RE: [sv-cc] Item 160: define 'linearize' multiple packed dimensions

Ralph, one question:

The new E.6.6 says that "C ranges may also be linearized,"
which in IEEE-speak means that they may not be
linearized in some cases. Since the new E.10.1 says
that the "Multiple packed dimensions are linearized,"
should E.6.6 say that "C ranges are also linearized"?

Regards,
Jim

---------------------------------------------------------
James H. Vellenga 978-262-6381
Engineering Director (FAX) 978-262-6636
Cadence Design Systems, Inc. vellenga@cadence.com
270 Billerica Rd
Chelmsford, MA 01824-4179
"We all work with partial information."
----------------------------------------------------------
  
 

] -----Original Message-----
] From: owner-sv-cc@eda.org [mailto:owner-sv-cc@eda.org] On
] Behalf Of Duncan, Ralph
] Sent: Monday, November 29, 2004 1:40 PM
] To: SV-CC
] Subject: [sv-cc] Item 160: define 'linearize' multiple packed
] dimensions
]
] We had hoped to make our DPI material on linearizing multiple
] packed array dimensions reference new LRM material defining the term.
]
] Since it is unlikely that another committee will produce such
] material
] in time, here is an attempt to integrate such material into our
] sections,
] as part of Mantis item 160.
]
] Adding a definition of array linearization to the DPI
] sections involves:
]
] . adding the definition, itself,
] . interweaving it with existing array material.
]
] The existing organization is:
] . E.6.1 mentions array 'linearization' and normalization' but does not
] define them.
] . E.6.5 "Normalized Ranges" -- defines 'normalization.'
] . E.6.6 "Mapping between SystemVerilog ranges and normalized
] ranges" --
] mentions
] both linearization and normalization.
]
] The proposed organization is:
] . E.6.1 -- add a cross-reference to an expanded section E.6.5.
] . E.6.5 -- Define both array terms here:
] . change title to "Normalized and linearized ranges"
] . add a paragraph defining 'linearization.'
] . E.6.6 -- Make it explicit that we are dealing with
] both concepts here:
]
] . change title to "Mapping between SystemVerilog ranges
] and C ranges,"
] . Add a sentence about linearization.
] . Add relevant cross-references.
]
] A proposal that reflects these suggestions has been uploaded to the
] Mantis database.
] (A copy of the proposal is attached to this email in HTML format).
]
] Thank you for considering this material.
]
] Ralph Duncan
]
]
]
]
Received on Mon Nov 29 12:03:27 2004