From: robk AT cyberway DOT com DOT sg (Rob Kramer)
Newsgroups: comp.os.msdos.djgpp
Subject: Re: DJGPP and Mapping physical memory from a PCI device
Date: Wed, 14 Jun 2000 11:51:34 +0800
Message-ID: <MPG.13b1821f161f6e6c98969c@news.cyberway.com.sg>
References: <Pine DOT SUN DOT 3 DOT 91 DOT 1000613133915 DOT 16888Z-100000 AT is>
Organization: Infologic Pte Ltd
X-Newsreader: Anawave Gravity v2.00
NNTP-Posting-Host: 203.116.147.127
Lines: 24
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp
Reply-To: djgpp AT delorie DOT com

In article <Pine DOT SUN DOT 3 DOT 91 DOT 1000613133915 DOT 16888Z-100000 AT is>, 
eliz AT is DOT elta DOT co DOT il says...
> 
> On Tue, 13 Jun 2000, Rob Kramer wrote:
> 
> >     /* Lock segment (dpmi 0x0600). */
> >     if (__dpmi_lock_linear_region (&info) == -1) {
> >         __dpmi_free_ldt_descriptor (selector);
> >         return false;
> >     }
> 
> I don't think this is required: the mapped region is already locked by
> the DPMI host.  In fact, with some DPMI hosts this call will actually
> fail, thus creating an illusion that the whole map operation didn't
> work.

Okidokie. It worked on cwsdpmi, so I never changed the code. I remember 
the segment unlocking function was actually a DPMI 1.0 call, which I 
found a bit strange..

Cheers,

	RoB