On Wed, Jan 20, 2010 at 09:11:00PM +0100, Kevin Steffensen wrote:

Hi,

> Thanks Tristan for working on implementing PSL support in GHDL. I've spent 
> the last couple of days messing about with it and I have some feedback. 

Thank you for using PSL.

> ** The first line of them manual "PSL implementation" page is unclear to me. 
> Should the PSL stuff go in separate files or in the VHDL files? I've written 
> the PSL stuff directly in the vhdl files and that works (as I would expect). 
> If the separate files referred to on the manual page are GHSL implementation 
> files I don't think it should be mentioned in the manual.

Oops, I forgot the 'not' before 'in separate files'.  Indeed, without 'not' it
is very confusing.

> **In the "PSL implementation" page it states "A PSL statement is considered 
> as a concurrent statement, and therefore is
> allowed only where processes are." At first I understood this to mean that 
> PSL statements were only allowed within processes. Maybe this should be 
> reworded. I don't have any good suggestions, though.

I have replaced this sentence with:
A PSL statement is considered as a process.  So it is not allowed within
a process.

I think this is clearer.

> **Is the PSL "cover" directive supported and if so how do I get the 
> information from it after a testbench run?

Not yet.  That's on my TODO list.  You can replace it with an assertion, which
fails when covered.

> **When analyzing a PSL assertion involving a specific bit of a 
> std_logic_vector or other array type (is that called a slice?) GHDL will 
> crash with an unhandled exception. I've filed this bug on the GNA site.

Thank you for filling the bug.  I will work on that too.

> **GtkWave 3.3.1 will crash with a seg fault when loading a ghw file from GHDL 
> 0.29. GtkWave 3.1.10 does not have this problem. I haven't checked the 
> versions in between. I don't know whether this is a GtkWave or GHDL problem 
> so I filed the bug with both bug trackers.

I think this will be fixed in gtkwave 3.3.2.

Thank you for your comments,
Tristan.

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

Reply via email to