On Thu, Mar 21, 2002 at 10:50:36 -0500, Taybin Rutkin wrote:
> On Thu, 21 Mar 2002, Paul Davis wrote:
> 
> > i noticed from luca's initial report on firm timers that switching the
> > caps lock LED on and off is a notable source of latency in the
> > kernel, on the order of 7msec.
> 
> Seems odd.  I would have expected the LED to be handled in the keyboard's
> hardware.  Maybe only certain models?  But I guess it must be universal
> since there is software to toggle the LEDs.

Certainly on AT (and I think PS2) keybaords the LED is controlled by the
CPU. It was a standard way to find out if the kernel had died.

- Steve

Reply via email to