delorie.com/archives/browse.cgi   search  
Mail Archives: djgpp/1999/02/14/06:05:30

Date: Sun, 14 Feb 1999 13:02:25 +0200 (IST)
From: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
X-Sender: eliz AT is
To: Engard Ferenc <fery AT pons DOT sote DOT hu>
cc: djgpp AT delorie DOT com
Subject: Re: pm and rm interrupt handlers vs. emm386
In-Reply-To: <Pine.LNX.3.96.990212164029.25362C-100000@Pons.sote.hu>
Message-ID: <Pine.SUN.3.91.990214130207.15332B-100000@is>
MIME-Version: 1.0
Reply-To: djgpp AT delorie DOT com

On Fri, 12 Feb 1999, Engard Ferenc wrote:

> My question is
> that it is enough to write a real-mode handler, or (if my program
> spend much time in protected mode), I need to write a protected-mode
> handler too for the best performance?

It's actually the other way around: you should install a PM handler
and keep your program in protected mode as much as possible, for the
best throughput.  That's what the FAQ tries to explain in section
18.11.

Only if the above is not enough to keep up with the interrupt rate,
should you consider installing a real-mode handler.

> While the interrupts are disabled (I
> suppose I need to disable it in a hardware irq-handler), other
> hardware interrupts are lost? (E.g., data coming from kbd/serial is
> lost?)

Interrupts are not lost, unless you leave them disabled long enough to
lose more than a single interrupt from a single source.  Otherwise,
they are just delayed until such time as you re-enable interrupts.

- Raw text -


  webmaster     delorie software   privacy  
  Copyright © 2019   by DJ Delorie     Updated Jul 2019