Hi Richard - Glad you got the runtime part worked out for your OOT block. CMake has its good and bad moments ...
As for the why the QA code fails, since you're on Linux do you need to set the LD_LIBRARY_PATH somewhere? I know on OSX we have to "make install" before doing "make test" because the library paths are messed up otherwise (NB: I'm working on a fix; just takes time to get done & tested). Another question: When you execute "make test" does the QA code get executed (and then fails), or is it skipped over? - MLD On Thu, Sep 10, 2015, at 04:27 PM, Richard Bell wrote: > I figured out what part of the problem was. I added a few OOT blocks > (gr_modtool add block1, block2, etc..) to my module and wrote the code > for one of them, with the other two in default gr_modtool created > states. I wanted to compile and test each one at a time, so I thought > I could just comment out the other ones in CMakeLists.txt like this: > > list(APPEND my_block_sources #my_block_source1_impl.cc > my_block_source2_impl.cc <-- only this gets compiled ) > > You cannot do this. Once I uncommented the other blocks in the > CMakeLists.txt file, the block I was testing runs in grc. However, I > still cannot run my qa python code for the block without the import > error related to swig. I cd to the python folder of my OOT module and > do the following > >> python qa_my_block.py > > Traceback (most recent call last): File "qa_my_block.py", line 24, in > <module> import my_block_swig as my_block ImportError: No module > named my_block_*swig* > > I've tried removing the build directory and recreating it. No luck. > Any suggestions? _______________________________________________ Discuss-gnuradio mailing list [email protected] https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
