Data Not Showing Up In SimVision

[Jake]

Hi,

I'm new to this whole NCVerilog-SimVision thing, and there's a
phenomena that I keep observing that is really confusing me (and
pissing me off). I choose nodes to display in the SimVision window,
and then start a simulation. The first time I run through the sim,
everything always works fine. However, if I want to observe different
nodes, after I rewind and restart the simulation, sometimes certain
nodes don't show up. It seems to be completely random as to when it
works and when it breaks. Hasn't anyone else observed this, and if
so, how did you fix it?

Thanks,
Jake

 

[Mike Treseler]

Jake wrote:

I'm new to this whole NCVerilog-SimVision thing, and there's a
phenomena that I keep observing that is really confusing me (and
pissing me off). I choose nodes to display in the SimVision window,
and then start a simulation. The first time I run through the sim,
everything always works fine. However, if I want to observe different
nodes, after I rewind and restart the simulation, sometimes certain
nodes don't show up. It seems to be completely random as to when it
works and when it breaks. Hasn't anyone else observed this, and if
so, how did you fix it?

In modelsim, I have to declare in advance
which nodes I want to log or view
when the sim is done.

-- Mike Treseler

 

[Ajeetha (www.noveldv.com)]

On Mar 17, 3:39 am, "Jake" <wegman.j...@gmail.com> wrote:

Hi,

I'm new to this whole NCVerilog-SimVision thing, and there's a
phenomena that I keep observing that is really confusing me (and
pissing me off). I choose nodes to display in the SimVision window,
and then start a simulation. The first time I run through the sim,
everything always works fine. However, if I want to observe different
nodes, after I rewind and restart the simulation, sometimes certain
nodes don't show up. It seems to be completely random as to when it
works and when it breaks. Hasn't anyone else observed this, and if
so, how did you fix it?

Thanks,
Jake

I believe what you are observing is somewhat common across tools.
Unless you explicitly state what you want to observe in an interactive
run "Prior to" run, tools tend to "guess" what you might need (based
on what's already added to Wave, expanded in hierarchy etc.) and
that's the root cause of this confusion. One guideline I keep
recommending is to dump as much as you need before run, and for
reasonably sized blocks that will be "all". In NCSIM TCL one can do:

database -open -shm waves -default
probe -create -shm top -depth all


I'm quite sure that there is a small GUI around doing this though I
don't remember it now.

This should make it fully predictable.

Good Luck
Ajeetha, CVC
www.noveldv.com