Jason,

this is a tough one. I have no ideas, other than 'throw in a couple of
ILAs and see where it's stuck'.

-- M

On 07/23/2018 12:03 PM, Jason Matusiak via USRP-users wrote:
> I have a flowgraph with a custom RFNoC block in the middle.  That block
> has 2 inputs and 2 outputs.  Just to get started and work out the MIMO
> functionality, all I do is cross the the inputs to the outputs (so input
> 0 comes out output 1).
>  
> What I am seeing is that if I run the flowgraph after loading a fresh
> bitfile, I don't get anything out of my output who should be passing the
> data from an RFNoC radio->DDC  and I see "timeout on chan 0" scrolling
> past on the screen.  But, if I run the same flowgraph a second time, it
> suddenly works fine (as well as subsequent runs afterwards until I
> reload the bitfile).  
>  
> It doesn't matter if I swap the inputs to the special block, it always
> reads "timeout on chan 0" and the output stream on my block that should
> be displaying that data is dead the first run after a reload.
>  
> I am guessing that "timeout on chan 0" is releated to either the Radio
> or DDC block as it is consistently 0 even when I move it around on my 2
> inputs (if I pull out the DDC, I see the same issues, so it seems to be
> something between the radio and my block).
>  
> That said, I don't see this issue if I run the rfnoc_ddc.grc example
> from Ettus which also uses a Radio and DDC.
>  
> I could believe that it has something to do with a reset in my block
> (like the fact that there are two clear_tx_seqnum flags), I am just
> confused why the issue moves around depending on where the rfnoc/ddc is
> connected (yet the timeout does not). 
>  
> If I remove the radio/ddc and use an rfnoc split stream to take my GR
> signal source, I don't see these issues.  So I cannot figure out what is
> not playing nicely between my block and the RFNoC radio. BUT... If I go
> from the radio into a split stream to my two inputs, it works right the
> first time, and then doesn't work again after that (so almost the
> reverse issue as I've started with, but I REALLY feel like something
> with my block dealing with 2 inputs and 2 outputs is causing this issue.
>  
>  
> 
> 
> _______________________________________________
> USRP-users mailing list
> USRP-users@lists.ettus.com
> http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
> 


_______________________________________________
USRP-users mailing list
USRP-users@lists.ettus.com
http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com

Reply via email to