Message-ID: <37F3FA6B.4B73AF3@virtualis.com>
Date: Fri, 01 Oct 1999 10:03:55 +1000
From: Alex Mendes da Costa <alexmdc AT virtualis DOT com>
X-Mailer: Mozilla 4.51 [en] (Win98; I)
X-Accept-Language: en
MIME-Version: 1.0
Newsgroups: comp.os.msdos.djgpp
Subject: Re: More Assembly!
References: <Pine DOT SUN DOT 3 DOT 91 DOT 990930160122 DOT 21365H-100000 AT is>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
NNTP-Posting-Host: a5-p05.syd.fl.net.au
X-Trace: 1 Oct 1999 10:07:30 +1000, a5-p05.syd.fl.net.au
Organization: Customer of First Link Internet Services, Sydney, Australia
Lines: 47
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp
Reply-To: djgpp AT delorie DOT com

Well, basically, I've got this bit of code written in assembly that I
want to use as an interrupt. I've only been working with DJGPP for about
a week so far, and I don't fully understand all this protected mode
stuff. When I wrote an interrupt in real mode I loaded it into memory
and then redirected the interrupt to that place in the memory. But I
guess its different for this.
Just say I have the following:

void interrupt()
{
  asm("
       blabla
      ");
}

and I want an interrupt to point to it. So I have to:
1. Save the __dpmi_get_protected_mode_interrupt_vector of the interrupt
I'm changing.
2. Now I don't understand this bit. It says call
__dpmi_lock_linear_region a series of times. But how do I know what
region I have to lock?
3. Then call __dpmi_set_protected_mode_interrupt_vector with _my_cs()
and my function offset. How do I get my function offset?

According to the FAQ that is all I have to do (apart from fix up the
interrupt at the end). So can anyone help me with the bits I don't get?

Thanks,
Alex


Eli Zaretskii wrote:
> 
> On Thu, 30 Sep 1999, Alex Mendes da Costa wrote:
> 
> > Sorry I keep asking questions!! I was wondering how its possible to load
> > a bit of assembly code that has been written in the inline assembler
> > into a place in memory, for use in an interrupt.
> 
> Not at run time.  You need to assemble the code separately, and put into
> your program an array of bytes that are the opcodes of the handler.  Then
> use memcpy or whatever to copy them.
> 
> But I really don't understand why do you need to go to such lengths?
> Can't you simply install your function, however it is written, as an
> interrupt handler?  (Section 18.9 of the DJGPP FAQ tells you how.)  Why
> do you need to copy it somewhere?