Trigger Warning at Virtuoso start up

You already asked about the first warnings a few days ago here, and I gave an answer.

The second three warnings are probably because you have some outdated bindkey definitions for the IC5141 waveform tool (which used the tool name "awv"), which is not present in IC615. So you probably have some functions:

hiSetBindKeys("awv" "<Key>A" "someFunction()")

or you're loading an old SKILL file containing such bindkey definitions.

Regards,

Andrew.

Hi Andrew and wgtkan,

I got 3 same warnings:

 *WARNING* No application registered for prefix awv.

Andrew had explained about these warnings. However, I cannot fix them. I think the problem is in .cdsinit file.

I don't know how to fix those problems. Can you explain more detail, please? Thank you so much.

Regards.

In my earlier post in this thread, I included a link to another post which suggested that these are probably coming from a call to deUnRegUserTriggers() in a context file loaded from the .cdsinit. It's unlikely that they're in the bindkeys - unless there's a call to this function in leBindKeys.il or schBindKeys.il - which seems unlikely.

The warnings are benign - although it is bad practice of the PDK to be unregistering the de triggers. This has been shown to break certain tools because they depend upon these triggers. In ancient history (IC446 and before) there used to be a limitation that you could only have one user trigger defined - but this was fixed over 10 years ago. So you really ought to contact IBM.

If you can't track down where they are coming from, and want a quiet life for users (no warnings), you could always add this at the beginning of the .cdsinit before any of the PDK stuff is loaded:

procedure(CCFdummy(args) t)
deRegUserTriggers("maskLayout" 'CCFdummy)
deRegUserTriggers("maskLayoutXL" 'CCFdummy)
deRegUserTriggers("schematicSymbol" 'CCFdummy)

Oh, hang on - you were asking about the "awv" messages (I got confused because you posted about the other warnings in another old thread). The awv warnings must be coming because of a call, somewhere, to an hiSetBindKey on "awv". You'll just have to hunt it down!

Try running "virtuoso -nocdsinit" and then gradually load the files loaded from the .cdsinit - or comment out pieces of the .cdsinit and see the effect.  Again, the warnings won't cause any problems in practice.

Regards,

Andrew.

Thank Andrew,

You're right. The .cdsinit call the bindkeys which are defined by pdk. And they conflict with Cadence bindkeys definition. I removed that code in .cdsinit file and the warning go aways. However, I think these warnings also cause problem. Before I fixed them, I could not see the properties of components (when I pressed "Q", I saw some paramters, but look like not enough). After I fixed, the properties of components look normal.

Again, thank for your instructions. 

Best regards,

Duy-Hien 

I don't think that's anything to do with the warnings. It might be something to do with the tool filter settings (Options->Tool Filter), but hard to tell without seeing what the problem is.

Regards,

Andrew.

Hi Andrew,

At first, I go to Option -> Bindkeys and I saw more than 200 bindkeys for schematic. I read the .cdsinit file again. And I found other codes load the bindkeys which are defined by pdk. I removed them. And the warnings go aways. Now, bindkeys for schematic are more than 100!

Regards

No, I meant that the warnings didn't have anything to do with the change in behaviour of the "Q" bindkey. That seemed a bit unlikely, unless they have a different definition of the Q bindkey which brings up a different property editor. Hard to tell without seeing the problem.

Regards,

Andrew.

I want to explain more detail about my problem. Before I fixed it, in a schematic, I chose a component A. When I clicked to device B, both devices are chosen (A and B, (similar to CTRL is hold)). So I though some hotkeys may conflict together. Therefore I removed some codes in .cdsinit. On the other hand, I saw the properties pop-up windows is different with photos in document. After I fixed, it almost looks ok.

Regards 

 IBM PDKs have some rather unusual bindkey choices (for some odd reason). First of all, each click is an incremental selection. It's possible that their Q bindkey calls a different function (turn on the Options->Log Filter and click all the checkboxes, and see what SKILL function gets called when hitting Q in the case with IBM's bindkeys and the standard Cadence bindkeys).

Regards,

Andrew.