RE: OOT blocks in 3.8 not working

[Robert Meyer]

Thank you Yun,

That didn't solve my problem, but I think there were some errors within the
YAML.  I used www.yamllint.com in order to make sure that all my blocks were
valid VAML files, since I am not as familiar with them.

I did make sure to run sudo ldconfig also


Something else that I noticed that was interesting:

In the python interpreter under the build/swig directory, I tried importing
different modules to see what would work and what wouldn't. 

>>>import gnuradio
>>>from gnuradio import gr
>>>import nuclear
/home/superuser/prefix/default/lib/libgnuradio-nuclear.so.1.0.0.git:
undefined smbol:
_ZN2gr7nuclear15radio_init_impl12ReceiveEventElN11EventObject19t_Notificatio
nTypesEii
>>>import nuclear
>>>nuclear.<tab to auto complete>
>>>nuclear.unicode_literals
_Feature((2, 6, 0, 'alpha', 2), (3, 0, 0, 'alpha', 0), 131072)
>>>nuclear
<module 'nuclear' from
'/home/superuser/prefix/default/lib/python3.6/dist-packages/nuclear/__init__
.py'>
>>>import howto
>>>howto.<tab to auto complete>
howto.high_res_timer_epoch(             howto.howto_swig
howto.square_ff_sptr_swigregister(
howto.high_res_timer_now(               howto.square_ff(
howto.square_ff_swigregister(
howto.high_res_timer_now_perfmon(       howto.square_ff_make(
howto.unicode_literals
howto.high_res_timer_tps(               howto.square_ff_sptr(   
>>>howto
<module 'howto' from
'/home/superuser/prefix/default/lib/python3.6/dist-packages/nuclear/__init__
.py'>
>>>howto.howto_swig
<module 'howto.howto_swig' from
'/home/superuser/prefix/default/lib/python3.6/dist-packages/howto/howto_swig
.py'>
>>>howto.unicode_literals
_Feature((2, 6, 0, 'alpha', 2), (3, 0, 0, 'alpha', 0), 131072)


It seems that maybe my nuclear_swig isn't getting generated, has errors, or
is being placed somewhere it can't be found?


-----Original Message-----
From: address@hidden <address@hidden> 
Sent: Thursday, December 12, 2019 1:21 AM
To: 'Robert Meyer' <address@hidden>
Subject: RE: OOT blocks in 3.8 not working

Hey Robert

Have you confirmed the yaml file of your OOT block?

I also suffered it when I start to use 3.8, as they changed xml to yaml.
It will not generate the necessary inputs and outputs in yaml file for you.

Or did you operate "sudo ldconfig" command after make install of the block? 

Best Regards
Yun

> -----Original Message-----
> From: Discuss-gnuradio
> [mailto:discuss-gnuradio-bounces+wen.yun=address@hidden] On 
> Behalf Of Robert Meyer
> Sent: Thursday, December 12, 2019 9:51 AM
> To: address@hidden
> Subject: OOT blocks in 3.8 not working
> 
> Hello,
> 
> 
> 
> I am trying move a project I have from gnuradio 3.7 to 3.8 and I am 
> having some difficulties.
> 
> 
> 
> I have went through the gr_modtool tutorial in 3.8, and was able to 
> get the square_fft block to build and run inside of gr-howto, but 
> making the same block in my module (gr-nuclear) doesn't work.
> 
> 
> 
> Some of the errors I suspected were from dependencies, and I think 
> that I have added them in correctly.  Those dependencies are UHD and
> libomnithread:
> 
> 
> 
> 
> 
> To include UHD I:
> 
> *     Added the following to the CMakeLists.txt based on gr-foo: (in the
> gr-nuclear's directory, not in /lib/)
> 
> 
> 
> I added this to the Compiler Specific setup section:
> 
> 
> 
> find_package(UHD)
> 
> if(NOT UHD_FOUND)
> 
>                 message(STATUS "UHD library not found, no support for 
> timed TX")
> 
>                 set(UHD_INCLUDE_DIRS "")
> 
>                 set(UHD_LIBRARY_DIRS "")
> 
>                 set(UHD_LIBRARIES "")
> 
> else()
> 
>                 message(STATUS "UHD library found.")
> 
>                 set(UHD_INCLUDE_DIRS "")
> 
>                 set(UHD_LIBRARY_DIRS "")
> 
>                 set(UHD_LIBRARIES "")
> 
> endif()
> 
> 
> 
> and I added this in the project setup section:
> 
> 
> 
> list(INSERT CMAKE_MODULE_PATH 0 "/path/to/uhd/module")
> 
> 
> 
> 
> 
> *     Added the following to the CMakeList.txt in the /lib/ directory
> 
> 
> 
> Added <${UHD_INCLUDE_DIRS}> to include_directories()
> 
> Added <$UHD_LIBRARY_DIRS}> to link_directories()
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> To include libomnithread4 I;
> 
> *     I had to purge all of libomnithread, and install the latest omniorb
> package.  After that I added the following to the 
> target_link_libraries() list in the CMakeLists.txt file in /lib/
> 
> 
> 
> /usr/lib/x86_64-linux-gnu/libomnithread.so.4
> 
> 
> 
> *     I was getting an error that omnithread.h was not in the include
> folder while building, so I moved omnithread.h to the include/nuclear/ 
> directory, and after that I was able to build the project without errors.
> 
> 
> 
> 
> 
> 
> 
> 
> 
> But once I try to launch my project I get "AttributeError: module
'nuclear'
> has no attribute 'radio_init', which is one of the blocks of my OOT
module.
> 
> 
> 
> I'm not sure if this is related, but while launching I also get the
warning:
> 
> 
> 
> "Warning: restarting the docstring loader (crashed while loading 
> 'uhd_amsg_source')"
> 
> 
> 
> When I go into the build/swig directory and run python3.6, I tried to 
> import my module the same way that __init__.py uses it and I get this 
> error
> 
> 
> 
> How I imported:
> 
>                 from .nuclear_swig import *
> 
> 
> 
> Error I get:
> 
> ModuleNotFoundError: No module named '__main__.nuclear_swig'; '__main__'
> is not a package
> 
> 
> 
> 
> 
> I have tried removing the '.' Which was recommended when I looked into 
> this problem online, but that did not solve it
> 
> 
> 
> 
> 
> Before include the two modules, trying to import .nuclear_swig was 
> segfaulting, so it seems like I have made at least some progress.
> 
> 
> 
> 
> 
> Does anyone have some ideas of what I could be missing/doing wrong?
> 
> 
> 
> Thank you
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> --
> This message has been scanned for viruses and dangerous content by 
> MailScanner <http://www.mailscanner.info/> , and is believed to be clean.

--
This message has been scanned for viruses and dangerous content by
MailScanner, and is believed to be clean.


-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.