gEDA-user: [icarus] completely open source fpga toolchain

[Stephen Williams]

Adam Megacz wrote:
> Stephen Williams <steve-TQXSILARuq3QT0dZR+AlfA at public.gmane.org> writes:
>> The v0.8 releases of Icarus Verilog have decent synthesis. The
>> synthesis is not at all Xilinx specific, but the code generators
>> are. But they needn't be. The FPGA target generates EDIF, so if
>> your intermediate form takes EDIF, the way to move forward is to
>> work on the fpga code generator to generate code for your device.
> 
> Hey, thanks for the advice.  I've got 0.8.4 generating EDIF that
> parses and flattens it into a netlist my PAR tools can handle.
> 
> One difficulty, though: the primitive cells that iverilog emits are
> pretty complex.  Is there any way to ask it to break down multipliers
> and adders into stuff no larger than a LUT4?

The reason it generates at that level is that in some cases it
is the appropriate level. For example, there *are* multipliers
in many modern FPGA primitive sets. It is a whole lot easer for
a code generator to break down a multiplier into gates, then to
merge up gates into a multiplier.

> Also, what's the difference between the tgt-edif and tgt-fpga
> directories?

Licensing, mostly. The tgt-edif target source code uses a BSD
style license, whereas the tgt-fpga target (which is currently
a superset) uses the GPL license.


-- 
Steve Williams                "The woods are lovely, dark and deep.
steve at icarus.com           But I have promises to keep,
http://www.icarus.com         and lines to code before I sleep,
http://www.picturel.com       And lines to code before I sleep."