Forum Discussion

Altera_Forum's avatar
Altera_Forum
Icon for Honored Contributor rankHonored Contributor
13 years ago

Quartus StateBased trigger condition bug

I have a simple 3 state based trigger state machine with 3 trigger conditions. All state based configurable run time options are checked so I can change values, etc. With the setup saved when I created the STP file, I can run and trigger just fine. I can change things in the state machine OK and save it just fine, but if I change anything in any of the 3 trigger conditions and save it, then Quartus complains and wants me to recompile! I'm an engineer at Western Digital and would like to make contact with someone so I can send some screen shots and files. This is not the only bug I have discovered in the tool. I am using Quartus II 32 bit SP2 web-based version 11.1

3 Replies

  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor

    I'm assuming you are using the state based triggering in Signal Tap.

    I have not used this mode much, so I can not help you with your issue, but my thought is it requires the design to be recompiled in this mode, because it has to add the logic to handle the changes associated with the triggering.

    In the simple triggering method you don't need to recompile because the associated edge detectors are already part of the logic that signaltap applies.

    If you still feel this is a bug, I would contact FAE first. They would be able work with you closer, and push the service request though channels for you.

    Pete
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor

    Hi Pete,

    I still believe its a bug. Here's why. Changing triggers does not require any recompile or place and route. Neither does changing the state machine comparison values IF you turn on the option(s) for allowing it which I did. The previous state machine I had was 2 instead of 3 states and I had 2 instead of 3 trigger conditions. It worked perfectly fine in that prior version whether I changed trigger conditions, state machine values, etc.. So, just increasing the number of trigger conditions and states should not matter. Now, here's what I found out this afternoon which confirms it a bug. With my new state machine/trigger conditions, I did a comparison of the .stp file, before I changed a single trigger condition, with that after changing the trigger condition and saving a new .stp file. I then did a comparison of the two files. There were several things that were different, but the one thing that stood out was the later file was missing a CRC string contained in the original file. So I copied the string and put it in the right location and all works well - until I go change the trigger condition and have to do this again! Definitely a bug. :) Can you get an FAE to contact me and I can send him some files.

    Thanks

    Kirk Weedman

    Verilog RTL Design Engineer

    Western Digital
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor

    Hi Kirk:

    I would go ahead and open up a Service Request on Altera's website.

    FAE's are assigned through your purchasing channel. If you buy devices directly from Altera, Altera should have an FAE assigned for your company.

    I'm just a design engineer like yourself. My FAE is in through Arrow, and is based in the Seattle area, so he probably wouldn't know who covers Western Digital.

    My suggestion, is talk to your purchasing person, and see who's the contact is. This person should know your FAE, and can give you the contact info.

    Get to know your FAE, because they'll help you push issues through the factory support..

    Pete