delorie.com/archives/browse.cgi   search  
Mail Archives: djgpp/2006/07/13/01:48:15

X-Authentication-Warning: delorie.com: mail set sender to djgpp-bounces using -f
From: Charles Sandmann <sandmann AT clio DOT rice DOT edu>
Newsgroups: comp.os.msdos.djgpp
Subject: Re: BUGs in DJGPP documentation & CWSDPMI please fix
Date: Thu, 13 Jul 2006 00:07:21 CDT
Organization: Rice University, Houston, TX
Lines: 49
Message-ID: <44b5d509.sandmann@clio.rice.edu>
References: <1152709411 DOT 653483 DOT 258660 AT h48g2000cwc DOT googlegroups DOT com>
NNTP-Posting-Host: clio.rice.edu
X-Trace: joe.rice.edu 1152768354 20940 128.42.105.3 (13 Jul 2006 05:25:54 GMT)
X-Complaints-To: abuse AT rice DOT edu
NNTP-Posting-Date: Thu, 13 Jul 2006 05:25:54 +0000 (UTC)
X-NewsEditor: ED-1.5.9
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp
Reply-To: djgpp AT delorie DOT com

> OK. I definitely will not release anything not perfectly legal.
> I am trying to develop a (dirty ?) workaround and achieved
> some promising (?) results so far.

We will certainly allow something, don't worry.  I just want to
make sure each request meets at least the spirit of either GPL
or my more liberal interpretion.

> Nevertheless, a CWSDSTUB.EXE with this SWAP flag working
> still would be useful. The latest "official" release of CWSDPMI
> dates from 2003, and contains CWSDPMI & CWSDPR0, but
> no CWSDSTUB at all. The version is correctly reported when
> printed to screen ("CWSDPMI -P" -> "CWSDPMI V0.90+ (6b)
> Copyright (C) 2003 CW Sandmann"), while the version at the
> beginning of the EXE remains "r5 Copyright (C) 2000" and
> the "__dpmi_get_capabilities" still reports 5,0.

This hurts my mind to think about the r6 test release stuff
at the same as bin patching something.  I'm confused, tired and
sleep deprived.  Sigh.  Sales meetings.

> "1) 4Mb regions are treated as locked or non-pageable.
> This means you cannot use virtual memory effectively
> with 4Mb pages."

> "1) Should 4Mb pages be "paged" to disk like 4K pages?
> This might take up to several seconds just for
> the I/O to page-out/page-in a 4Mb page."

> "2) If we don't allow paging of 4M pages,
> should this only be available when swapping is disabled?
> Or in the ring 0 version?

> My suggestion would be a CWSDSTUB with no Ring0, no swapping, and
> version and flags reported correctly.

We'll fix something, hang in there.  I have considered other
hacks (like making 4K pages only used in the first 128MB of 
address space).  Let's just say the DPMI spec didn't really
think about 4MB pages very much.

> In any case, I'll remain legal with CWSDPMI keeping Causeway DOS
> Extender (Public Domain) in mind.

I have no desire to stop you distributing anything, I just want to
make sure the users are protected by being able to replicate the changes
or do a clean build to reproduce the changes.  I've granted dozens
of special requests in the past (my goal is wide, useful distribution)
but I like to typically handle out of public forums.

- Raw text -


  webmaster     delorie software   privacy  
  Copyright © 2019   by DJ Delorie     Updated Jul 2019