Andrea Arcangeli wrote:
> For doing that you should run some code also on the other CPU. Why do you
> want to do that? It's slower. Storing 64bit information is much faster.
>
But I am going to do that only at interrupt time, instead of 64bit
arithmetic in do_gettimeofday(). Also in this way I don't have to store
current time.
I also think that the right place to store rdtsc is when we update xtime
(is it update_wallclock ?)
that way delay_at_interrupt is not needed and we have less discrepancy
between
xtime and gettimeofday()
> btw, the per-cpu fields you added should really be cacheline aligned or
> you'll have cacheline pingpong each time the timer irq happens on another
> CPU.
Which one should I use ____cacheline_aligned or __cacheline_aligned ?
Boris
-
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/
This archive was generated by hypermail 2b29 : Wed Mar 15 2000 - 21:00:16 EST