problem with IRQ 8 (RTC) [SOLVED]

[Ferrarius]

hey everyone,

I'm having a to me unexplainable problem with IRQ 8. When this interrupt is called the computer/bochs locks up. Bochs as output gives the following:
check_cs(0x0255): conforming code seg descriptor dpl > cpl, dpl = 3 cpl = 1.

The thing is I've not done anything with regard to ring changing yet. and the descriptor has a DPL of 0.

Code: Select all

;##INT 0x28  -  Real Time Clock##

dw rtc

dw 0x08

dw 0x8E00

dw 0x02

the OS is loaded in at 0x20000, hence the last dw 0x02. When calling the interrupt with int 0x28 it does work correctly.

does anyone have an idea where this problem might just come from?


Solved: A small off by one error was the problem. Please close

[earlz]

I would like to point out that many computers have buggy implementations of the RTC. a lot of computers(Dells) either have a broken IRQ 8, or a very slow IRQ8.. you should compensate for this or else you'll have a whole bunch of computers your system doesn't work on

[johnsa]

I've encountered the irq8 issue before on my dell machines, but reading/writing the cmos/rtc values still seems to be ok. I'm assuming that even though the rtc periodic interrupt is dodgy we can still safely use those ports to get the system time at boot and sync it again on shutdown? (during the os the time is maintained through a different mechanism of course for accuracy and features).