Error when comparing 2 rtl files on conformal lec tool

comparing 2 rtl files on conformal lec

Hi,

I am facing the issue where we are getting the error while comparing 2 rtl files in conformal lec tool
Both the rtl files are present in the same location and are same rtl set.
Still getting the issue where it is saying that unmapped points (extra) - 657 along with 654 unreachable points !!

Can anyone let me know the cuase for this.

REgards
Limitless

Re: comparing 2 rtl files on conformal lec

Why are you comparing 2 RTL files using conformal?

Re: comparing 2 rtl files on conformal lec

sharath666 said:

Why are you comparing 2 RTL files using conformal?

Click to expand...

I am doing it for testing purpose.
Is there a known issue that conformal will throw up an error if we are comparing 2 rtl files.

Re: comparing 2 rtl files on conformal lec

We usually use Conformal to test whether a RTL is the same as the corresponding netlist. I don't see any sense in comparing 2 RTL files. What do you mean by testing purpose?

Re: comparing 2 rtl files on conformal lec

sharath666 said:

We usually use Conformal to test whether a RTL is the same as the corresponding netlist. I don't see any sense in comparing 2 RTL files. What do you mean by testing purpose?

Click to expand...

No, We can use Conformal for Comparing 2 RTL's like for example you can compare between a Normal RTL and a scan-inserted RTL file. Here the case is with normal RTL file and another RTL file on which verification has been run and it is already freezed.

So now we are seeing some Extras points in the log file ? I need to know what is the cause for this.
Below is the snippet which i am attaching here for more details:

1974 unmapped points reported
================================================================================
Golden:
--------------------------------------------------------------------------------
Unmapped points DFF DLAT Total
--------------------------------------------------------------------------------
Unreachable 584 70 654
================================================================================
Revised:
--------------------------------------------------------------------------------
Unmapped points DFF DLAT E Total
--------------------------------------------------------------------------------
Unreachable 584 70 9 663
Not-mapped 0 0 657 657

Regards
Limitless_21

Re: comparing 2 rtl files on conformal lec

limitless_21 said:

No, We can use Conformal for Comparing 2 RTL's like for example you can compare between a Normal RTL and a scan-inserted RTL file.

Click to expand...

I would tell you reconsider the reply from sarath666!
Are you sure about what a scan-inserted RTL is and how is it obtained?

Basically what a LEC tool (I refer here to Cadence Conformal LEC tool since I had used it) does is to compare between two versions (golden and revised). One of them is the RTL and the other is the netlist. If I remember correctly what I had read in the Cadence LEC docu 3 yrs ago, the Cadence LEC tool internally converts the RTL to netlist and then does the comparison with the fed-in netlist. This is a standard ASIC design flow.

So your statement of comparing two RTLs using LEC tool is a bit strange to me.

Re: comparing 2 rtl files on conformal lec

Moreover if you already have a verified and frozen RTL, why don't you use the same instead of checking whether some other RTL is equivalent to this or not.
I don't know what a scan inserted RTL is. I have heard of only a scan inserted netlist.

Re: comparing 2 rtl files on conformal lec

Some time but rarely, I used LEC to compare two RTL files, as someone asked for what?
First, no one could claim their test suite cover 100% of their codes. The test suit cover the verification plan which indicate which functionalities must be checked. So re-run the test suit could not guarantee the RTL modification and could need more time than just execute LEC on a single file.
I usually used LEC to guarantee the functionality did not change between the two files when I do backend for external customer, and to show them, their RTL code could be changed to improve the synthesis qor, generally I did not have their test suite.