On 5/7/11 12:05 PM, Kevin O'Connor wrote:
On Sat, May 07, 2011 at 08:48:45PM +0200, Sven
Do you know how long it takes for your keyboard
to become responsive?
(You can pass "-n" to tools/readserial.py to have it report wall times
instead of adjusted times.)
It takes something about 0.8s, according to serial
Does adding "msleep(800)" to the top of
Please don't do things like that. Probing the hardware is much better
than a fixed delay.
The various i8042 reimplementations out there are quite ugly and each in
its specific way.
Maybe we should look at the coreboot mainboard type in order to decide
whether to poll the hardware or drop out on the first 0xfe answer. Or at
least make the behavior configurable in Kconfig?
The logs you posted show the i8042 controller
responsive when SeaBIOS
starts talking to it. So, I wonder if it's the keyboard that is still
powering up (instead of the superIO).
Not unlikely. It might also just be some crappy implementation of EC