Messages in this thread | | | Date | Fri, 29 Mar 1996 14:08:49 +0200 | From | Andries.Brouwer@cwi ... | Subject | Re: 1.3.79 - The Keyboard that never returned... |
| |
Colten Edwards:
: Wonder if this is related to my dead keyboard syndrome. my no concernable : reason my keyboard will lockup completely dead. Reset button time. tasks : are still running as I watched a kernel re-compile finish once. The : screen will blank but nothing on the keyboard is active. There are no : messages written in syslog. This will usually occur if I leave the room : for a minute and usually when in an editor with half finished mail :) But : it has happened at a prompt as well. and it's really bothering me. : Somedays it doesn't happen at all, but other days like today it occured : 3-4 times.
: Pentium 100 Gigabyte with 256k of cache 16meg of mem.. Old old mitsumi AT : keyboard. PCI triton. 4 IDE drives and a soundblaster. ATI mach-32 ISA with : 2meg of video. APM is disabled, but everything else is set to its fastest : settings.
A well-known phenomenon, that I used to see a few times a year. With 1.3.78 however, I saw it 6 times in 6 hours. It is definitely a timing question, although I do not understand the details. If you can login remotely, it is easy to solve: writing two bytes to port 60 gets your keyboard running again [of course, assuming that you see the same problem that I saw with 1.3.78]. Otherwise you may be able to use the mouse to start the program below.
Thus, typing on the spot, no guarantees given:
#include <fcntl.h>
outkbd(int fd, char z) { char w; do { lseek(fd, 0x64, 0); read(fd, &w, 1); } while ((w & 2) == 2 ); lseek(fd, 0x60, 0); write(fd, &z, 1); }
int main(){ int fd = open("/dev/port", O_RDWR); if (fd < 0) { /* need to be root */ perror("/dev/port"); exit(1); } outkbd (fd, 0); /* one is not enough */ outkbd (fd, 0); /* now the keyboard is working again */ return 0; }
| |