Xref: news-dnh.mv.net comp.os.msdos.djgpp:1447
Path: news-dnh.mv.net!mv!news.sprintlink.net!newsfeed.internetmci.com!news.msfc.nasa.gov!bcm.tmc.edu!newsfeed.rice.edu!rice!news!sandmann
From: Charles Sandmann <sandmann AT clio DOT rice DOT edu>
Newsgroups: comp.os.msdos.djgpp
Subject: Re: CLI asm instruction
Date: Tue, 08 Aug 1995 20:23:44 CDT
Organization: Rice University, Houston, Texas
Lines: 12
References: <DCzGMs DOT Ius AT jade DOT mv DOT net> <302779a0 DOT sandmann AT clio DOT rice DOT edu> <408or5$bk0 AT st-james DOT comp DOT vuw DOT ac DOT nz>
Reply-To: sandmann AT clio DOT rice DOT edu
Nntp-Posting-Host: clio.rice.edu
To: djgpp AT sun DOT soe DOT clarkson DOT edu
Dj-Gateway: from newsgroup comp.os.msdos.djgpp

> If CLI is so slow then how are interrupts disabled inside interrupt 
> routines normally? Or do they just take the performance hit.

Interrupt routines in PM are by nature slow, since they almost always involve
a ring change, a stack change, and usually a change to RM and back.
The interrupt flag gets cleared during all of the ring 0 processing,
instead of being handled in an exception, so doesn't add anything there.

But CLI is only slow if the DPMI runs occasionally at IOPL != CPL, which 
multitasking OS's would do.  If the DPMI always runs as IOPL == 3 (like
CWSDPMI) then you always get direct access to the flag and you can wedge
the machine at will ...