https://open-source-silicon.dev logo
Join Slack
Powered by
<@U016Q6CGUCB> I saw the example of configuring LA...
# shuttle
h
@User I saw the example of configuring LA they are doing it after the 1 is written into the 
reg_mprj_xfer
m
necessary for the gpio setup, not needed for LA
h
@Matt Venn I want to send a signal through the LA when the GPIO pads are setup
m
I'm doing the same
my reset line is high to reset, so I send high then low signal
the trace you showed in the other thread is the logic analyser input signals, not the output
hmm, think I'm wrong about la_data_in and out I just checked and here's my trace
image.png
you have to set reg_la[1234]_ena as low to work as an output. if high then they are inputs
but it seems you're doing that
h
@Matt Venn yes I am doing that. Still my 
la_oen
does not change from 0xFFFFFFFF
m
how long are you waiting?
for me it takes 170us or so shown by the simulator
h
umm, @Matt Venn i used the testbench from the io_ports folder. Can you share your tb where you are waiting so that I can check mine as well?
m
check the link I already sent
h
I checked, my gtkwave is maxing out to 134700 ns
m
so probably not long enough for the c code to run
h
If i try replicating your testbench, i get errors like these: Net USER_VDD3V3 is not defined in this context.
I saw your pull request and changed the testbench accordingly and it worked. But I guess your tb uses the declarations that causes these errors.
m
also this branch is way out of date with caravel/develop
h
@Matt Venn you are running your programs on the 
develop
branch of the caravel?
m
no
on my branch I linked
h
@Matt Venn I can't seem to go to 170us on gtkwave. Even changed the repeat loop of 1000 clock cycles from 15 to 35 in the testbench.
m
post a link to your testbench
and post a screenshot of what happens on the command line after the testbench finishes
h
Sorry, the git is not configured properly. I have attached the testbench here.
ibtida_test_tb.v
Here is the output on terminal
m
so your test is only waiting for wait(mprj_io_0 == 8'h05); and then ends
you need to change that line or remove the finish if you want it to keep running
h
Yup now go it! Who would have thought it was such a silly mistake. Thanks @Matt Venn!
Increasing the clock cycles showed that I was receiving the actual signal.
m
np
Open in Slack
PreviousNext
Powered by