Date: Sun, 28 Sep 1997 18:47:14 +0200 (IST)
From: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
To: Oberhumer Markus <k3040e4 AT c210 DOT edvz DOT uni-linz DOT ac DOT at>
cc: dj AT delorie DOT com, djgpp-workers AT delorie DOT com, malcolm AT manawatu DOT gen DOT nz,
        Charles Sandmann <sandmann AT clio DOT rice DOT edu>
Subject: Re: [malcolm AT manawatu DOT gen DOT nz: Fork source code.]
In-Reply-To: <199709261425.RAA26936@is.elta.co.il>
Message-ID: <Pine.SUN.3.91.970928184638.423O-100000@is>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Precedence: bulk


On Fri, 26 Sep 1997, Oberhumer Markus wrote:

> >	* Charles once told me that there are many bugs and subtleties
> >	  in the way different DPMI hosts implement functions 0900h
> >	  and 0901h.  This code uses these heavily and seems to rely
> >	  on the fact that no interrupt will arrive when the virtual
> >	  interrupts are disabled.  Will this assumtion hold?  What,
> >	  if any, are other implications, for the case of this code,
> >	  of whatever problems there are in the different
> >	  implementations of 0900h out there?
> 
> You can easily lock at least Windows 3.1 with some calls to 0900/0901.
> 
> I'm always using the following macros in my programs:
> #define disable() __asm__ __volatile__("cli \n");
> #define enable()  __asm__ __volatile__("sti ; cld \n");

Charles, what do you think about this?  

(My message quoted by Thomas refers to an implementation of `fork'
that creates multiple threads in the same program by using a
user-defined software signal generated on each timer tick.  It was
submitted to DJ for inclusion in v2.02.)