Newsgroups: comp.os.msdos.djgpp
From: Peter Berdeklis <peter AT atmosp DOT physics DOT utoronto DOT ca>
Subject: Help! __dpmi_lock_linear_region wont!
Message-ID: <Pine.SGI.3.91.970109173452.10525A-100000@chinook.physics.utoronto.ca>
Nntp-Posting-Host: chinook.physics.utoronto.ca
Sender: news AT info DOT physics DOT utoronto DOT ca (System Administrator)
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Organization: University of Toronto - Dept. of Physics
Date: Thu, 9 Jan 1997 23:11:00 GMT
Lines: 54
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp


I'm trying to lock some DOS memory for a real mode interrupt handler and a 
bunch of DMA buffers, but every call to __dpmi_lock_linear_region returns 
a -1.  Why?

This is what I'm doing for all the DOS memory regions (properly allocated 
with a call to __dpmi_allocate_dos_memory - no error returned).


void example()
{
__dpmi_meminfo vector_mem;
int real_mode_int_segment, real_mode_int_selector;
int paragraphs = (handler_size + 15)>>4;

real_mode_int_segment = __dpmi_allocate_dos_memory( paragraphs, 
						&real_mode_int_selector );
if( real_mode_int_segment == -1 )
	exit(-1);	// no exit, allocation OK

vector_mem.handle = 0;	// should be don't care for this call, right?
vector_mem.size = handler_size;
vector_mem.address = real_mode_int_segment << 4;  // dos memory always in
						// paragraphs so offset=0,
						// right?

if( __dpmi_lock_linear_region( &vector_mem ) == -1 )
	exit(-1);	// region not locked, exit
}


I do the same thing for the buffers, and every call to lock memory 
returns an error (I don't actually exit, so I know this).  Why would the 
DPMI host refuse to lock the memory?  I'm using CWSDPMI.  Am I making the 
call correctly?  Is the handle a don't care for this (and most, I think) 
calls?

By the way, according to go32-v2.exe I have:

DPMI memory available: 1483 Kb
DPMI swap space available: 44905 Kb

(Yes, I know that pretty sparse on the memory, but it's a small program 
on an old machine.)

Finally, I know that I could lock all the memory with the crt0 flags, but 
I'ld rather make this work the right way. :)

Thanks in advance for your help.

---------------
Peter Berdeklis
Dept. of Physics, Univ. of Toronto