About result of rt/hrt patch, is it reasonable?

["Lee David" <live4thee@xxxxxxxxx>]

Hi, all

I downloaded rt-preempt patch and hrt patch, after successfully compiled a new
kernel I ran the cyclictest three times each. As to rt vs. hrt, the result
seemed to be not that different as showed in Gleixner and Niehaus's OLS paper
"Hrtimers and Beyond: Transforming the Linux Time Subsystem."

Was my result below make sense? Or did I missed something?
Any comments and advices are welcome.
----------------------------------------------------------------------
* The following result is from the paper:
All tests have been run on a Pentium III 400MHz based PC.
Test case: clock_nanosleep(TIME_ABSTIME), interval 10000 microseconds,
10000 loops, no load.

Commandline: cyclictest -t 1 -p 80 -n -i 10000 -l 10000
Kernel 	        min 	max 	avg
2.6.16 	        24 	4043 	1989
2.6.16-hrt5 	12 	94 	20
2.6.16-rt12 	6 	40 	10

* Here is my result:
Celeron M 370 (1.5 GHZ) w/ 512M memory.
I turned on "HPET Timer Support" for all the three cases.
For rt, I chose "Complete real-time preemption".

kernel          min     max     avg
2.6.16          8       4038    2027
               8       4055    1990
               8       4018    1991
2.6.16-hrt6     16      94      62
               17      78      61
               11      72      63
2.6.16-rt29     14      79      62
               11      74      62
               11      71      63
----------------------------------------------------------------------
rt-preempt patch:
http://people.redhat.com/mingo/realtime-preempt/older/patch-2.6.16-rt29

hrt patch:
http://www.tglx.de/projects/hrtimers/2.6.16/patch-2.6.16-hrt6.patch

Cyclictest:
http://tglx.de/projects/misc/cyclictest/cyclictest-v0.11.tar.bz2

Thanks in advance,
D.L.

--
Kernelnewbies: Help each other learn about the Linux kernel.
Archive:       http://mail.nl.linux.org/kernelnewbies/
FAQ:           http://kernelnewbies.org/faq/