ftok(const char *path, char id) but /usr/include/sys/ipc.h declares
> ftok as
>
> ftok(const char *, int);
>
> so basically in the source file the second arg is a char, but the header
> file (ipc.h) declares the second arg to be an int.
True, that's a bug. I don't think it has any actual effect, since
the function definition has the old pre-ANSI form:
key_t
ftok(path, id)
const char *path;
char id;
which means that "id" is assumed to be passed as an int, even though
only the low-order byte of it is used. But it should be fixed.
John
--
John Polstra jdp@polstra.com
John D. Polstra & Co., Inc. Seattle, Washington USA
"Self-knowledge is always bad news." -- John Barth
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-bugs" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199808270602.XAA08550>