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: DPMI Interrupt latency ?
Date: Tue, 29 Jan 2002 21:22:34 CST
Organization: Rice University, Houston TX
Lines: 21
Message-ID: <3c5766fa.sandmann@clio.rice.edu>
References: <Pine DOT SUN DOT 3 DOT 91 DOT 1020129153726 DOT 8440D-100000 AT is> <slrna5de5e DOT 5cd DOT kala AT ganga DOT sankya DOT com>
NNTP-Posting-Host: clio.rice.edu
X-Trace: joe.rice.edu 1012361146 25557 128.42.105.3 (30 Jan 2002 03:25:46 GMT)
X-Complaints-To: abuse AT rice DOT edu
NNTP-Posting-Date: 30 Jan 2002 03:25:46 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

> That makes sense - and would explains why my handler didn't get called.
> My confusion was only regarding whether the DPMI server would accept 
> INT 8D blindly as a hardware interrupt or if it would actually read the
> ISR from the PIC  to determine that the interrupt was generated through
> hardware before deciding whether to forward it or not. The latter seems
> to be true at least with CWSDPMI. ( Wonder if PMODE/DJ checks this too -
> I suspect it may not, that would explain one major source of the faster
> response - Reading the ISR requires 2 I/O instructions !)

Sorry I don't have ttime to look at this completely - I'm out of town on
business trip.  It doesn't read the ISR.  It just hooks the real mode
interrupt table and directs it to a mode jump.  Yes, it proably must be
the real HW interrupt (which is revectored for the low PIC).  It will
work if you find right one.