From: Zack Weinberg <zack AT wolery DOT cumb DOT org>
Date: Mon, 1 May 2000 00:09:28 -0700
To: Laurynas Biveinis <lauras AT softhome DOT net>
Cc: gcc-patches AT gcc DOT gnu DOT org, DJGPP Workers <djgpp-workers AT delorie DOT com>
Subject: Re: Minor DJGPP fixes
Message-ID: <20000501000928.J11780@wolery.cumb.org>
References: <390C769C DOT 812EAB7D AT softhome DOT net>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
User-Agent: Mutt/1.1.12i
In-Reply-To: <390C769C.812EAB7D@softhome.net>; from lauras@softhome.net on Sun, Apr 30, 2000 at 09:08:28PM +0300
Reply-To: djgpp-workers AT delorie DOT com
Errors-To: nobody AT delorie DOT com
X-Mailing-List: djgpp-workers AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk

On Sun, Apr 30, 2000 at 09:08:28PM +0300, Laurynas Biveinis wrote:

> 2) C preprocessor (read_include_file() in cppfiles.c) fails to do 
>    anything under DJGPP because DJGPP does not support 
>    fcntl(..., F_SETFL, ...).
>    I made a fix similar to code present in libgcc2.c, __bb_exit_func(),
>    with new target macro.

...
> 	* cppfiles.c (read_include_file): call fcntl(fd, F_SETFL, 0) only if
> 	macro TARGET_HAS_F_SETFL defined.

This is the wrong approach, because:

- fcntl() working is a characteristic of the _host_, not the target.
- You should determine if it works using an autoconf test.
- If it doesn't work, you need to open the file in blocking mode, not
  just elide the fcntl(); otherwise the code reading the file will
  break.

This fcntl() is there to avoid a problem that never happens in real
life: someone does #include </dev/rmt0> and the preprocessor gets
wedged because there's no tape.  Further, I think if someone actually
does include /dev/rmt0, we'll get stuck anyway, because the check
later on for a sensible file permits all character devices.

Therefore, I'd be willing to dump the fcntl() call entirely and open
the file in blocking mode on all hosts.  Does anyone else have an
opinion?

zw