Date: Wed, 2 Dec 1998 10:55:17 +0200 (IST)
From: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
X-Sender: eliz AT is
To: "Salvador Eduardo Tropea (SET)" <salvador AT inti DOT gov DOT ar>
cc: djgpp-workers AT delorie DOT com
Subject: Re: bug in stat?
In-Reply-To: <m0zktQM-000S60C@inti.gov.ar>
Message-ID: <Pine.SUN.3.91.981202105450.24125N-100000@is>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Reply-To: djgpp-workers AT delorie DOT com


On Tue, 1 Dec 1998, Salvador Eduardo Tropea (SET) wrote:

> /dev/prn -> v2.01 and v2.02: is a device
> prn -> v2.02: is a device
> /foo/prn -> both says that's a normal file but DOS
> understands it as a device.
> 
>   The questions are:
>   Is just me? (I mixed stat from v2.02 and the rest of 2.01)
>   Is that easy to solve? I didn't tried because I know Eli
> knows all the guts of stat and I guess he can find the
> solution much more quickly and cleanly.

I would gladly help, but I don't see the problem: "/foo/prn" works
for me in both patched 2.01 libc and in the latest 2.02 beta.  (I
don't have stock 2.01 to try.)

Maybe you mixed versions in a way that broke something?  You need to
use the entire libc.a and the include directory from v2.02, to get
consistent results (use -I and -L switches to gcc to force it to
search the correct directories).  Some parts of this puzzle are solved
in places other that stat.c proper (I think _truename and _put_path
have something to do with how the devices are handled).