* Poor cycletest results for rt6
@ 2011-07-29 11:10 Hinko Kocevar
2011-07-29 11:44 ` N, Mugunthan V
` (2 more replies)
0 siblings, 3 replies; 4+ messages in thread
From: Hinko Kocevar @ 2011-07-29 11:10 UTC (permalink / raw)
To: linux-rt-users
Hi,
I've build linux-3.0-rt6 for my ARM platform (XCEP). Running cycletest
on vanilla, preempt and rt6 kernel build reviles the following results.
I would expect better rt6 performance.
Note that last test actually looks quite promising, but it happened once
in 5 test runs :(
Linux xcep 3.0.0 #1 Thu Jul 28 21:53:19 CEST 2011 armv5tel GNU/Linux
hinko@hinkopc ~ $ ssh root@192.168.11.132 /shared/utils/cyclictest -t 10
-p 80 -n -D 30
WARNING: Most functions require kernel 2.6
policy: fifo: loadavg: 5.61 2.17 0.82 7/44 323
T: 0 ( 295) P:80 I:1000 C: 28983 Min: 17 Act: 201 Avg: 191 Max:
3043
T: 1 ( 296) P:79 I:1500 C: 19323 Min: 20 Act: 237 Avg: 241 Max:
3340
T: 2 ( 297) P:78 I:2000 C: 14494 Min: 23 Act: 198 Avg: 276 Max:
3569
T: 3 ( 298) P:77 I:2500 C: 11597 Min: 22 Act: 287 Avg: 248 Max:
3364
T: 4 ( 299) P:76 I:3000 C: 9664 Min: 24 Act: 181 Avg: 193 Max:
3096
T: 5 ( 300) P:75 I:3500 C: 8283 Min: 26 Act: 508 Avg: 305 Max:
3398
T: 6 ( 301) P:74 I:4000 C: 7247 Min: 26 Act: 404 Avg: 265 Max:
3454
T: 7 ( 302) P:73 I:4500 C: 6442 Min: 24 Act: 126 Avg: 365 Max:
3421
T: 8 ( 303) P:72 I:5000 C: 5797 Min: 27 Act: 593 Avg: 291 Max:
3291
T: 9 ( 304) P:71 I:5500 C: 5270 Min: 24 Act: 860 Avg: 285 Max:
3230
Linux xcep 3.0.0-preempt #1 PREEMPT Thu Jul 28 21:55:54 CEST 2011
armv5tel GNU/Linux
WARNING: Most functions require kernel 2.6
policy: fifo: loadavg: 5.57 4.68 4.29 8/40 2104
T: 0 ( 2092) P:80 I:1000 C: 28997 Min: 18 Act: 44 Avg: 207 Max:
969
T: 1 ( 2093) P:79 I:1500 C: 19332 Min: 24 Act: 38 Avg: 214 Max:
968
T: 2 ( 2094) P:78 I:2000 C: 14499 Min: 24 Act: 28 Avg: 164 Max:
879
T: 3 ( 2095) P:77 I:2500 C: 11599 Min: 26 Act: 190 Avg: 297 Max:
1032
T: 4 ( 2096) P:76 I:3000 C: 9664 Min: 27 Act: 156 Avg: 227 Max:
878
T: 5 ( 2097) P:75 I:3500 C: 8286 Min: 26 Act: 213 Avg: 256 Max:
1011
T: 6 ( 2098) P:74 I:4000 C: 7248 Min: 24 Act: 261 Avg: 177 Max:
692
T: 7 ( 2099) P:73 I:4500 C: 6442 Min: 26 Act: 139 Avg: 263 Max:
967
T: 8 ( 2100) P:72 I:5000 C: 5797 Min: 31 Act: 38 Avg: 279 Max:
795
T: 9 ( 2101) P:71 I:5500 C: 5270 Min: 28 Act: 391 Avg: 242 Max:
971
Linux xcep 3.0.0-rt6 #1 PREEMPT RT Fri Jul 29 11:41:14 CEST 2011
armv5tel GNU/Linux
4/5 iterations
WARNING: Most functions require kernel 2.6
policy: fifo: loadavg: 5.38 2.57 1.14 9/46 425
T: 0 ( 404) P:80 I:1000 C: 28989 Min: 15 Act: 141 Avg: 278 Max:
49660
T: 1 ( 405) P:79 I:1500 C: 19326 Min: 15 Act: 340 Avg: 342 Max:
50406
T: 2 ( 406) P:78 I:2000 C: 14492 Min: 16 Act: 395 Avg: 378 Max:
49655
T: 3 ( 407) P:77 I:2500 C: 11594 Min: 15 Act: 466 Avg: 334 Max:
50703
T: 4 ( 408) P:76 I:3000 C: 9661 Min: 16 Act: 214 Avg: 295 Max:
50687
T: 5 ( 409) P:75 I:3500 C: 8280 Min: 16 Act: 223 Avg: 337 Max:
52270
T: 6 ( 410) P:74 I:4000 C: 7247 Min: 16 Act: 370 Avg: 353 Max:
52213
T: 7 ( 411) P:73 I:4500 C: 6442 Min: 16 Act: 187 Avg: 313 Max:
52567
T: 8 ( 412) P:72 I:5000 C: 5798 Min: 16 Act: 336 Avg: 373 Max:
49518
T: 9 ( 413) P:71 I:5500 C: 5270 Min: 17 Act: 179 Avg: 401 Max:
51195
1/5 iterations
WARNING: Most functions require kernel 2.6
policy: fifo: loadavg: 3.30 1.93 0.89 6/39 386
T: 0 ( 374) P:80 I:1000 C: 28067 Min: 16 Act: 104 Avg: 168 Max:
492
T: 1 ( 375) P:79 I:1500 C: 18713 Min: 19 Act: 233 Avg: 162 Max:
535
T: 2 ( 376) P:78 I:2000 C: 14035 Min: 24 Act: 54 Avg: 238 Max:
653
T: 3 ( 377) P:77 I:2500 C: 11601 Min: 24 Act: 205 Avg: 245 Max:
597
T: 4 ( 378) P:76 I:3000 C: 9668 Min: 26 Act: 135 Avg: 382 Max:
703
T: 5 ( 379) P:75 I:3500 C: 8287 Min: 24 Act: 40 Avg: 248 Max:
609
T: 6 ( 380) P:74 I:4000 C: 7251 Min: 25 Act: 138 Avg: 198 Max:
592
T: 7 ( 381) P:73 I:4500 C: 6446 Min: 25 Act: 114 Avg: 255 Max:
683
T: 8 ( 382) P:72 I:5000 C: 5800 Min: 34 Act: 204 Avg: 274 Max:
630
T: 9 ( 383) P:71 I:5500 C: 5273 Min: 26 Act: 108 Avg: 284 Max:
682
--
Best regards,
Hinko Kocevar
^ permalink raw reply [flat|nested] 4+ messages in thread
* RE: Poor cycletest results for rt6
2011-07-29 11:10 Poor cycletest results for rt6 Hinko Kocevar
@ 2011-07-29 11:44 ` N, Mugunthan V
2011-07-29 11:46 ` N, Mugunthan V
2011-07-29 14:33 ` Hinko Kocevar
2 siblings, 0 replies; 4+ messages in thread
From: N, Mugunthan V @ 2011-07-29 11:44 UTC (permalink / raw)
To: Hinko Kocevar, linux-rt-users@vger.kernel.org
Hi,
I've build linux-3.0-rt6 for AM3517 and run the cyclic test with the following commands
./hackbench -P -l -1 -g 10 &
./cyclictest -t1 -p80 -n -h 400
---
Regards,
Mugunthan V N.
^ permalink raw reply [flat|nested] 4+ messages in thread
* RE: Poor cycletest results for rt6
2011-07-29 11:10 Poor cycletest results for rt6 Hinko Kocevar
2011-07-29 11:44 ` N, Mugunthan V
@ 2011-07-29 11:46 ` N, Mugunthan V
2011-07-29 14:33 ` Hinko Kocevar
2 siblings, 0 replies; 4+ messages in thread
From: N, Mugunthan V @ 2011-07-29 11:46 UTC (permalink / raw)
To: Hinko Kocevar, linux-rt-users@vger.kernel.org
Hi,
I've build linux-3.0-rt6 for AM3517 and run the cyclic test with the
following commands
./hackbench -P -l -1 -g 10 &
./cyclictest -t1 -p80 -n -h 400
The test results are poot with the max latency of 5231
---
Regards,
Mugunthan V N.
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: Poor cycletest results for rt6
2011-07-29 11:10 Poor cycletest results for rt6 Hinko Kocevar
2011-07-29 11:44 ` N, Mugunthan V
2011-07-29 11:46 ` N, Mugunthan V
@ 2011-07-29 14:33 ` Hinko Kocevar
2 siblings, 0 replies; 4+ messages in thread
From: Hinko Kocevar @ 2011-07-29 14:33 UTC (permalink / raw)
To: linux-rt-users
On 07/29/2011 01:10 PM, Hinko Kocevar wrote:
> Hi,
>
> I've build linux-3.0-rt6 for my ARM platform (XCEP). Running cycletest
> on vanilla, preempt and rt6 kernel build reviles the following results.
> I would expect better rt6 performance.
>
> Note that last test actually looks quite promising, but it happened once
> in 5 test runs :(
Mike Galbraith pointed me into the right direction with the heads up
about the ksoftirqd/0 process that hogged the CPU for a short moment,
enough to make cycletest 'Max' column numbers high.
Reading some googled results about the ksoftirqd made me suspicious
about the network traffic
-http://www.gossamer-threads.com/lists/quagga/dev/17725
Since I'm running my board of the NFS rootfs I opted for using -q switch
with the cycletest. The result was 5/5 test runs with best 'Max' column :)
$ ssh root@192.168.11.132 /shared/utils/cyclictest -t 10 -p 80 -n -D 30 -q
WARNING: Most functions require kernel 2.6
T: 0 ( 707) P:80 I:1000 C: 29001 Min: 16 Act: 129 Avg: 203 Max:
504
T: 1 ( 708) P:79 I:1500 C: 19333 Min: 16 Act: 217 Avg: 204 Max:
602
T: 2 ( 709) P:78 I:2000 C: 14500 Min: 16 Act: 89 Avg: 163 Max:
575
T: 3 ( 710) P:77 I:2500 C: 11600 Min: 16 Act: 195 Avg: 222 Max:
590
T: 4 ( 711) P:76 I:3000 C: 9666 Min: 16 Act: 395 Avg: 358 Max:
697
T: 5 ( 712) P:75 I:3500 C: 8285 Min: 16 Act: 115 Avg: 263 Max:
767
T: 6 ( 713) P:74 I:4000 C: 7249 Min: 16 Act: 265 Avg: 262 Max:
637
T: 7 ( 714) P:73 I:4500 C: 6443 Min: 17 Act: 74 Avg: 220 Max:
734
T: 8 ( 715) P:72 I:5000 C: 5797 Min: 17 Act: 353 Avg: 294 Max:
730
T: 9 ( 716) P:71 I:5500 C: 5264 Min: 18 Act: 207 Avg: 264 Max:
696
--
Hinko Kocevar
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2011-07-29 14:33 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2011-07-29 11:10 Poor cycletest results for rt6 Hinko Kocevar
2011-07-29 11:44 ` N, Mugunthan V
2011-07-29 11:46 ` N, Mugunthan V
2011-07-29 14:33 ` Hinko Kocevar
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.