Message-ID:	<3958561A.E3D9736D@softhome.net>
Date:	Tue, 27 Jun 2000 09:22:02 +0200
From:	Laurynas Biveinis <lauras AT softhome DOT net>
X-Mailer:	Mozilla 4.73 [en] (Win98; U)
X-Accept-Language:	lt,en
MIME-Version:	1.0
To:	djgpp-workers AT delorie DOT com
Subject:	Re: Patch: chown() preparation for symlinks
References:	<395643BB DOT 7109B0DA AT softhome DOT net> <200006261453 DOT KAA11494 AT envy DOT delorie DOT com> <3957BF10 DOT 7EBB2B8D AT softhome DOT net> <200006262111 DOT RAA02148 AT envy DOT delorie DOT com>
Reply-To:	djgpp-workers AT delorie DOT com

DJ Delorie wrote:
> 
> > assumption wrong? Or is it general policy to #include <libc/stubs.h>
> > everywhere in libc sources to play safe?
> 
> I've got a program (or had, when I was working under MS-DOS) that
> checked libc.a for "cleanness".  It told me when I needed stubs.h and
> when I didn't.  If you can get that program working again (um,
> tests/libclink/check.cc and friends) then it will tell you when you
> need it and when you don't.

If I understand correctly, the following from its output means 
that libc/stubs.h is missing:

Impure stub chown.o(access) -> stub0000.o
Impure stub fsext.o(close) -> stub0002.o
Impure stub tempnam.o(access) -> stub0000.o
Impure stub malloc.o(sbrk) -> stub0031.o
Impure stub fwrite.o(isatty) -> stub0008.o

I've just included libc/stubs.h in chown.c and the first line of 
output disappeared. I'll investigate other problems mentioned there.

Laurynas

- Raw text -