Re: gEDA: A Plea for Programming Help (Icarus Verilog)

[John Sheahan <jrsheahan-at-NOSPAMoptushome_dot_com_dot_au>]

I'll volunteer to help with this.
regressions in perl + verilog is something I've done before.

just thinking though - this is formal verification land,
rather than simulation.  could reducing the code work?
john




On Sun, Mar 23, 2003 at 08:34:09PM -0800, Stephen Williams wrote:
> 
> Hi,
> 
> Looking for a job that pays less then nothing?-)
> 
> I'm looking for Perl/Verilog/Xilinx programmers who are interested
> in the exciting (read: boring as hell) task of writing a regression
> test rig for Xilinx synthesis. What I'm looking for is a one-button
> regression test a la vvp_reg.pl in the ivtest test suite that tests
> Xilinx synthesis.
> 
> It turns out with the Xilinx tool chain that the results of "map"
> can be fed to ngd2ver to make a Verilog simulation of the EDIF
> design that was written out by the -tfpga target of Icarus Verilog.
> This can be used to make automatic regression tests. So I imagine
> that the test rig would take a synthesizeable Verilog source, run
> it through Icarus Verilog synthesis, ngdbuild, map, and ngd2ver
> to get Verilog out. It would then combine the output Verilog with
> another Verilog file, the test bench, to run the Verilog that
> went through the whole synthesis and mapping path.
> 
> The current test rigs in the ivtest project (in sourceforge.net)
> are written in perl. Even though I continue to avoid learning
> perl, it is preferable that perl remain the standard for test rigs
> in the test suite, unless someone demonstrates good results with
> expect or the like.
> 
> With the work I'm doing on Virtex-II code generation (available
> via CVS) I'm getting to the point where the -tfpga target is far
> too complex to test ad hoc. The regression test suite for simulation
> has been invaluable, partly because I did *not* write most of the
> programs that it contains, but also because the test rig runs auto-
> matically. I need something like that for Synthesis, and for Xilinx
> synthesis we can use ngd2ver.
> 
> It's OK (even preferable) if it runs under Cygwin.
> 
> If Xilinx ever releases their Linux native tools, I would choose
> to move over to them.
> -- 
> Steve Williams                "The woods are lovely, dark and deep.
> steve at icarus.com           But I have promises to keep,
> steve at picturel.com         and lines to code before I sleep,
> http://www.picturel.com       And lines to code before I sleep."
>