Hi,

I've discovered some interesting points about the rtc and related 
interrupts and the use of hpet in the kernel config, but I am quite 
unsure which is best to choose:

In character devices config there is a option called "Enhanced Real Time 
Clock Support" (RTC).
Disabling this option makes another option in Character devices visible, 
right under "HPET": "HPET control RTC IRQ".

And there is even a whole Real Time Clock class in the kernel config 
which is relatively new and it seems it allows one to have multiple real 
time clocks.
Loading the "Enhanced RTC" driver and then trying to load the rtc-cmos 
module of the rtc class produces a problem:

rtc_cmos: probe of 00:08 failed with error -16

Without the rtc module loaded, rtc-cmos loads just fine:

rtc_cmos 00:08: rtc core: registered rtc_cmos as rtc0
rtc0: alarms up to one year

Vice versa the rtc driver won't load any longer with rtc-cmos driver 
already present:

FATAL: Error inserting rtc 
(/lib/modules/2.6.24-rc5-shutdown/kernel/drivers/char/rtc.ko): 
Input/output error

So it seems that those rtc-stuff is somehow related to HPET as a 
clocksource and interrupts as well - but which combination should I choose?
Should I avoid using the rtc driver and having HPET control the RTC IRQ 
instead and if yes, should I load rtc-cmos instead of rtc or shouldn't I 
use any of the rtc drivers for best HPET operation???

This is quite confusing to me since I can't find out what effects which 
driver will have considering IRQs and powersaving.
The fact that the description of the rtc driver mentions something about 
reading the clock in a SMP-fashioned way and because I am running a 
Core2 Duo I am running a SMP kernel as well.

So can anyone tell me what options I should use and which I shouldn't?

Peter

_______________________________________________
Power mailing list
[email protected]
http://www.bughost.org/mailman/listinfo/power

Reply via email to