X-Authentication-Warning:	delorie.com: mailnull set sender to djgpp-bounces using -f
From:	Charles Sandmann <sandmann AT clio DOT rice DOT edu>
Newsgroups:	comp.os.msdos.djgpp
Subject:	Re: DS and ESP
Date:	Tue, 19 Feb 2002 11:01:15 CST
Organization:	Rice University, Houston TX
Lines:	34
Message-ID:	<3c7284db.sandmann@clio.rice.edu>
References:	<a4rfgu$sbg$1 AT news DOT tpi DOT pl>
NNTP-Posting-Host:	clio.rice.edu
X-Trace:	joe.rice.edu 1014138624 1439 128.42.105.3 (19 Feb 2002 17:10:24 GMT)
X-Complaints-To:	abuse AT rice DOT edu
NNTP-Posting-Date:	19 Feb 2002 17:10:24 GMT
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

> I'm developing some kind of simple API (as a library) to install
> protected mode interrupts under DJGPP. I dont't use standard DPMI
> functions to install interrupts. Instead i create some alias
> descriptors in LDT and install my service routines directly in IDT
> (it works under DOS and CWSDPMI only but its sufficient for me).

Interesting that you are using this feature.
I allowed users to put anything into the LDT so they could also set up
protected mode services - such as a ring 0 routine which could be 
called from their ring 3 program.
But I never knew if anyone had done anthing with it.  Interesting.

>    In this stub routine I assume 2 things:
>  a) ds segment always contains selector for .data segment.

You should probably load ds_alias instead, but since you have 
control of when people call your interrupt, this could also be 
an interface requrirement.

>  b) similar problem is about esp (I belive that ss is valid ).
>     My stub routine assumes that valid program data on stack are
>     pointed by esp and higher values. 

If this is really important to you have your routine run at a 
different priv level so you get swapped to a different stack.  This
might be a case you want to use the DPMI facilities, since CWSDPMI
will swap stacks for you and not touch the user stack on an exception
or a fault.

>     I belive that GCC doesn't generate code like above and
>     there is no such incoherenties in DJGPP's libc.

There were some a long time ago in the library, but I believe they 
all have been fixed.

- Raw text -