> On Mon, 15 Mar 1999, Philip Gladstone wrote:
>
> > Under 2.0.3x, if you run with a serial console, then you will find that
> > printk's
> >
> > have interrupts disabled while they run. Thus a 100character printk takes
> > about 100ms (10 ticks lost) on a 9600 bps line.
>
> hm, this case looks legitimate. It's not fixed in 2.2, it cannot be
> 'fixed' because it's not really broken. printk is definitely unique in
> this regard. I couldnt find a better solution but to not print out the
> debugging messages when the serial console is running. (people running a
> serial console are mostly debugging the kernel anyway)
I think we can do something about this, actually. We can cause printk to
buffer for low severity levels. In fact, we can dynamically decide to
buffer based on string length, baud rate, and message priority, trying to
always stay below a timer tick.
People running serial may in fact be mostly doing embedded development and
might actually care to know that serial console is destroying their
latency.
-- "Love the dolphins," she advised him. "Write by W.A.S.T.E.."
- 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/