Hi all,

> 
> To implement libraries which allow some sort of system interaction beyond
> reading/writing text files one must go beyond the confines of VHDL, and
> VHPI is the way to go. That was my main point.

I guess noone would object.
I'm kind of an impatient software guy, so I've been looking for the most
efficient way to marry GUIs/Python/C/VHDL for a while. In the process I
found some inspiring stuff (thanks, whygee!) as well as undocumented
excellent features that the commercial tools don't offer (or maybe I
just didn't spend enough money..).
At the moment, I have a pretty complex SoC design running as a "virtual
board" within GHDL and the 'ghdlex' extensions (co-simulation of HW and
SW via VHPI), including JTAG so that I can remote-debug the entire SoC
with an 'off the shelf' Gnu debugger. I've got quite a gain from this
method personally, and it seems it's matching the trend of the 'big
guys' too, so I'm happy to be able to demonstrate some stuff at the
'embedded world 2012'. Sneak peak: http://section5.ch/doc/jtag/movies/

Anyhow, the problem might be, that some people not very inclined to
'reverse engineering' are put off by the lack of documentation or things
not following the 'standard'. I've started documenting some stuff for
ghdlex in doxygen, but that's not even 20% of the pole.

I might be ignorant, but I see no point in implementing "new standards"
that don't immediately boost us factor 10x in solving typical
development problems. As far as I can see, GHDL already is one of the
best standard compliant tools around.
[ Pretty please, don't touch the running system ].
But I'd sure like to hear more opinions about limitations with respect
to the current VHPI interface. The ones I've seen so far could be worked
around easily. So the priority '9' task in my opinion is, writing more
docs/examples/appnotes about what the current stuff can do...

Cheers,

- Martin

_______________________________________________
Ghdl-discuss mailing list
[email protected]
https://mail.gna.org/listinfo/ghdl-discuss

Reply via email to