Date: Wed, 3 May 2000 11:46:44 +0300 (IDT)
From: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
X-Sender: eliz AT is
To: djgpp AT delorie DOT com
Subject: Re: 3rd Try: Maybe an asm problem? (Problems linking)
In-Reply-To: <5jotgscf7s5houa8btrj60spmai2h4f8c5@4ax.com>
Message-ID: <Pine.SUN.3.91.1000503114453.4851E-100000@is>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Reply-To: djgpp AT delorie DOT com
Errors-To: nobody AT delorie DOT com
X-Mailing-List: djgpp AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk


On Tue, 2 May 2000, Damian Yerrick wrote:

> On Tue, 2 May 2000 13:16:07 +0300 (IDT), Eli Zaretskii
> <eliz AT is DOT elta DOT co DOT il> wrote:
> 
> >You are right.  Some library functions in DJGPP really do that for
> >you.  But if library functions don't do that, the application should.
> 
> The library should do it for all functions that block.

This is not possible with some of the functions, for which there's no
way of checking for a resource without blocking.

In addition, if an application wants to issue a low-level system call,
the library has no business interfering with it.

There are functions that by definition always wait for something.  These 
functions *should* call __dpmi_yield.  However, I think most, if not all,
such functions already do that.

So what you suggest is not always possible or desirable.  But if you
think that there are some library functions which should do this but
don't in the current version, please feel free to submit bug reports
and/or patches.