On Thu, 04 Nov 2010 18:00:02 +0100, Thomas Sailer wrote:
> On Thu, 2010-11-04 at 17:32 +0100, whygee wrote:
>> hah, that's another issue. But assuming that it was not,
>> what good reason do I have to not write a VHPIDIRECT-based
>> package with these essential operations ?
> None IMO...
>> Has anybody done it before ?
> Not that I know of.
still no answer on comp.lang.vhdl, no trace on google...
the deal is sealed.
That will be a nice addition to http://ygdes.com/GHDL/
> I thought about doing it, but then again I do not want to become ghdl
> only at the moment.
OK, that's a good reason.
then again, if nobody moves first, nobody will follow :-/
> So for me to use something like this, I would need a
> compatibility package that's implemented in terms of standard VHDL and
> that doesn't suck too hard.
haha you have a good point :-)
> That said, I haven't gotten around writing such a package...
well there are 3 "methods" for such a feature :
- direct compiler support (see below)
- VHDL package (GHDL only, with and I won't care about 64-bit
compatibility)
- portable, slow, stupid version that works everywhere :
good luck swallowing it :-/ (the speed argument is completely
lost because it is as slow or slower than std_ulogic_vector)
>> Could there be (later) a patch to GHDL that provides AND/OR/XOR
>> to integers so GCC does not have to make a function call
>> just of one boolean instruction ?
> Sure, if you find someone competent enough in gcc frontends written in
> Ada... 8-)
haha ! If at least I was fluent with the source code of GHDL,
it would be possible... I don't want to step on Tristan's feet
so now I prefer to make a simple GHDL-only package that provides
the first functionalities and features. We'll see later how that
evolves, but we can't if there is no prototype of any kind.
and there we go...
> Tom
yg
_______________________________________________
Ghdl-discuss mailing list
[email protected]
https://mail.gna.org/listinfo/ghdl-discuss
