From: "Charles Sandmann" <sandmann AT clio DOT rice DOT edu>
Newsgroups: comp.os.msdos.djgpp
Subject: Re: Accesing from 0xFFFF:F000 to 0xFFFF:FFFF physical memory address
Date: Fri, 13 Oct 2000 10:52:29
Organization: Aspen Technology, Inc.
Lines: 13
Message-ID: <39e6e96d.sandmann@clio.rice.edu>
References: <39E5E00A DOT 39A80E46 AT gtwparc DOT padova DOT ccr DOT it> <7263-Thu12Oct2000193356+0300-eliz AT is DOT elta DOT co DOT il> <39E6C1EB DOT 58AED5F3 AT gtwparc DOT padova DOT ccr DOT it> <8011-Fri13Oct2000132503+0300-eliz AT is DOT elta DOT co DOT il> <39E70A59 DOT 1ABE9F69 AT gtwparc DOT padova DOT ccr DOT it> <39E72636 DOT B47FE5E6 AT gtwparc DOT padova DOT ccr DOT it>
NNTP-Posting-Host: dcloan.hou.aspentech.com
X-Trace: selma.aspentech.com 971453633 22924 10.32.115.107 (13 Oct 2000 16:13:53 GMT)
X-Complaints-To: postmaster AT aspentech DOT com
NNTP-Posting-Date: 13 Oct 2000 16:13:53 GMT
X-NewsEditor: ED-1.5.8
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp
Reply-To: djgpp AT delorie DOT com

> It's strange but if I try to read from 0xFFFC 0000 to 0xFFFF EFFF the
> program is OK!!! I have a problem to acces last 4K!!

Uh, oh.  This sounds like a physical mapping bug.  The final address in this case
(start address + size) would overflow to zero.  So the final address would be 
less than the starting, and nothing would get mapped.

This looks like a CWSDPMI bug.  In r4 and still in r5.  I think I see a fix.
Well, better to find it this week than next week ... 

ftp://clio.rice.edu/djgpp/fffffix.zip  (contains a new CWSDPMI r5 image).

Please test.