[sv-bc] RE: [sv-ac] 22.10 bind review

Lisa, 

 

I do not see "variation" in 3.8.

 

This is 3.8:

 

3.8 Configurations

SystemVerilog provides the ability to specify design configurations,
which specify the binding information

of module instances to specific SystemVerilog source code.
Configurations utilize libraries. A library is a

collection of modules, interfaces, programs, primitives, packages, and
other configurations. Separate library

map files specify the source code location for the cells contained
within the libraries. The names of the

library map files are typically specified as invocation options to
simulators or other software tools reading in

SystemVerilog source code.

See Clause 32 for details of configurations.

 

Nevertheless, SV-BC people agreed with you about the term's meaning in
the context of bind.

 

However, they did not agree about it being a problem.

 

See http://www.eda.org/sv-bc/hm/5989.html 

and http://www.eda.org/sv-bc/hm/5993.html .

 

Regards,

Shalom

 

________________________________

From: Lisa Piper [mailto:piper@cadence.com] 
Sent: Monday, June 04, 2007 8:12 PM
To: Bresticker, Shalom
Cc: sv-ac@eda.org; sv-bc@eda.org
Subject: RE: [sv-ac] 22.10 bind review

 

Hi Shalom,
 
I disagree that there is any issue with the sentence:
 
"It shall be an error to use noninstance-based binding if the design
contains more than one variation of the target module, program, or
interface. This can occur in the presence of configuration library
mapping or
nonstandard functionality such as provided by the `uselib directive."
 

What the LRM talks about is multiple variations of a module( i.e.
potentially different implementations of module "child"  are compiled in
different libraries). And using verilog configurations, the elaborator
can bind instance c1 to one variantion of child and can bind instance c2
to different variation, as controlled by the "uselib".  In such cases,
LRM prescribes the instance based binding.  Refer the LRM section 3.8
that talks about configurations and the use of libraries.  It uses the
the word "variation" and then clarifies that it is referencing
configuration library mapping.  I think this is important so that
implementations of bind do not have to track library information as
well.  I think (not sure) that the bind syntax would need to change to
support it too.

 
Lisa

 

 

________________________________

From: owner-sv-ac@eda.org [mailto:owner-sv-ac@eda.org] On Behalf Of
Bresticker, Shalom
Sent: Friday, June 01, 2007 9:45 AM
To: Lisa Piper; john.havlicek@freescale.com
Cc: sv-ac@eda.org; sv-bc@eda.org
Subject: RE: [sv-ac] 22.10 bind review

 

Thanks.

 

That still has to be updated to Draft 3.

 

Also, that Mantis does not deal with the 'variation' issue.

 

Regards,

Shalom

 

________________________________

From: Lisa Piper [mailto:piper@cadence.com] 
Sent: Friday, June 01, 2007 4:35 PM
To: Bresticker, Shalom; john.havlicek@freescale.com
Cc: sv-ac@eda.org
Subject: RE: [sv-ac] 22.10 bind review

 

This was fixed in SV-AC Mantis 1722 which has not yet been incorporated
in draft 3. 

 

Lisa

 

________________________________

From: Bresticker, Shalom [mailto:shalom.bresticker@intel.com] 
Sent: Friday, June 01, 2007 6:40 AM
To: Lisa Piper; john.havlicek@freescale.com
Cc: sv-ac@eda.org
Subject: RE: [sv-ac] 22.10 bind review

 

Also see thread beginning http://www.eda-stds.org/sv-bc/hm/5983.html .

 

Shalom

 

________________________________

From: owner-sv-ac@server.eda.org [mailto:owner-sv-ac@server.eda.org] On
Behalf Of Lisa Piper
Sent: Friday, June 01, 2007 1:29 AM
To: john.havlicek@freescale.com
Cc: sv-ac@server.eda.org
Subject: [sv-ac] 22.10 bind review

 

Hi John,

I confirmed that the 22.10 Text on Bind was copied from Draft1 to Draft3
correctly.  However, I also happened to notice that the Clause 16
description in section 1.6 (page 26) needs to be updated to exclude
assertion binding.  It should probably be added to the Clause 22
description as well.  

Another think that stands out now is that the title "Binding properties
to scopes or instances" is not generic enough.  Bind was moved because
it does not  have to only be used for assertions. 

Also, I noticed that Syntax 22-3 - Module item syntax replicates the
bind directive from Annex A.  Mantis 1722 added a note to the annex A
bind syntax but did not note that this reference also needs to be
udpated.   The other changes of 1722 are not yet incorporated either.
Perhaps we should open another Mantis item to change the title and not
the note for Syntx 22-3.

Lisa


-- 
This message has been scanned for viruses and 
dangerous content by MailScanner <http://www.mailscanner.info/> , and is

believed to be clean. 
-- 
This message has been scanned for viruses and 
dangerous content by MailScanner <http://www.mailscanner.info/> , and is

believed to be clean. 

-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.