home *** CD-ROM | disk | FTP | other *** search
- From: asickels@bonnie.ics.uci.edu (Alan Sickels)
- Subject: Re: Reverse Num-Lock problems
- Date: 19 Apr 89 23:52:10 GMT
- Summary: It's internal
-
- bobc@killer.Dallas.TX.US (Bob Calbridge) writes:
- >Here's a problem that has started to pop up lately at our school district.
- >Under certain circumstances the numeric key pad acts contrary to what the
- >numlock light says. We suspect anything from the keyboard to the software.
-
- I've got an XT clone and it happens quite frequently to me. This is what
- is happening: The true NUM-LOCK state on PC/XT's and clones is stored in
- memory in bit 5 of byte 0040:0017H. With the AT and 386's, the keyboard
- communication is two way and the keyboard lights reflect the true state
- of the memory location. The PC/XT's can't do that so it's kept track of
- independent of the true state. This also annoyed me so I wrote a quick
- assembly program to turn all the indicator bits off (i.e. NUM-LOCK,
- CAPS-LOCK, and SCROLL-LOCK) and just make sure all the lights are off
- when I run it. Source code included.
-
- MOV AX,0040H
- MOV DS,AX
- MOV BX,0017H
- AND BYTE PTR [BX],8FH
- MOV AH,4CH
- INT 21H
-
- Alan Sickels