DIVA error after upgrade to lastest USR

Discussion in 'Cadence' started by tritue, Feb 9, 2006.

  1. tritue

    tritue Guest

    Hi all,

    We are just upgrade IC5141 to latest revision then our Diva rul do not work
    any more on linux version. the solaris version work fine

    linux version is 5.10.41.500.3.30
    Solaris version is 5.10.41.500.3.31
    The previous version work fine in both linux/solaris.
    The error is :
    Parsing drcExtractRules of "divaEXT2.rul"...

    error: Layer LATEMIT1 is the common serial number source for connect
    layers:

    LATBASE LATCOLL LATEMIT1 LATEMIT LATGATE.

    info: If short location check is desired later on, check the
    saveInterconnect

    statement for correctness:

    Interconnect layer STRAP, PSD, NSD, PWCONN, PSUB, NWCONN, NWELL, PLCT,
    POLY2, POLY, PPIN, NPIN, LATGATE, AACT, BASE, EMITTERA, LATEMIT, LATEMIT1,
    LATCOLL, LATBASE, BIPDUM, EECT, EEC, EES, EED, BIPCT, BIPDEF, NWSHCT,
    NWSHOT, V1M, V2M, V3M, SUBCT, substrate not saved.

    Errors exist in the rules file "divaEXT2.rul".

    \o Verification program terminated.





    I never see this error before, any idea ?

    Thank
     
    tritue, Feb 9, 2006
    #1
  2. You have probably been seeing that in the form of a warning for quite
    some time now. It recently became an error. Using selection commands
    like geomInside to create the connected layers in geomConnect can lead
    to shapes on different layers having the same shape id (serial) number.
    Since geomConnect uses serial numbers as the key in a set association
    table, those shapes will be considered connected, even if they should
    not.

    You can fix your rule deck, or you can get the latest ISR and ignore the
    warning. If you don't fix the rule deck, you may have incorrect
    connectivity. If you file a PCR and the warning is being issued, R&D
    will correct the rule deck to remove the warning and reject your PCR
    when the "bug" goes away.
     
    Edward Kalenda, Feb 10, 2006
    #2
  3. tritue

    tritue Guest

    Thank you,

    I have to fix the Diva rule deck for each of process that we used. But It
    works.
    Basically we have to add a geomOr() for each layer that generated by
    geomEnclose/Inside to re-serialize the number then it work .
     
    tritue, Feb 10, 2006
    #3
Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.