Needless to say, that last part should have been ". it does prove useful in some debugging situations ." And this is after I had my eyes fixed. Sheesh. Todd. Description: cid:EAFF2D52-4B63-4A05-9D24-B96BE375B7E0@eau.wi.charter.com Todd Westerhoff VP, Software Products Signal Integrity Software Inc. . <http://www.sisoft.com/> www.sisoft.com 6 Clock Tower Place . Suite 250 . Maynard, MA 01754 (978) 461-0449 x24 . <mailto:twesterh@xxxxxxxxxx> twesterh@xxxxxxxxxx "Three in the morning and I'm still awake, So I picked up a pen and a page . " -Sidewalk Prophets From: ibis-macro-bounce@xxxxxxxxxxxxx [mailto:ibis-macro-bounce@xxxxxxxxxxxxx] On Behalf Of Todd Westerhoff Sent: Thursday, March 08, 2012 9:22 AM To: 'IBIS-ATM' Subject: [ibis-macro] Re: Question about GetWave_Exists Arpad, The original intent, if I remember correctly, was to provide clear-text documentation on what the model did and didn't do. You're correct that the EDA tool could query the entry points itself and I don't think we considered that when the text was first written. But - the point stands, the .ami file provides clear text that the user can see. Approved BIRD 120.1 is more explicit. Beginning at line 448, we see things like this: | Step 5. If Tx GetWave_Exists is True the output of Step 4 is presented to | the Tx model's AMI_GetWave function and the Tx AMI_GetWave | function is executed. The output of the Tx AMI_GetWave function | is passed on to Step 6. This is clear - the value declared in the .ami file is to be used to control the analysis flow. No matter what anyone was (or wasn't) thinking when this text was written, this is what the text-to-be-included in IBIS 5.1 says. There is an obvious implication here; the reverse of the original question you proposed. If the .ami file claims Getwave exists and it actually doesn't, an EDA tool that blindly follows the directive will attempt to call an entry point that doesn't exist. The existing text could be augmented to include that consideration and recommend that the tool verify the entry point actually exists BEFORE calling it. I consider this to be good programming practice that everyone has probably incorporated already, or will after this discussion - without any further activity on the committee's part. Bottom line, I think [this is my opinion], the EDA tool should do what the .ami file says - otherwise, why bother having the parameter at all? While the ability to turn Getwave access on/off for a model that supports Getwave is an unusual application, I agree with David - it does provide useful in some debugging situations and is worth keeping around - besides, that's what the current text says. Thanks for the question, Todd. Description: cid:EAFF2D52-4B63-4A05-9D24-B96BE375B7E0@eau.wi.charter.com Todd Westerhoff VP, Software Products Signal Integrity Software Inc. . www.sisoft.com <http://www.sisoft.com/> 6 Clock Tower Place . Suite 250 . Maynard, MA 01754 (978) 461-0449 x24 . twesterh@xxxxxxxxxx "Three in the morning and I'm still awake, So I picked up a pen and a page . " -Sidewalk Prophets From: ibis-macro-bounce@xxxxxxxxxxxxx [mailto:ibis-macro-bounce@xxxxxxxxxxxxx] On Behalf Of Muranyi, Arpad Sent: Wednesday, March 07, 2012 6:42 PM To: 'IBIS-ATM' Subject: [ibis-macro] Re: Question about GetWave_Exists Ambrish, Even though I tend to agree with you, I wonder what was the purpose of putting this parameter into the AMI specification. The reason I wonder is because any programmer skillful in the art knows how to obtain the entry points to the functions in a DLL, i.e. they can figure it out programmatically whether the function exists or not, and if they are careful programmers, they will not call the function if the pointer to it is a null. Having that in mind, who is this information in the .ami file targeted to, and for what purpose? Thanks, Arpad =================================================== From: Ambrish Varma [mailto:ambrishv@xxxxxxxxxxx] Sent: Wednesday, March 07, 2012 4:22 PM To: Muranyi, Arpad; 'IBIS-ATM' Subject: RE: Question about GetWave_Exists Arpad, The parameter value was supposed to be an answer to the question Getwave Exists in the AMI model or not? So I do believe that the .ami file was written incorrectly. Thanks, -Ambrish. Ambrish Varma | Member of Consulting Staff P: 978.262.6431 <http://www.cadence.com> www.cadence.com _____ From: ibis-macro-bounce@xxxxxxxxxxxxx [mailto:ibis-macro-bounce@xxxxxxxxxxxxx] On Behalf Of Muranyi, Arpad Sent: Wednesday, March 07, 2012 5:05 PM To: 'IBIS-ATM' Subject: [ibis-macro] Question about GetWave_Exists Hello everyone, I would like to ask a question about the GetWave_Exists AMI parameter. This is what I see in the specification on pg. 144: | GetWave_Exists: | | GetWave_Exists is of usage Info and type Boolean. It tells | the EDA platform whether the "AMI_GetWave" function is | implemented in this model. Note that if Init_Returns_Impulse | is set to "False", then Getwave_Exists MUST be set to "True". I came across a model recently in which the AMI DLL does have a GetWave function, but the .ami file says GetWave_Exists = False. The vendor says that they want to use this AMI parameter as a switch to be able to tell the EDA tool whether to invoke the GetWave function or not. The way I read the above specification snippet is that this parameter supposed to tell the tool whether this function exists in the DLL or not. If this parameter was to be used to control whether the EDA tool should make use of it or not, we should have given this parameter a different name, something like this: Use_GetWave_Function_If_Exists I am curious to hear what our experts have to say about this. Is the model's .ami file written incorrectly? Should we advise the model maker to not use this AMI parameter this way? How should the EDA vendor handle this situation? Should the tool execute the GetWave function when it exists, regardless of what this parameter says, or should it only execute the GetWave function if this parameter is set to True, regardless of whether the GetWave function exists in the DLL or not J. Thanks, Arpad ==============================================================