Simulation Problem with Archaric VHDL model

[Kuan Zhou]

Hi,

I am using Cadence 5.0.33.011305 on Fedora 2 to simulate my
schematics. But spectre gives me the following warning message:

Warning from spectre during AHDL read-in.
"/cad/cds/IBM_PDK/cmrf7sf/relML/Spectre/models/mvcap.va", line 11:
Warning: (Archaic Syntax) Does not comply with the Accellera Verilog-AMS
2.0 Standard and beyond. To comply with the current standard, use a
larger number in place of `inf'. `inf' is now used only to specify
ranges.
"/cad/cds/IBM_PDK/cmrf7sf/relML/Spectre/models/mvcap.va", line 19:
WarningArchaic Syntax) Does not comply with the Accellera Verilog-AMS
2.0 Standard and beyond. To comply with the current standard, assign a
value to `T'. A variable that is never assigned a value is
considered a digital variable in Verilog-AMS and is therefore not pure
Verilog-A.


How can I remove such warning messages? I used the same Cadence on
Redhat 8 in another machine and got no warning message. But redhat 8
doesn't support my SATA harddisk and audio chip.

Kuan

 

[Andrew Beckett]

On Sat, 19 Mar 2005 12:04:29 -0500, Kuan Zhou <koy2@cisunix.unh.edu> wrote:

Hi,

I am using Cadence 5.0.33.011305 on Fedora 2 to simulate my
schematics. But spectre gives me the following warning message:

Warning from spectre during AHDL read-in.
"/cad/cds/IBM_PDK/cmrf7sf/relML/Spectre/models/mvcap.va", line 11:
Warning: (Archaic Syntax) Does not comply with the Accellera Verilog-AMS
2.0 Standard and beyond. To comply with the current standard, use a
larger number in place of `inf'. `inf' is now used only to specify
ranges.
"/cad/cds/IBM_PDK/cmrf7sf/relML/Spectre/models/mvcap.va", line 19:
WarningArchaic Syntax) Does not comply with the Accellera Verilog-AMS
2.0 Standard and beyond. To comply with the current standard, assign a
value to `T'. A variable that is never assigned a value is
considered a digital variable in Verilog-AMS and is therefore not pure
Verilog-A.


How can I remove such warning messages? I used the same Cadence on
Redhat 8 in another machine and got no warning message. But redhat 8
doesn't support my SATA harddisk and audio chip.

Kuan

Actually, it's not VHDL - but Verilog-A. It's nothing to do with Fedora 2, but
the fact that the Verilog-A parser has become overly strict. Many of these
cases are not strictly correct.

Solution 11170754 talks about this. Currently a number of PCRs have been filed
for these issues, but the fixes are not available yet.

I'm surprised the problem happens on one Linux machine but not another. Are
you sure you're using the same version?

Andrew.

 

[Kuan Zhou]

Hi Andrew,

Sorry for the confusion. I checked the version again. The Cadence I
installed on Fedora is released in 2005. The Cadence on Redhat 8 is
released in 2004.I will install the older Cadence on Fedora to
avoid this problem.

Kuan

On Sun, 20 Mar 2005, Andrew Beckett wrote:

On Sat, 19 Mar 2005 12:04:29 -0500, Kuan Zhou <koy2@cisunix.unh.edu> wrote:

Hi,

I am using Cadence 5.0.33.011305 on Fedora 2 to simulate my
schematics. But spectre gives me the following warning message:

Warning from spectre during AHDL read-in.
"/cad/cds/IBM_PDK/cmrf7sf/relML/Spectre/models/mvcap.va", line 11:
Warning: (Archaic Syntax) Does not comply with the Accellera Verilog-AMS
2.0 Standard and beyond. To comply with the current standard, use a
larger number in place of `inf'. `inf' is now used only to specify
ranges.
"/cad/cds/IBM_PDK/cmrf7sf/relML/Spectre/models/mvcap.va", line 19:
WarningArchaic Syntax) Does not comply with the Accellera Verilog-AMS
2.0 Standard and beyond. To comply with the current standard, assign a
value to `T'. A variable that is never assigned a value is
considered a digital variable in Verilog-AMS and is therefore not pure
Verilog-A.


How can I remove such warning messages? I used the same Cadence on
Redhat 8 in another machine and got no warning message. But redhat 8
doesn't support my SATA harddisk and audio chip.

Kuan

Actually, it's not VHDL - but Verilog-A. It's nothing to do with Fedora 2, but
the fact that the Verilog-A parser has become overly strict. Many of these
cases are not strictly correct.

Solution 11170754 talks about this. Currently a number of PCRs have been filed
for these issues, but the fixes are not available yet.

I'm surprised the problem happens on one Linux machine but not another. Are
you sure you're using the same version?

Andrew.