What other devices are on this thing? My guess is that it has a bad
ide isa controller that locks the bus.
To test: modify the measurement rtl code to just do something like
do forever
do 1000 times
read cycle counter
read an io from serial device
check difference in times
done
wait
done
then the module could report on the longest times to get to the io-bus.
run tar cf /usr /dev/null
during the measurement.
I suspect that you'll see a long delay sometimes, and there is not
a thing that RTL can do about that. If you don't see that, then
we can try something else.
On Tue, Mar 23, 1999 at 12:23:30PM +0100, Bernhard Nebel wrote:
> Hi all,
>
> I've installed RTLinux v1.1 on a Toshiba Libretto CT100 notebook
> (Pentium 166 MMX) because we want to capture data from the serial
> interface at 500K Baud. Unfortunately, my module (based on rt_com) still
> loses bytes from the serial interface (i.e., I get the OVERRUN error
> from the UART). Running the "measure" example on the notebook shows that
> the maximal delay time for the scheduler can be as worse as 300
> microseconds. My own PC at home shows always a maximal delay of 20
> microseconds.
>
> Now my question: Could it be the case that the hardware is responsible
> for this behavior (i.e., because the notebook does not have a 2nd level
> cache)? Is there perhaps a cure?
>
> -- Bernhard
> --- [rtl] ---
> To unsubscribe:
> echo "unsubscribe rtl" | mail [EMAIL PROTECTED] OR
> echo "unsubscribe rtl <Your_email>" | mail [EMAIL PROTECTED]
> ----
> For more information on Real-Time Linux see:
> http://www.rtlinux.org/~rtlinux/
--- [rtl] ---
To unsubscribe:
echo "unsubscribe rtl" | mail [EMAIL PROTECTED] OR
echo "unsubscribe rtl <Your_email>" | mail [EMAIL PROTECTED]
----
For more information on Real-Time Linux see:
http://www.rtlinux.org/~rtlinux/
