# Problem installing new module

**URL:** <https://www.ruby-forum.com/t/problem-installing-new-module/184345>\
**Category:** GNU Radio\
**Created:** [March 1, 2010, 9:56am UTC](https://www.ruby-forum.com/t/problem-installing-new-module/184345 "2010-03-01T09:56:20Z")\
**Posts on this page:** 7\
**Page:** 1

<div class="post-metadata">

**Author:** ![sagar\_k](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/s/e79b87/32.png) [@sagar\_k](https://www.ruby-forum.com/u/sagar_k)\
**Post date:** [March 1, 2010, 9:56am UTC](https://www.ruby-forum.com/t/problem-installing-new-module/184345/1 "2010-03-01T09:56:20Z")

</div>

Hi All,

I am new to GNURadio. I am having trouble installing new module in the  
GNURadio.

I have successfully installed the GNURadio at location

/opt/local/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/site-packages

When i try to install the new module, I follow the following steps.

1. ./bootstrap
2. ./configure
3. make check

After make check I get following error

## ====================================================================== ERROR: test\_001\_sink\_ff ( **main**.qa\_goodney)

Traceback (most recent call last):  
File “./qa\_goodney.py”, line 42, in test\_001\_sink\_ff  
self.tb.run ()  
File  
“/opt/local/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/site-packages/gnuradio/gr/top\_block.py”,  
line 103, in run  
self.start()  
File  
“/opt/local/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/site-packages/gnuradio/gr/top\_block.py”,  
line 97, in start  
self.\_tb.start()  
File  
“/opt/local/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/site-packages/gnuradio/gr/gnuradio\_swig\_py\_runtime.py”,  
line 1455, in start  
return \_gnuradio\_swig\_py\_runtime.gr\_top\_block\_sptr\_start(self)  
RuntimeError: unable to resolve input endpoint sink\_ff(0):0

* * *

Am I missing something?

Thanks,  
Sagar

---

<div class="post-metadata">

**Author:** ![sagar\_k](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/s/e79b87/32.png) [@sagar\_k](https://www.ruby-forum.com/u/sagar_k)\
**Post date:** [March 1, 2010, 9:57am UTC](https://www.ruby-forum.com/t/problem-installing-new-module/184345/2 "2010-03-01T09:57:30Z")

</div>

I am running MAC OS 10.5

---

<div class="post-metadata">

**Author:** ![sagar\_k](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/s/e79b87/32.png) [@sagar\_k](https://www.ruby-forum.com/u/sagar_k)\
**Post date:** [March 1, 2010, 4:42pm UTC](https://www.ruby-forum.com/t/problem-installing-new-module/184345/3 "2010-03-01T16:42:33Z")

</div>

On Mon, Mar 1, 2010 at 00:41, Sagar K. [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid)  
wrote:

> RuntimeError: unable to resolve input endpoint sink\_ff(0):0

The runtime doesn’t like a reference you are passing to a ‘connect’  
call. Check what you are putting in the connect call that you are  
wiring to ‘sink\_ff’ input 0. It should be a properly instantiated gr  
block or hierarchical block.

Johnathan

---

<div class="post-metadata">

**Author:** ![sagar\_k](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/s/e79b87/32.png) [@sagar\_k](https://www.ruby-forum.com/u/sagar_k)\
**Post date:** [March 5, 2010, 11:35pm UTC](https://www.ruby-forum.com/t/problem-installing-new-module/184345/4 "2010-03-05T23:35:53Z")

</div>

The exact same code works on two other machines, both running OSX 10.5.  
The  
sink is a locally created modules, so it is entirely possible that there  
is  
something missing in the instantiation that is being caught on only on  
my  
machine and not at others (for what reason I cannot fathom).

As an aside, another thing that might be helpful is that my python  
installation is partially broken: I have to move the installed .so .la  
files  
(as well as the python files) ouput from a make install from the  
/usr/local/lib onto the python script directory (as found during the  
configure step)  
/opt/local/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/site-packages.

On Mon, Mar 1, 2010 at 7:39 AM, Johnathan C. \<

---

<div class="post-metadata">

**Author:** ![sagar\_k](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/s/e79b87/32.png) [@sagar\_k](https://www.ruby-forum.com/u/sagar_k)\
**Post date:** [March 7, 2010, 2:31am UTC](https://www.ruby-forum.com/t/problem-installing-new-module/184345/5 "2010-03-07T02:31:16Z")

</div>

On Mar 5, 2010, at 5:24 PM, Sagar K. wrote:

> The exact same code works on two other machines, both running OSX  
> 10.5. The sink is a locally created modules, so it is entirely  
> possible that there is something missing in the instantiation that  
> is being caught on only on my machine and not at others (for what  
> reason I cannot fathom).

On my agenda to try out some of my blocks on 10.6 to see if they work  
(they do on 10.5). That said, I don’t have a sink or source block,  
just “through” blocks. Hence, it might be a source / sink issue, or  
it might be some odd change to 10.6. I’ll try to get to this next week.

> As an aside, another thing that might be helpful is that my python  
> installation is partially broken: I have to move the  
> installed .so .la files (as well as the python files) ouput from a  
> make install from the /usr/local/lib onto the python script  
> directory (as found during the configure step) /opt/local/Library/  
> Frameworks/Python.framework/Versions/2.6/lib/python2.6/site-packages.

Moving these files should not be necessary if PYTHONPATH is set  
correctly. That said, moving them around shouldn’t cause issues in  
any direct sense (they’re just shared libraries) … but if this  
Python is installed oddly, or partially broken, then you probably want  
to fix this issue before trying to correct other related issues. Just  
my advice on debugging. - MLD

---

<div class="post-metadata">

**Author:** ![sagar\_k](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/s/e79b87/32.png) [@sagar\_k](https://www.ruby-forum.com/u/sagar_k)\
**Post date:** [March 10, 2010, 3:38am UTC](https://www.ruby-forum.com/t/problem-installing-new-module/184345/6 "2010-03-10T03:38:16Z")

</div>

Hi all,

The same error I am also getting for the basic  
gr-howto-write-a-block-3.2

## ====================================================================== ERROR: test\_001\_square\_ff ( **main**.qa\_howto)

Traceback (most recent call last):  
File “./qa\_howto.py”, line 42, in test\_001\_square\_ff  
self.tb.run ()  
File  
“/opt/local/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/site-packages/gnuradio/gr/top\_block.py”,  
line 103, in run  
self.start()  
File  
“/opt/local/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/site-packages/gnuradio/gr/top\_block.py”,  
line 97, in start  
self.\_tb.start()  
File  
“/opt/local/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/site-packages/gnuradio/gr/gnuradio\_swig\_py\_runtime.py”,  
line 1455, in start  
return \_gnuradio\_swig\_py\_runtime.gr\_top\_block\_sptr\_start(self)  
RuntimeError: unable to resolve input endpoint square\_ff(0):0

## ====================================================================== ERROR: test\_002\_square2\_ff ( **main**.qa\_howto)

Traceback (most recent call last):  
File “./qa\_howto.py”, line 54, in test\_002\_square2\_ff  
self.tb.run ()  
File  
“/opt/local/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/site-packages/gnuradio/gr/top\_block.py”,  
line 103, in run  
self.start()  
File  
“/opt/local/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/site-packages/gnuradio/gr/top\_block.py”,  
line 97, in start  
self.\_tb.start()  
File  
“/opt/local/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/site-packages/gnuradio/gr/gnuradio\_swig\_py\_runtime.py”,  
line 1455, in start  
return \_gnuradio\_swig\_py\_runtime.gr\_top\_block\_sptr\_start(self)  
RuntimeError: unable to resolve input endpoint square2\_ff(1):0

* * *

Ran 2 tests in 0.059s

# FAILED (errors=2) FAIL: run\_tests

# 1 of 1 test failed

make[3]: \*\*\* [check-TESTS] Error 1  
make[2]: \*\*\* [check-am] Error 2  
make[1]: \*\*\* [check-recursive] Error 1  
make: \*\*\* [check-recursive] Error 1

So I think its more of configuration issue specific to my machine (MAC  
OS  
10.5). GNURadio works perfectly. Only when I try to add new block, I get  
same error… I am stuck… Any help will be appreciated.

Thanks,  
sagar

On Mon, Mar 1, 2010 at 7:39 AM, Johnathan C. \<

---

<div class="post-metadata">

**Author:** ![sagar\_k](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/s/e79b87/32.png) [@sagar\_k](https://www.ruby-forum.com/u/sagar_k)\
**Post date:** [March 12, 2010, 3:40am UTC](https://www.ruby-forum.com/t/problem-installing-new-module/184345/7 "2010-03-12T03:40:36Z")

</div>

Sagar resolved this issue off-list, without my help but with some good  
discussion. That said, I got curious about what might be going one,  
and after some hacking around on my Mac’s installs of GNU Radio I  
think the issue is that you have multiple versions installed and this  
particular “howto” is linking with the incorrect version. For  
example, suppose I have GNU Radio installed by MacPorts into /opt/  
local, and from GIT into /usr/local . I’ve setup my shell environment  
variables choose /opt/local first, so when I configure and build  
“howto” from GIT it all works just fine, but using the “framework”  
from 3.2.2 (since that’s what MacPorts currently installs). But “make  
check” fails in exactly the way you found.

If I then do:

sudo port deactivate gnuradio-core  
…/configure  
make  
make check

then it all works correctly (\*), using just the GIT-installed  
“framework” in /opt/local . I can’t explain why “make check” would  
fail under these circumstances, since the python QA code hasn’t  
changed (in functionality, just names for the imported module) – just  
the structure of where the files are located. Maybe someone else has  
a clue? - MLD

(\*) ‘run\_tests.in’ needs to be fixes to work for Mac OS X: since the  
files were separated out into multiple subdirectories (to allow for  
disabling SWIG/Python), the file now needs to know the library search  
path (via DYLD\_LIBRARY\_PATH, as also done in the main GNU Radio  
run\_tests.sh.in file). Probably needs a similar fix for Windows  
users, if anyone is testing that these days.
