|
From: <by...@nc...> - 2018-10-29 21:19:48
|
You don't need the documentation for the .gtkw format: just save it
from gtkwave. It's probably easiest to add the FST signals manually by
having two gtkwave sessions open: one is LXT2 and one is FST. Once you
get FST looking right, save it and only use it for FST.
Do *not* convert files as this will make the problem worse. Generate
FST directly from Icarus and once it works, forget that LXT2 ever
existed. :-)
For transaction filters, see transaction.c in the examples/ directory.
It's a little bit tricky getting transaction filters to work.
-Tony
-----------------------------------------From: "Ron Burkey"
To: gtk...@li...
Cc:
Sent: Monday October 29 2018 4:22:02PM
Subject: Re: [Gtkwave-users] Problem with viewing FST generated by
Icarus Verilog in GTKWAVE
Okay, I'm starting to see it, but I'm not there yet. I'm sure the
documentation for the .gtkw format and for how to write these
transactions filters is right out in the open somewhere, but I can't
find it. (And I didn't write this code I'm working with, so I have no
past experience to call upon.) Could you possibly point me in the
right direction to find the documentation? Thanks!
On 10/29/2018 01:46 PM, Ron Burkey wrote:
Redo it how?
On 10/29/2018 01:39 PM, by...@nc... [1] wrote:
Signal names are different due to LXT2 limitations. Please redo
your .gtkw file and it should work.
LXT2 has not been in active development for at least nine years. I
am considering dropping LXT/LXT2/VZT support from gtkwave in order to
force their removal from other tools.
-Tony
----------------------------------------- From: "Ron Burkey"
To: Gtk...@li... [2]
Cc:
Sent: Monday October 29 2018 12:07:38PM
Subject: [Gtkwave-users] Problem with viewing FST generated by Icarus
Verilog in GTKWAVE
I've partially switched over from using LXT2 files to using FST files
in running simulations in Icarus, but can't do so entirely because of
a little problem I'm having viewing the FST dumpfiles in GTKWAVE.
The problem can be reproduced with the files at this Google Drive
link:
[3]https://drive.google.com/open?id=1sif-JU89qnlCuG_Rn3dWDskXoo9Rs6_j
[4]. To see the problem, you have to view agc.fst in GTKWAVE, using
test.gtkw, making sure that instruction_trans_decoder.py is in the
same directory. What you should see is what's in screenshot.png (which
I got by viewing agc.lxt2), but what I actually get is an error
message from GTKWAVE reading:
ERROR: Group End encountered with no matching start. Ignoring.
What's going on in this example is that a Python script
(instruction_trans_decoder.py) has been provided that decodes the
signals in the dumpfile, and displays them in GTKWAVE as CPU opcodes.
This works when I view the LXT2 file, but fails when I view the FST
file. The signals in the dumpfile are fine, as far as I know, but the
decoded instructions simply won't show up when viewing the FST file.
This also fails if I use fst2vcd and vcd2lxt2 to create an LXT2 file
from the FST file, and then try to view the newly-created LXT2 file.
Is this a problem with the format of my gtkw file? With GTKWAVE? With
Icarus?
-- Ron
P.S., I'm using the latest Icarus release candidate and GTKWAVE
3.3.95.
_______________________________________________ Gtkwave-users mailing
list Gtk...@li... [5]
[6]https://lists.sourceforge.net/lists/listinfo/gtkwave-users [7]
Links:
------
[1] mailto:by...@nc...
[2] mailto:Gtk...@li...
[3] https://drive.google.com/open?id=1sif-JU89qnlCuG_Rn3dWDskXoo9Rs6_j
[4] https://drive.google.com/open?id=1sif-JU89qnlCuG_Rn3dWDskXoo9Rs6_j
[5] mailto:Gtk...@li...
[6] https://lists.sourceforge.net/lists/listinfo/gtkwave-users
[7] https://lists.sourceforge.net/lists/listinfo/gtkwave-users
|
