[Discuss OpenCPI] Duration for Unit Test on Zed

Miller, Peter PeterM at signalscape.com
Wed Jun 27 17:38:17 EDT 2018


Did you guys respond to this? If my unit -test.xml file has:
<case OnlyPlatforms='xsim' duration='180'>
<input port='in' file="../../../test_vectors/PassbandIQ_NonHT_MCS0_Fs20MHz.rx1.raw" />
<output port='out' file='../../../test_vectors/PassbandIQ_NonHT_MCS0_Fs20MHz_RxOutVec.bin' />
</case>
<case OnlyPlatforms='zed' duration='1'>
<input port='in' file="../../../test_vectors/PassbandIQ_NonHT_MCS0_Fs20MHz.rx1.raw" />
<output port='out' file='../../../test_vectors/PassbandIQ_NonHT_MCS0_Fs20MHz_RxOutVec.bin' />
</case>

Running "make runonly" will cause xsim and zed each to run on both case 00 and case 01
Running "make runonly OnlyPlatform="zed" Cases="case00.00" " will run case00 on zed
So the " case OnlyPlatforms" element in the xml appears to be ignored.


-----Original Message-----
From: Miller, Peter
Sent: Monday, June 25, 2018 1:42 PM
To: 'James Kulp' <jek at parera.com>; discuss at lists.opencpi.org
Subject: RE: [Discuss OpenCPI] Duration for Unit Test on Zed

Understood. When generating at 20M complex samples/sec. continuously you get 80MB per sec. When I left duration=120 (for xsim) I got a big file out when run on zed. I can stick nanosleep() in the ACI but ocpirun was where this came from.

-----Original Message-----
From: discuss <discuss-bounces at lists.opencpi.org> On Behalf Of James Kulp
Sent: Monday, June 25, 2018 1:34 PM
To: discuss at lists.opencpi.org
Subject: Re: [Discuss OpenCPI] Duration for Unit Test on Zed

Just to be clear, the "duration" and "timeout" are seconds since they are meant to be a gross measure of time from the top level of the testing framework, which may even be a system that is distant from the actual execution.

I.e. it is rare that all the machinery being involved are fast enough for the difference between 1 second and 1ms  and 1 us to matter much.

That said, we could certainly change the data type to a float and allow sub-second times.

Having a timeout that corresponds to actual real-time execution is different and harder.




On 6/25/18 9:40 AM, Miller, Peter wrote:
> Thank you, sir, that will do the trick! Perhaps for the next release of AV you might consider changing duration to microseconds.
>
> From: aponchak <aponchak at office.geontech.com>
> Sent: Monday, June 25, 2018 9:25 AM
> To: Miller, Peter <PeterM at signalscape.com>
> Cc: discuss at lists.opencpi.org
> Subject: Re: [Discuss OpenCPI] Duration for Unit Test on Zed
>
>
> 1) duration and timeout are size_t (unsigned integer types), so the smallest possible is 1 sec. This variables are defined in ocpigen/src/tests.cxx, so you might try changing them to support your needs, < 1 sec.
>
> 2) Reference CDG, Section 13.3.5. Create a unique case for each platform and limit the duration.
>
> <tests>
>
> ----global (for all test cases): I/O and properties----
>
> <case OnlyPlatforms='xsim' Duration='120'>
>    ----case: I/O and properties----
> </case>
>
> <case OnlyPlatforms='zed' Duration='1'>
>    ----case: I/O and properties----
> </case>
>
> </tests>
> On 2018-06-25 08:30, Miller, Peter <PeterM at signalscape.com<mailto:PeterM at signalscape.com>> wrote:
>
> We have a unit test for a component (phy_tx_802dot11g.test) to run on xsim and on zed. The component does not emit a ZLM and so we use duration to limit the test. In xsim this is 120 seconds <case duration='120'> But for zed we need to limit it to about 100 usec. It appears that duration is in integer seconds as case duration='0.001' runs forever and the file gets big fast.
>
>
>
>
>
>    1.  How can we limit the duration of a unit test in microsec for zed?
>
>    2.  How can we write a unit tester with cases specific to zed and to xsim?
>
>
>
> Sincerely,
>
> Peter B. Miller
>
> Potomac: (301) 765-9668
>
>
>
> --------------------------- This email and any files transmitted with it are confidential and intended solely for the use of Signalscape, Inc. and the addressed individual or entity. If you have received this email in error please delete it. Information in this email may be subject to the Privacy Act of 1974 and any unauthorized review, use, disclosure, or distribution is strictly prohibited. Any views or opinions presented in this email are solely those of the author and do not necessarily represent those of the company.
>
> -------------- next part --------------
>
> An HTML attachment was scrubbed...
>
> URL:
> <http://lists.opencpi.org/pipermail/discuss_lists.opencpi.org/attachme
> nts/20180625/5b59f618/attachment.html>
>
> _______________________________________________
>
> discuss mailing list
>
> discuss at lists.opencpi.org<mailto:discuss at lists.opencpi.org>
>
> http://lists.opencpi.org/mailman/listinfo/discuss_lists.opencpi.org
>
> --------------------------- This email and any files transmitted with it are confidential and intended solely for the use of Signalscape, Inc. and the addressed individual or entity. If you have received this email in error please delete it. Information in this email may be subject to the Privacy Act of 1974 and any unauthorized review, use, disclosure, or distribution is strictly prohibited. Any views or opinions presented in this email are solely those of the author and do not necessarily represent those of the company.
> -------------- next part -------------- An HTML attachment was
> scrubbed...
> URL:
> <http://lists.opencpi.org/pipermail/discuss_lists.opencpi.org/attachme
> nts/20180625/1003ec18/attachment.html>
> _______________________________________________
> discuss mailing list
> discuss at lists.opencpi.org
> http://lists.opencpi.org/mailman/listinfo/discuss_lists.opencpi.org



_______________________________________________
discuss mailing list
discuss at lists.opencpi.org
http://lists.opencpi.org/mailman/listinfo/discuss_lists.opencpi.org
--------------------------- This email and any files transmitted with it are confidential and intended solely for the use of Signalscape, Inc. and the addressed individual or entity. If you have received this email in error please delete it. Information in this email may be subject to the Privacy Act of 1974 and any unauthorized review, use, disclosure, or distribution is strictly prohibited. Any views or opinions presented in this email are solely those of the author and do not necessarily represent those of the company.


More information about the discuss mailing list