| Author |
Message |
ywguo
Joined: 09 Jan 2005 Posts: 18
|
10 Oct 2005 3:15 no_notifier |
|
|
|
|
Hello,
I need to disable timing check for several instances on running post-layout simulation. The simulator is NC-Verilog.
Does anybody know how to disable timing check (setup time and hold time) for only several instances in NC-Verilog?
Thanks
Yawei
|
|
| Back to top |
|
 |
aji_vlsi
Joined: 10 Sep 2004 Posts: 640 Helped: 72 Location: Bangalore, India
|
10 Oct 2005 4:00 no timing check |
|
|
|
|
Hi,
Perhaps you know this already: ncelab has a -notimingcheck - but that is a global switch - truns off for ALL instances.
If your concern is X-propagation due to timing violation (and not the violation iself), try using no_notifier flag, again global.
For instance specific stuff few ideas:
1. If you like this no_notifier kind of thing, it is fairly easy to create a TCL script to force notifier reg of each FF instance (that you are interested) to 0.
2. There is a PLI to do precisey what you wanted, see:
http://www.project-veripage.com/notc.php
3. IIRC, NC supports a system task named $disable_warning - to supress warnings from individual instances, search in their manual.
HTH
Ajeetha
--
www.noveldv.com
Interested in expert PSL/SVA training in Bangalore?
Visit www.noveldv.com/cvc.html
|
|
| Back to top |
|
 |
wls
Joined: 26 Jul 2001 Posts: 104 Location: Singapore
|
10 Oct 2005 5:55 ncverilog disable_warning |
|
|
|
|
Hello aji_vlsi. If we disable the timing checks , is it going to make the post simulation less accurate ? Some eda simulation tools , rounded off negative timing checks for certain fab. Is it allowable , lets say for .35um ? Does it affect the accuracy of the timing info ?
How do we confirm or say , okay it can be neglect ? If STA tools like PT check alright but post simulation have message saying that no negative timing checks done. And the value os rounded to zero ? So is it valid , the post simulation if functional checks alright .
Thx in advance.
|
|
| Back to top |
|
 |
aji_vlsi
Joined: 10 Sep 2004 Posts: 640 Helped: 72 Location: Bangalore, India
|
10 Oct 2005 11:16 ncverilog no_notifier |
|
|
|
|
Hi,
| wls wrote: |
Hello aji_vlsi. If we disable the timing checks , is it going to make the post simulation less accurate ?
|
Yes indeed, but that's where user intervens and says: i know this interface well, in real life these sginals don't toggle this way, hence let me ignore/turn them off.
| Quote: |
Some eda simulation tools , rounded off negative timing checks for certain fab. Is it allowable , lets say for .35um ? Does it affect the accuracy of the timing info ?
|
Many simulators do provide a way to enable -ve timing checks (-neg_thck ?)
I would prefer not to ignore -ve delays - unless you know what and why you are doing that.
| Quote: |
How do we confirm or say , okay it can be neglect ? If STA tools like PT check alright
|
Rely more on STA, do DTA (Dynamic, simulation) only for asynchronous stuff, sanity check and some functional tests.
| Quote: |
but post simulation have message saying that no negative timing checks done. And the value os rounded to zero ? So is it valid , the post simulation if functional checks alright .
Thx in advance. |
As I mentioned, there are tool options to make these warnings disappear.
HTH
Ajeetha
--
www.noveldv.com
Interested in expert PSL/SVA training in Bangalore?
Visit www.noveldv.com/cvc.html
|
|
| Back to top |
|
 |
Google AdSense

|
10 Oct 2005 11:16 Ads |
|
|
|
|
|
|
| Back to top |
|
 |
ywguo
Joined: 09 Jan 2005 Posts: 18
|
11 Oct 2005 8:02 negative timing check |
|
|
|
|
Hi, Ajeetha,
Thanks very much. I really need to prevent the X state propagation.
I try the option -nonotifier for ncelab. Now it doesn't shown any timing violation and X state at the waveform.
Yawei
|
|
| Back to top |
|
 |