Re: [Discuss-gnuradio] Build GR 3.8 from source

You need scipy for the polar code tests. Not sure about the qa_pmt failure.

sudo apt-get install python3-scipy

Ron

On 7/22/19 13:51, Barry Duggan wrote:
> Hi Michael,
>
> Success! Based on Ron's suggestion, I was able to finish the last half
> of the make in 3 hours.
>
> Then I did a 'make test' and got the following failures:
>
> """
> 98% tests passed, 6 tests failed out of 362
>
> Total Test time (real) = 1465.99 sec
>
> The following tests FAILED:
>     150 - qa_pmt (Failed)
>     243 - qa_polar_decoder_sc (Failed)
>     244 - qa_polar_decoder_sc_list (Failed)
>     245 - qa_polar_decoder_sc_systematic (Failed)
>     246 - qa_polar_encoder (Failed)
>     247 - qa_polar_encoder_systematic (Failed)
> Errors while running CTest
> make: *** [Makefile:86: test] Error 8
> """
>
> I don't think it will impact what I need, but thought you might want
> to know.
>
> I really appreciate all the help you have provided.
>
> Best regards,
> ---
> Barry Duggan
>
>
> On 2019-07-22 09:02, Michael Dickens wrote:
>> Good morning, Barry! Progress ... keep it coming!
>>
>> Reviewing your cmake log: My guess is that you just need to give it
>> longer. A -lot- longer. The point of delay is when C++ is compiling
>> the SWIG-generated Python interface for gr::filter. SWIG C++ code is
>> horrendous to read through, and takes a log time to compile when it of
>> any reasonable amount of complexity. If your RPi can't handle the
>> compilation, it'll crash -- not just hang, but totally crash. Way back
>> with the RPi 1, we did get GR to compile on it ... but it took like 2
>> days! Patience sometimes works here!
>>
>> All of that said: Did you look at Ron's recommendation? The RPi is
>> notoriously difficult to compile GR directly on; sometimes you need to
>> tweak file / memory settings to get "reasonable" compiles.
>>
>> Also: You could try using OpenEmbedded to build GR on your host
>> computer, cross-compiled for your RPi. That would take some work, but
>> might be worth the effort as once it works it would be easily
>> repeatable!
>>
>> Hope this is useful! - MLD
>>
>> On Mon, Jul 22, 2019, at 7:30 AM, Barry Duggan wrote:
>>> Good Morning!
>>>
>>> Well, good news and bad news. The good news is that it got past the 13%
>>> mark with no problems. The bad news is that it got stuck again in an
>>> even tighter loop at 53%. I couldn't even move the cursor or interrupt
>>> it with a control-C. The activity light was on solid.
>>>
>>> I started at 17:05 and at 20:35 it was at 53%, so I projected it should
>>> finish around midnight. I left it running overnight, but it was
>>> still at
>>> 53%. The output didn't capture the last of what was on the screen.
>>>
>>> Let me know what to try next. Thank you for your help.
>>>
>>> Best regards,
>>> ---
>>> Barry Duggan
>>>
>>>
>>> On 2019-07-21 13:41, Michael Dickens wrote:
>>> > Hi Barry - OK; good progress! Can you do the following, from the
>>> > top-level "build" directory (looks like "/home/pi/gnuradio/build"):
>>> > {{{
>>> > make clean
>>> > make VERBOSE=ON 2>&1 | tee make_out.txt
>>> > }}}
>>> > and then walk away & let it go until it's done. You'll be able to see
>>> > the build going along via the 'tee' command. Please be patient! It
>>> > might work; it might not! Either way on an RPi it'll take some
>>> time to
>>> > build whatever it can! - MLD
>>> >
>>>
>>> <snip>
>>>
>>> Attachments:
>>> * make_out.txt
>
> _______________________________________________
> Discuss-gnuradio mailing list
> Discuss-gnuradio@gnu.org
> https://lists.gnu.org/mailman/listinfo/discuss-gnuradio

_______________________________________________
Discuss-gnuradio mailing list
Discuss-gnuradio@gnu.org
https://lists.gnu.org/mailman/listinfo/discuss-gnuradio