Stats

  • Locked Locked
  • Replies 1
  • Subscribers 92
  • Views 3442
  • Members are here 0
This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Query nets/cells causing DRVs using Tcl or dbGet commands to automate ECO flows

Luca Pacher

Dear all,

I'm currently facing some issues with signoff timing closure for a highly congested block design.

In short,  the flow runs fine up to signoffTimeDesign in Innovus, however I'm not able to find an "automated" way to fix a few residual DRVs (max_tran and max_cap) in my design apart from invoking ECO flow commands such as ecoChangeCell or ecoAddRepeater "by hand" in interactive Innovus sessions according to reports generated using timeDesign, signoffTimeDesign, reportTranViolations or reportCapViolations commands.

I would like to know if there are flows that allows to "automate" a little bit these small fixes in such cases, mainly due to the fact that net names or cell instance names causing these few residual DRVs usually change from run to run.

What I'm searching for is basically a Tcl flow that allows to:

1. identify nets or cells violating DRVs such as max_cap and max_tran using Tcl or dbGet commands

2. obtain a database pointer to these net/cell objects

3. automatically run ECO commands such as ecoChangeCell or ecoAddRepeater  targeting nets or cell names extracted from the database pointer

Just to be clear, I'm not asking how to generate DRVs reports or how to fix DRVs using the default setOptMode/optDesign -drv flow, I would like to automate fixes for residual DRVs without the need of fixing them one after the other "by hand" using ECO commands interactively.

Any suggestion would be more than appreciated !

Cheers

Luca

Parents
  • Dimo M

    Hi Luca,

    this article contains an exemplary script for fixing transition violations: Fixing transition violation by adding buffer on route

    It expects the net names as an input, though.  To identify the nets that need fixing, you can do this:

    # First, redirect the output of the report to a variable
    redirect -variable cap_vio_report "reportCapViolation"
    # Extract the violating nets from the report
    set max_cap_violated_nets [dict values [regexp -all -inline {\n([^\s#]+)\n} $cap_vio_report]]

    redirect -variable tran_vio_report "reportTranViolation"
    set max_tran_violated_nets [dict values [regexp -all -inline {\n([^\s#]+)\n} $tran_vio_report]]

    You might need to adjust the regular expression. Now the group ([^\s#]+) is matching all non-space characters that are not hash (#), I am not sure if this covers all the corner cases. 
    The report's format is such that for every violating net there is an entry with only the net name on a new line. The regexp expression is using this to extract the net names from the report.
    Due to the way the regexp returns the matches, I also added "dict values" to extract every second entry from the returned list.

    BR
    Dimo

Reply
  • Dimo M

    Hi Luca,

    this article contains an exemplary script for fixing transition violations: Fixing transition violation by adding buffer on route

    It expects the net names as an input, though.  To identify the nets that need fixing, you can do this:

    # First, redirect the output of the report to a variable
    redirect -variable cap_vio_report "reportCapViolation"
    # Extract the violating nets from the report
    set max_cap_violated_nets [dict values [regexp -all -inline {\n([^\s#]+)\n} $cap_vio_report]]

    redirect -variable tran_vio_report "reportTranViolation"
    set max_tran_violated_nets [dict values [regexp -all -inline {\n([^\s#]+)\n} $tran_vio_report]]

    You might need to adjust the regular expression. Now the group ([^\s#]+) is matching all non-space characters that are not hash (#), I am not sure if this covers all the corner cases. 
    The report's format is such that for every violating net there is an entry with only the net name on a new line. The regexp expression is using this to extract the net names from the report.
    Due to the way the regexp returns the matches, I also added "dict values" to extract every second entry from the returned list.

    BR
    Dimo

Children
No Data

Community Guidelines

The Cadence Design Communities support Cadence users and technologists interacting to exchange ideas, news, technical information, and best practices to solve problems and get the most from Cadence technology. The community is open to everyone, and to provide the most value, we require participants to follow our Community Guidelines that facilitate a quality exchange of ideas and information. By accessing, contributing, using or downloading any materials from the site, you agree to be bound by the full Community Guidelines.