Stats

  • Locked Locked
  • Replies 4
  • Subscribers 65
  • Views 15999
  • 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

Slow simulation caused by Assertions

nwang

I am using -profile to investigate why my simulation is so slow, and I found the warning:

ncsim: *W,FLSTRT

the explanation of it is: The assertion is spending a significant amount of time starting new
        attempts that immediately terminate.  In most assertions, such activity
        can be minimized by optimizations in ncsim; something about this
        assertion is preventing those optimizations from taking effect. 

I want to know what prevent the optimization, how can I fix it?

 My assertion is fairly simple as below:

 property check_conflict;
      @(posedge clk) ((a && b && !c) ##1 (d && c)) |-> ##[0:2]  S;
   endproperty
   conflict:assert property (disable iff (!rst_an) check_conflict)
         else `uvm_error("ERROR",$psprintf("conflict at %0t \n",$time))

 

Further, I don't quite understand what it means in the warning explaination "starting new attempts". In the case of my assertion, does it mean new attempts are started by satifying the first part of the expression (a && b && !c) is true, but terminated by (d&&c) is not true after ##1?  From what I see in the waveform, my undertanding is not right. Can someone explain?

 Could anyone help? Really appreciated.

  • You should check out the Coding Style Guidelines for Maximizing Assertion Performance guide. You can find it by launching 'cdnshelp' and going to Incisive Enterprise Simulator -> Assertion-Based Verification -> Assertion Writing Guide -> Maximizing Assertion Performance.

    Tim 

  •  Thank you very much, Tim.

    That's the document I need to study.

  •  Hi Tim,

    I have a further question.

    In the report ncprof.out, I have over 20% Stream Counts hits on SSS_MT_SVHASSIGN method, is it normal ?

    And over 60% on a function which is checking memory data. Although the test case is doing a lot of memory read/write continuously, does this high percetage look OK?

    Thank you.

  • These method entries are optimized assembly routines which are mostly event simulator internal operations. I would suggest trying the "-iprof" option that provide more detail than the standard "-profile" option. you'll need to use the "iprof" command to analyze the output from "-iprof". It is best to review the iprof help by searching for iprof in cdsnhelp.

     Tim  

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.