>> I don't have kernel sources handy to look through, but the above
>> sounds suspiciously like something's reprogramming the timer
>> interrupt frequency from under us, with the result that the system
>> is getting timer interrupts at the wrong rate - that would explain
>> all of the above symptoms in one go...
> But only if the drift was consistently in one direction, which it
> is not AFAIU.
Why?
All it needs is a configuration error somewhere, and either the kernel
or the BIOS maps some important data port to the address of one of the
relevant timer's preset registers, and one gets a totally random
frequency for the timer interrupt.
As an example, assume that the error resulted in the sound card's data
register being mapped that way, and the user decides to play the
voyager theme track, or some other lengthy piece of music. Guess what
frequency your timer interrupt occurs at...
Don't laugh too hard, I met this very scenario not too long ago, and
the ONLY symptom was that the clock was erratic, sometimes gaining
madly, at other times losing equally madly - swings between gaining
and losing over 2 seconds a minute were possible! The clue that helped
me track it down was that the change in clock rate only occurred when
some sort of sound occurred, be it a beep or whatever...
> Also, if it was a RH screwup, it would happen everywhere, not at
> random in either direction for identical machines installed the
> same way.
Again, why?
Best wishes from Riley.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/