[EMAIL PROTECTED] wrote:
> I'd love to see a solid timing study showing that. Can you do
> ping -f someone &
> while true; do tar cf /dev/null /dev/hda_x ; done; &
> while true; do make -j 4 bzImage ; done; &
> and still get sub ms latencies?
> That would be cool.
Nice challange. I took it, in a slightly different way, as follows:
ping - f server
do make -j 4 bzImage
while "true"; do ls -aR /; done
while "true"; do cat /proc/interrupts; done
while running 3 hard real time tasks at 10 Khz in oneshot mode, amid 12
Linux SCHED_FIFO, in USER SPACE as a nonroot user .
The highest priority hard real time task measures its scheduling
latency. After testing for 1 hour I got the following results:
MAX JITTER 30 us,
AVERAGE JITTER 10 us.
The oneshot was not calibrated, if such you could have taken away some 7
us from the above figures.
The machine is a UP PII 233 Mhz.
The test is running while I'm writing this message, with a good alpha
response but with sluggish mouse based cut-pastes. Nonetheless by
looking at the looping "cat /proc/interrupts" above I can see that my
box is taking many hundreds of interrupts per second without any problem
and the ping -f dots are below 3 lines.
The result above can be obtained thank to a joint cooperative effort of
a group of friends.
Ciao, Paolo
-- [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/
