Messages in this thread |  | | From | Herbert Rosmanith <> | Subject | 1.99.14: keyboard dead | Date | Fri, 7 Jun 1996 11:21:46 +0200 (MET DST) |
| |
fyi,
1.99.14 still gives me a dead keyboard, that is, no VC switch, no LEDs that flash on/off (pressing CAPS/NUM/..). however, I still can log in from my LAN.
Emil sent me the following programm:
> I have got the same symptoms, though only when my GUS is playing any sound > for a while. It looks like missed interrupt for me because reading a single > character directly from keybord I/O returns it to life. It needn't to be > your case but why not trying this when logged from another machine. > This works for me: > > (*---------cut here---------*) > > #include <fcntl.h> > > int main() > { > int fd; > unsigned char w; > > fd=open("/dev/port",O_RDONLY); > if(fd<0) printf("Error 1"); > lseek(fd,0x60,0); > read(fd,&w,1); > close(fd); > printf("value=%x\n",w); > } > (*--------cut here-----------*) >
I started the program, and the keyboard remained dead. So, I started it a couple of times, then the keyboard returned to live. typing "dmesg" shows a "keyboard error".
hardware: asus ap4 i486/100 2x16MB real aha2940 cpf1060s (scsi-disk) seagate (for m$/dos only) soundblaster sb16 ide-cdrom drive (on /dev/hdb) ne2000 compatible
emil reported that the problem occurs when he starts his GUS soundcard. maybe my sb16 is causing the problem ? I have sb16 support compiled in, but (since I have put my boxes to a different place) I am not using it. will disable the sb16 code and tell you if that fixes it.
/herp
|  |