[Xenomai-help] latency in kernel mode

[Xenomai-help] latency in kernel mode

[Joachim Meyer]

Hi

I still don't know what to do with the latency test in kernel mode. Every others test works fine except for the IRQ test, but I was told this is normal for an non x86 architecture. Can anybody give me a hint what I can do? What would probably be the problem? How important is it, that that test won't work? Should I ignore it and start with programming some real-time-stuff? But if I start now and get problems later I would always think, the failed test could be the reason for the problem too....

Here's the output:
---------------------------------------------------------------------------------------------
bash-3.00# cd /usr/xenomai/testsuite/latency/
bash-3.00# ./run -- -t1
*
*
* Type ^C to stop this application.
*
*
== Sampling period: 100 us
== Test mode: in-kernel periodic task
== All results in microseconds
latency: failed to start in-kernel timer benchmark, code -25
---|------------|------------|------------|--------|-------------------------
RTS|  267928.028|  805352.732|      32.768|       0|    00:00:10/00:00:10
bash-3.00#
----------------------------------------------------------------------------------------------

Thanks for your help and patience.
Joachim
____________________________________________________________________
Ihre Messenger, Communities und E-Mails jetzt in einem Programm!
WEB.DE MultiMessenger http://www.produkte.web.de/messenger/?did=3071

Re: [Xenomai-help] latency in kernel mode

[Gilles Chanteperdrix]

Joachim Meyer wrote:
> bash-3.00# ./run -- -t1
> *
> *
> * Type ^C to stop this application.
> *
> *
> == Sampling period: 100 us
> == Test mode: in-kernel periodic task
> == All results in microseconds
> latency: failed to start in-kernel timer benchmark, code -25
> ---|------------|------------|------------|--------|-------------------------
> RTS|  267928.028|  805352.732|      32.768|       0|    00:00:10/00:00:10
> bash-3.00#
> ----------------------------------------------------------------------------------------------
> 
> Thanks for your help and patience.

Enable the timerbench driver in kernel configuration. If you want to
enable several testing drivers and everything to work "out of thebox",
enable them as module, and do not load the modules by yourself, let the
run script load the apropriate module. Otherwise, specify the correct
testing driver index to the test you run.

-- 


					    Gilles.

Re: [Xenomai-help] latency in kernel mode

[Breno Carneiro Pinheiro]

[-- Attachment #1: Type: text/plain, Size: 1699 bytes --]

Hi guys, I wonder do the same test as described below on my system. Where's
the timerbench driver? I was looking for this option but I couldn't find it
on my .conifg.
Is it necessary to do what Joachim said at previous email?:

" I enabled the timerbench module, and installed it with the other modules
(IRQ Benchmark and Context Switch).
The Switchtest test works, so I think the modules are correct installed and
the tests are able to load them...."

Thanks,

Breno

2008/6/25 Gilles Chanteperdrix <gilles.chanteperdrix@xenomai.org>:

> Joachim Meyer wrote:
> > bash-3.00# ./run -- -t1
> > *
> > *
> > * Type ^C to stop this application.
> > *
> > *
> > == Sampling period: 100 us
> > == Test mode: in-kernel periodic task
> > == All results in microseconds
> > latency: failed to start in-kernel timer benchmark, code -25
> >
> ---|------------|------------|------------|--------|-------------------------
> > RTS|  267928.028|  805352.732|      32.768|       0|    00:00:10/00:00:10
> > bash-3.00#
> >
> ----------------------------------------------------------------------------------------------
> >
> > Thanks for your help and patience.
>
> Enable the timerbench driver in kernel configuration. If you want to
> enable several testing drivers and everything to work "out of thebox",
> enable them as module, and do not load the modules by yourself, let the
> run script load the apropriate module. Otherwise, specify the correct
> testing driver index to the test you run.
>
> --
>
>
>                                            Gilles.
>
> _______________________________________________
> Xenomai-help mailing list
> Xenomai-help@domain.hid
> https://mail.gna.org/listinfo/xenomai-help
>

[-- Attachment #2: Type: text/html, Size: 2501 bytes --]

Re: [Xenomai-help] latency in kernel mode

[Gilles Chanteperdrix]

Breno Carneiro Pinheiro wrote:
> Hi guys, I wonder do the same test as described below on my system.
> Where's the timerbench driver? I was looking for this option but I
> couldn't find it on my .conifg.

As documented in section 1.2 of README.INSTALL, all Xenomai options are
available under the menu "Real-time susbsytem" in Linux kernel
configuration.

> Is it necessary to do what Joachim said at previous email?:
> 
> " I enabled the timerbench module, and installed it with the other
> modules (IRQ Benchmark and Context Switch). The Switchtest test
> works, so I think the modules are correct installed and the tests are
> able to load them...."

If you mean install the modules then no, you do not need to install
Xenomai modules, but if you do not do so, you will have to pass their
full path to insmod, and Xenomai "run" scripts will not be able to find
them by themselves.

-- 


					    Gilles.

Re: [Xenomai-help] latency in kernel mode

[Sebastian Smolorz]

Joachim Meyer wrote:
> I still don't know what to do with the latency test in kernel mode.
[...]
> Here's the output:
> ---------------------------------------------------------------------------------------------
> bash-3.00# cd /usr/xenomai/testsuite/latency/
> bash-3.00# ./run -- -t1
> *
> *
> * Type ^C to stop this application.
> *
> *
> == Sampling period: 100 us
> == Test mode: in-kernel periodic task
> == All results in microseconds
> latency: failed to start in-kernel timer benchmark, code -25

Did you compile and load the kernel driver xeno_timerbench.ko?

-- 
Sebastian

Re: [Xenomai-help] latency in kernel mode

[Joachim Meyer]

Sorry, for my late answer.

I enabled the timerbench module, and installed it with the other modules (IRQ Benchmark and Context Switch).
The Switchtest test works, so I think the modules are correct installed and the tests are able to load them.
Does this mean the timerbench module isn't ok? Or are there other possible reasons, why the latency test still says:

bash-3.00# ./run -- -t 1
*
*
* Type ^C to stop this application.
*
*
== Sampling period: 100 us
== Test mode: in-kernel periodic task
== All results in microseconds
latency: failed to start in-kernel timer benchmark, code -25
---|------------|------------|------------|--------|-------------------------
RTS|  267928.028|  805352.732|      32.768|       0|    00:00:07/00:00:07
bash-3.00#

I just want to add that I'm using a ppc405 on an Xilinx Virtex2pro (XUP Board), Linux v2.6.23xlnx, a rootFS from eldk4.1 and Xenomai 2.4.2.

Are there any other importent options in compiling the kernel. I thought perhaps there are good default values, which will do it for me in the beginning. 
My Real-time sub-system config: 
[*] Xenomai
 <*> Nucleus
 [*] Pervasive real-time support in user-space
 [ ] Interuppt shield support
 [*] Priority coupling support
 [*] optimize as pipeline head
 (32) number of pipe devices
 (512) Number of registry slots
 (128) Size of the system heap (Kb)
 [*] Statistics collection
 [ ] Debug Support
 [ ] Shared interuppts
 Timing --> [ ] Enable periodic timing
                  (0) Scheduling latency (ns)
 Scalability --> [ ] 0(1) scheduler
                       Timer indexing method (Linear) --> (X) Linear
                                                                               ( ) Tree
                                                                               ( ) Hash
Machine --> No options!
Interfaces --> Not in the mood to print all of this, but if it would help, i 'would do it.
Drivers --> only the three testing drivers as modules

Another Question I have is about the Output of my switchbench test:

bash-3.00# ./run -- -n 5000
*
*
* Type ^C to stop this application.
*
*
== Sampling period: 100 us
== Do not interrupt this program
RTH|       lat min|     lat avg|     lat max|        lost
RTD|      -29560|        5810|       49723|        4352
bash-3.00#

Is it normal that the lost number is that high? If not, what can i do?

Thanks for all your help.
Joachim
_________________________________________________________________________
In 5 Schritten zur eigenen Homepage. Jetzt Domain sichern und gestalten! 
Nur 3,99 EUR/Monat! http://www.maildomain.web.de/?mc=021114

Re: [Xenomai-help] latency in kernel mode

[Wolfgang Grandegger]

Joachim Meyer wrote:
> Sorry, for my late answer.
> 
> I enabled the timerbench module, and installed it with the other modules (IRQ Benchmark and Context Switch).
> The Switchtest test works, so I think the modules are correct installed and the tests are able to load them.
> Does this mean the timerbench module isn't ok? Or are there other possible reasons, why the latency test still says:
> 
> bash-3.00# ./run -- -t 1
> *
> *
> * Type ^C to stop this application.
> *
> *
> == Sampling period: 100 us
> == Test mode: in-kernel periodic task
> == All results in microseconds
> latency: failed to start in-kernel timer benchmark, code -25
> ---|------------|------------|------------|--------|-------------------------
> RTS|  267928.028|  805352.732|      32.768|       0|    00:00:07/00:00:07
> bash-3.00#
> 

What does the following command show:

   # cat /proc/xenomai/rtdm/rttest*/information

You may need to use the option "-D 1".

> I just want to add that I'm using a ppc405 on an Xilinx Virtex2pro (XUP Board), Linux v2.6.23xlnx, a rootFS from eldk4.1 and Xenomai 2.4.2.
> 
> Are there any other importent options in compiling the kernel. I thought perhaps there are good default values, which will do it for me in the beginning. 
> My Real-time sub-system config: 
> [*] Xenomai
>  <*> Nucleus
>  [*] Pervasive real-time support in user-space
>  [ ] Interuppt shield support
>  [*] Priority coupling support
>  [*] optimize as pipeline head
>  (32) number of pipe devices
>  (512) Number of registry slots
>  (128) Size of the system heap (Kb)
>  [*] Statistics collection
>  [ ] Debug Support
>  [ ] Shared interuppts
>  Timing --> [ ] Enable periodic timing
>                   (0) Scheduling latency (ns)
>  Scalability --> [ ] 0(1) scheduler
>                        Timer indexing method (Linear) --> (X) Linear
>                                                                                ( ) Tree
>                                                                                ( ) Hash
> Machine --> No options!
> Interfaces --> Not in the mood to print all of this, but if it would help, i 'would do it.
> Drivers --> only the three testing drivers as modules

The default values are good to start with.

> Another Question I have is about the Output of my switchbench test:
> 
> bash-3.00# ./run -- -n 5000
> *
> *
> * Type ^C to stop this application.
> *
> *
> == Sampling period: 100 us
> == Do not interrupt this program
> RTH|       lat min|     lat avg|     lat max|        lost
> RTD|      -29560|        5810|       49723|        4352
> bash-3.00#
> 
> Is it normal that the lost number is that high? If not, what can i do?

A period of 100us is tough for your system. What worst case latencies do 
you measure with the latency program under load. It will be more than 
100us, I guess.

Wolfgang.

Re: [Xenomai-help] latency in kernel mode

[Sebastian Smolorz]

Joachim Meyer wrote:
> Sorry, for my late answer.
> 
> I enabled the timerbench module, and installed it with the other modules (IRQ Benchmark and Context Switch).
> The Switchtest test works, so I think the modules are correct installed and the tests are able to load them.
> Does this mean the timerbench module isn't ok? Or are there other possible reasons, why the latency test still says:
> 
> bash-3.00# ./run -- -t 1
> *
> *
> * Type ^C to stop this application.
> *
> *
> == Sampling period: 100 us
> == Test mode: in-kernel periodic task
> == All results in microseconds
> latency: failed to start in-kernel timer benchmark, code -25

-ENOTTY. This is strange because it means that the IO-Control 
RTTST_RTIOC_TMBENCH_START is unknown by the timerbench driver. Do you 
happen to use different versions of Xenomai in kernel and userland? It 
could also help if you modify your latency and timerbench code to print 
out the values of RTTST_RTIOC_TMBENCH_START in order to detect a 
discrepancy between kernel and userland. If you don't know how to do 
this please tell us so we can provide you with a patch.

-- 
Sebastian

Re: [Xenomai-help] latency in kernel mode

[Joachim Meyer]

Hi

You were right, I really used different versions.....I fixed it now, so I use only 2.4.2. 

Latency:
-----------
bash-3.00# ./run -- -t 1
./run: line 1: dirname: command not found
*
*
* Type ^C to stop this application.
*
*
== Sampling period: 100 us
== Test mode: in-kernel periodic task
== All results in microseconds
warming up...
RTT|  00:00:01  (in-kernel periodic task, 100 us period, priority 99)
RTH|-----lat min|-----lat avg|-----lat max|-overrun|----lat best|---lat worst
RTD|       4.270|       6.190|      32.050|       0|       4.270|      32.050
RTD|       4.270|       6.238|      50.750|       0|       4.270|      50.750
RTD|       4.270|       6.175|      49.760|       0|       4.270|      50.750
RTD|       4.270|       6.206|      63.770|       0|       4.270|      63.770
RTD|       4.270|       6.210|      48.220|       0|       4.270|      63.770
---|------------|------------|------------|--------|-------------------------
RTS|       4.270|       6.203|      63.770|       0|    00:00:06/00:00:06
bash-3.00#

Switchbench:
-----------------
bash-3.00# ./run -- -n 5000 -p 200
./run: line 1: dirname: command not found
*
*
* Type ^C to stop this application.
*
*
== Sampling period: 200 us
== Do not interrupt this program
RTH|     lat min|     lat avg|     lat max|        lost
RTD|      49.070|      53.050|      79.350|           1
bash-3.00#


Thank you all.

I think now, that all test work fine I can start programm my RT programms. I you have any recommendations of things I should do in addition to the tests of the testsuite, please feel free to mail.

Thanks all again 
Greez
Joachim 

_______________________________________________________________________
EINE FÜR ALLE: die kostenlose WEB.DE-Plattform für Freunde und Deine
Homepage mit eigenem Namen. Jetzt starten! http://unddu.de/?kid=kid@domain.hid

Re: [Xenomai-help] latency in kernel mode

[Joachim Meyer]

Hi

The exact Place (at my 2.6.23 Kernel and 2.4.2 Xenomai) in the config-menu is:
---> Real-time sub-system
        ---> Drivers
                ---> Testing drivers
Here you find the 3 drivers you should best compile as Modules and install to the proper place so you can run the tests easily with the run scripts.

Greez
Joachim
_________________________________________________________________
WEB.DE schenkt Ihnen jeden Monat einen hochkarätigen Blockbuster 
von maxdome! Jetzt anmelden unter http://www.blockbuster.web.de

Re: [Xenomai-help] latency in kernel mode

[Joachim Meyer]

Hi

Sorry my answer took so long. I appreciate your offer. But I don't know mcuh about Xenomai, so the test results doesn't say much to me, perhaps you can tell me more. I was more interested to verify with the tests, that I installed Xenomai right. Here the output of some tests. Most times I used default values because, like I said before, I don't know the purpose of most tests:

I use a XUPv2p Board with an Virtex II pro (XC2VP30)
-------------------------------------------
bash-3.00# cat /proc/cpuinfo
processor       : 0
cpu             : Virtex-II Pro
clock           : 300MHz
revision        : 8.160 (pvr 2001 08a0)
bogomips        : 297.98
machine         : Xilinx XUPV2P
plb bus clock   : 100MHz
bash-3.00#
------------------------------------------

clocktest
-------------------------------------------
bash-3.00# cd clocktest/ && ./run -- -T 10 && cd ..
./run: line 1: dirname: command not found
*
*
* Type ^C to stop this application.
*
*
== Tested clock: 0 (CLOCK_REALTIME)
CPU      ToD offset [us] ToD drift [us/s]      warps max delta [us]
--- -------------------- ---------------- ---------- --------------
  0             817999.8           -0.002          0            0.0
bash-3.00#
---------------------------------------------

cyclic
----------------------------------------
bash-3.00# cd cyclic/ && ./run -- -l 50000 && cd ..
./run: line 1: dirname: command not found
*
*
* Type ^C to stop this application.
*
*
0.02 0.07 0.08 1/22 2006

T: 0 ( 2006) P:99 I:    1000 C:   49215 Min:      31 Act:     110 Avg:      43 Max:     145
bash-3.00#
--------------------------------------------


latency in both modes
------------------------------------------------------------
bash-3.00# cd latency && ./run -- -T 30 && cd ..
./run: line 1: dirname: command not found
*
*
* Type ^C to stop this application.
*
*
== Sampling period: 100 us
== Test mode: periodic user-mode task
== All results in microseconds
warming up...
RTT|  00:00:01  (periodic user-mode task, 100 us period, priority 99)
RTH|-----lat min|-----lat avg|-----lat max|-overrun|----lat best|---lat worst
RTD|      18.403|      19.820|      63.993|       0|      18.403|      63.993
RTD|      17.163|      20.216|      95.543|       0|      17.163|      95.543
RTD|      18.383|      20.016|      95.523|       0|      17.163|      95.543
RTD|      16.633|      21.550|      74.733|       0|      16.633|      95.543
RTD|      18.333|      20.013|     106.983|       0|      16.633|     106.983
RTD|      18.333|      19.990|      89.173|       0|      16.633|     106.983
RTD|      18.403|      20.003|      81.633|       0|      16.633|     106.983
RTD|      18.363|      19.976|      74.103|       0|      16.633|     106.983
RTD|      18.333|      20.190|      70.833|       0|      16.633|     106.983
RTD|      18.363|      19.980|      69.953|       0|      16.633|     106.983
RTD|      18.333|      20.006|      71.613|       0|      16.633|     106.983
RTD|      18.403|      19.970|      88.343|       0|      16.633|     106.983
RTD|      18.333|      19.996|      71.413|       0|      16.633|     106.983
RTD|      18.383|      20.156|     107.603|       0|      16.633|     107.603
RTD|      18.333|      20.000|      71.473|       0|      16.633|     107.603
RTD|      18.363|      19.993|      74.183|       0|      16.633|     107.603
RTD|      18.333|      20.010|      75.823|       0|      16.633|     107.603
RTD|      18.403|      19.983|      74.233|       0|      16.633|     107.603
RTD|      18.383|      20.193|      75.563|       0|      16.633|     107.603
RTD|      18.403|      19.996|     108.583|       0|      16.633|     108.583
RTD|      18.333|      20.020|      73.563|       0|      16.633|     108.583
RTT|  00:00:22  (periodic user-mode task, 100 us period, priority 99)
RTH|-----lat min|-----lat avg|-----lat max|-overrun|----lat best|---lat worst
RTD|      17.283|      20.020|      87.283|       0|      16.633|     108.583
RTD|      18.333|      20.250|      75.183|       0|      16.633|     108.583
RTD|      18.333|      20.166|      75.213|       0|      16.633|     108.583
RTD|      18.333|      20.020|     109.183|       0|      16.633|     109.183
RTD|      18.333|      20.003|      74.113|       0|      16.633|     109.183
RTD|      18.333|      19.996|      68.613|       0|      16.633|     109.183
RTD|      18.333|      19.983|      74.223|       0|      16.633|     109.183
RTD|      18.333|      20.196|      81.173|       0|      16.633|     109.183
---|------------|------------|------------|--------|-------------------------
RTS|      16.633|      20.093|     109.183|       0|    00:00:30/00:00:30
bash-3.00# cd latency && ./run -- -t 1 -T 30 && cd ..
./run: line 1: dirname: command not found
*
*
* Type ^C to stop this application.
*
*
== Sampling period: 100 us
== Test mode: in-kernel periodic task
== All results in microseconds
warming up...
RTT|  00:00:01  (in-kernel periodic task, 100 us period, priority 99)
RTH|-----lat min|-----lat avg|-----lat max|-overrun|----lat best|---lat worst
RTD|       4.270|       4.961|      32.380|       0|       4.270|      32.380
RTD|       4.270|       5.166|      41.780|       0|       4.270|      41.780
RTD|       4.270|       5.038|      36.790|       0|       4.270|      41.780
RTD|       4.270|       5.880|      43.340|       0|       4.270|      43.340
RTD|       4.270|       5.552|      39.830|       0|       4.270|      43.340
RTD|       4.270|       5.050|      57.710|       0|       4.270|      57.710
RTD|       4.270|       5.050|      44.120|       0|       4.270|      57.710
RTD|       4.270|       5.048|      56.770|       0|       4.270|      57.710
RTD|       4.270|       5.039|      44.170|       0|       4.270|      57.710
RTD|       4.270|       5.528|      43.090|       0|       4.270|      57.710
RTD|       4.270|       5.045|      35.190|       0|       4.270|      57.710
RTD|       4.270|       5.033|      46.260|       0|       4.270|      57.710
RTD|       4.270|       5.040|      39.260|       0|       4.270|      57.710
RTD|       4.270|       5.046|      43.910|       0|       4.270|      57.710
RTD|       4.270|       5.065|      45.320|       0|       4.270|      57.710
RTD|       4.270|       5.030|      42.820|       0|       4.270|      57.710
RTD|       4.270|       5.043|      42.790|       0|       4.270|      57.710
RTD|       4.270|       5.035|      42.440|       0|       4.270|      57.710
RTD|       4.270|       5.047|      44.110|       0|       4.270|      57.710
RTD|       4.270|       5.057|      40.040|       0|       4.270|      57.710
RTD|       4.270|       5.033|      52.410|       0|       4.270|      57.710
RTT|  00:00:22  (in-kernel periodic task, 100 us period, priority 99)
RTH|-----lat min|-----lat avg|-----lat max|-overrun|----lat best|---lat worst
RTD|       4.270|       5.041|      45.520|       0|       4.270|      57.710
RTD|       4.270|       5.143|      44.180|       0|       4.270|      57.710
RTD|       4.270|       5.030|      40.780|       0|       4.270|      57.710
RTD|       4.270|       5.542|      31.540|       0|       4.270|      57.710
RTD|       4.270|       5.037|      58.200|       0|       4.270|      58.200
RTD|       4.270|       5.054|      39.280|       0|       4.270|      58.200
RTD|       4.270|       5.039|      34.010|       0|       4.270|      58.200
RTD|       4.270|       5.043|      41.780|       0|       4.270|      58.200
---|------------|------------|------------|--------|-------------------------
RTS|       4.270|       5.128|      58.200|       0|    00:00:30/00:00:30
bash-3.00#
---------------------------------------------------------------------------------


switchbench
-----------------------------------------------------------------------------------
bash-3.00# cd switchbench/ && ./run -- -p 200 -n 100000 && cd
./run: line 1: dirname: command not found
*
*
* Type ^C to stop this application.
*
*
== Sampling period: 200 us
== Do not interrupt this program
RTH|     lat min|     lat avg|     lat max|        lost
RTD|      43.916|      54.773|      91.860|           1
bash-3.00#
--------------------------------------------------------------------------------------


switchtest
----------------------------------------------------------------------------------
bash-3.00# cd switchtest/ && ./run -- -T 30 -n && cd ..
./run: line 1: dirname: command not found
*
*
* Type ^C to stop this application.
*
*
== Threads: sleeper-0 rtk-1 rtk-2 rtup-3 rtup-4 rtus-5 rtus-6 rtuo-7 rtuo-8
RTT|  00:00:01
RTH|ctx switches|-------total
RTD|        1791|        1791
RTD|        1782|        3573
RTD|        1791|        5364
RTD|        1692|        7056
RTD|        1800|        8856
RTD|        1791|       10647
RTD|        1785|       12432
RTD|        1815|       14247
RTD|        1800|       16047
RTD|        1800|       17847
RTD|        1800|       19647
RTD|        1800|       21447
RTD|        1800|       23247
RTD|        1800|       25047
RTD|        1800|       26847
RTD|        1800|       28647
RTD|        1785|       30432
RTD|        1800|       32232
RTD|        1797|       34029
RTD|        1800|       35829
RTD|        1800|       37629
RTT|  00:00:22
RTH|ctx switches|-------total
RTD|        1800|       39429
RTD|        1800|       41229
RTD|        1800|       43029
RTD|        1800|       44829
RTD|        1800|       46629
RTD|        1791|       48420
RTD|        1800|       50220
RTD|        1800|       52020
RTD|        1584|       53604
bash-3.00#
------------------------------------------------------------------------------------

Greez
Joachim
_____________________________________________________________________
Der WEB.DE SmartSurfer hilft bis zu 70% Ihrer Onlinekosten zu sparen!
http://smartsurfer.web.de/?mc=100071&distributionid=000000000066