[USRP-users] another issue with RFNoC testbench

Jason Matusiak jason at gardettoengineering.com
Fri Mar 18 10:11:22 EDT 2016


I have another issue that seems to be user error (or more than likely 
misunderstanding).  I am using the moving_average testbench as my 
starting point.  Instead of having "TEST_LENGTH" for the for loop, I 
have the length of a file size that I am reading in and using as my data 
points.

My issue is that my testbench seems to "stop" after the 529th sample 
every time.  Looking at the different signals, it appears as if the 
downstream block from my noc_block under test (which is setup by the 
scripts) is de-asserting o_tready, so my block stops trying to process 
while it waits for that to go high.

I have tried for a bit to figure out why it is stopping there, but can't 
seem to find a setting that handles it.  If I add in delays, the test 
will run longer time-wise, but still stops at the same sample number.  
It is as if the downstream simulated block is full and needs to be 
flushed (but I could find no way to do that in the scripts).

Any ideas?




More information about the USRP-users mailing list