On Tue, Jul 27, 2004 at 03:47:04PM +0200, Philippe Gerum wrote:
> > using fusion-0.3-2.6-i386/testsuite/latency
> > 
> > > min = 5185 ns, max = 79593 ns, avg = 10617 ns, overrun = 0
> > > min = 5185 ns, max = 79593 ns, avg = 10797 ns, overrun = 0
> > > min = 5185 ns, max = 79593 ns, avg = 11028 ns, overrun = 0
> > > min = 5185 ns, max = 79593 ns, avg = 10979 ns, overrun = 0
> > > min = 5185 ns, max = 79593 ns, avg = 9906 ns, overrun = 0
> > > min = 5185 ns, max = 79593 ns, avg = 9144 ns, overrun = 0
> > > min = 5185 ns, max = 79593 ns, avg = 11602 ns, overrun = 0

> What would be interesting is having a distribution of the peaks for
> fusion/latency. I wonder if this is a one shot problem at startup, or a
> continuous jitter spreaded all over the test.

# at startup:
min = 5073 ns, max = 17235 ns, avg = 5930 ns, overrun = 0
min = 5073 ns, max = 19038 ns, avg = 5989 ns, overrun = 0
min = 5073 ns, max = 19038 ns, avg = 6062 ns, overrun = 0
min = 5073 ns, max = 19038 ns, avg = 6057 ns, overrun = 0

# while typing this email:
min = 5073 ns, max = 29922 ns, avg = 6042 ns, overrun = 0
min = 5073 ns, max = 29922 ns, avg = 6015 ns, overrun = 0
min = 5073 ns, max = 29922 ns, avg = 5963 ns, overrun = 0
min = 5073 ns, max = 29922 ns, avg = 6016 ns, overrun = 0
min = 5073 ns, max = 29922 ns, avg = 6126 ns, overrun = 0

# with ping -f
# and while true; do echo `seq 1 46`; done
min = 5073 ns, max = 32894 ns, avg = 8447 ns, overrun = 0
min = 5073 ns, max = 45732 ns, avg = 8667 ns, overrun = 0
min = 5073 ns, max = 45732 ns, avg = 8419 ns, overrun = 0
min = 5073 ns, max = 45732 ns, avg = 8386 ns, overrun = 0
min = 5073 ns, max = 45732 ns, avg = 8389 ns, overrun = 0

# dd
min = 5073 ns, max = 45732 ns, avg = 8897 ns, overrun = 0
min = 5073 ns, max = 55220 ns, avg = 9098 ns, overrun = 0
min = 5073 ns, max = 55220 ns, avg = 9374 ns, overrun = 0
min = 5073 ns, max = 67279 ns, avg = 9068 ns, overrun = 0
min = 5073 ns, max = 67279 ns, avg = 8722 ns, overrun = 0
min = 5073 ns, max = 67279 ns, avg = 8447 ns, overrun = 0
....
min = 5073 ns, max = 67279 ns, avg = 8751 ns, overrun = 0
min = 5073 ns, max = 72061 ns, avg = 10184 ns, overrun = 0
min = 5073 ns, max = 72061 ns, avg = 10651 ns, overrun = 0


> In any case, there is at least a ~4 us inherent jitter as shown by min
> values which is not correctly taken in account by the calibration code
> for either/both timer and scheduling. But this obviously does not
> explain all.

> At the very least, did you experiment any unbounded jitter while
> testing?

No - the jitter increases as I put more load on my system.

Marc

-- 
#!/bin/sh
set - `type $0` 'tr "[a-zA-Z]" "[n-za-mN-ZA-M]"';while [ "$2" != "" ];do \
shift;done; echo 'frq -a -rc '`echo "$0"| $1 `'>$UBZR/.`rpub signature|'`\
echo $1|$1`'`;rpub "Jr ner fvtangher bs obet. Erfvfgnapr vf shgvyr!"'|$1|sh

Reply via email to