Quoting Thomas Sailer <[email protected]>:

> Hi Tristan,
>
> thanks a lot for taking a shot at modifying the grt data structures to
> make them more efficient with many non-sensitized processes!
>
> I've tried svn130 on one of my larger testbenches. Compared to svn126,
> it takes a lot longer (more runtime), and then it aborts with the
> following messages:
> $ ./tbench --trace-processes
> Now is 1178670ns +0
> ./tbench:internal error: process in timeout
> ./tbench:error: simulation failed
>
> Do you have any idea what could have gone wrong?

No.  I tested this change with my testsuite and there was no regression.
Maybe the change is bogus or the test is wrong.

Can you send to me your design (or a reproducer) ?

Thanks,
Tristan.

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

Reply via email to