Mail Archives: djgpp/1997/02/05/21:28:34
nikki AT gameboutique DOT co (nikki) wrote:
>> D6 is an "undocumented" (newly-documented) opcode, called SALC (Set AL
>> with Carry). It seems its behaviour is already described in your
>> comments.
>ack undocumented :( do you know anywhere it might be 'newly documented' as i
>don't have any reference to it anywhere and wouldn't mind knowing some more
>about it. doesn't seem to be on intels sheet :(
It may be documented in the Pentium Pro programmer's manual. See also
http://www.x86.org/
>> This handler really only updates a table with which keys are down, and
>> it doesn't even do that well. Instead of wasting your time, consider
>yeh i noticed :( doesn't appear to push or pop anything either which seems
>bad practice for an interrupt handler.
The "interrupt" directive in Turbo Pascal already does all of the
pushing and popping, loading DS, etc, so this is not a problem.
>> using already-existing better-written keyboard handlers instead. I
>> think there's pointers to some in the FAQ.
>yeh, but i don't actually need the code as a matter of life and death :) it's
>just an exercise in writing it really :) i don't mind wasting time as i have
>nothing better to do :)
Sorry, I haven't actually ever done this using DPMI myself, so someone
else will have to help you with that.
However, I understand that when you're in protected mode, a protected
mode handler is faster. In real mode, a real mode handler is faster.
Since djgpp runs in protected mode, you need a protected mode handler.
Only if you also spend much time in real mode (disk access etc), is
when you should consider a real mode handler in addition to the
protected mode one, if you sorely need the extra speed.
- Raw text -