首页资源分类FPGA/CPLD > Verilog-2001

Verilog-2001

已有 445109个资源

下载专区

上传者其他资源

    文档信息举报收藏

    标    签:Verilog

    分    享:

    文档简介

    IEEE-Verilog-2001,PDF文档

    文档预览

    IEEE Standards IEEE Std 1364-2001 (Revision of IEEE Std 1364-1995) IEEE Standard Verilog® Hardware Description Language IEEE Computer Society Sponsored by the Design Automation Standards Committee Published by The Institute of Electrical and Electronics Engineers, Inc. 3 Park Avenue, New York, NY 10016-5997, USA 28 September 2001 Print: SH94921 PDF: SS94921 IEEE Std 1364-2001 Version C (Revision of IEEE Std 1364-1995) IEEE Standard Verilog® Hardware Description Language Sponsor Design Automation Standards Committee of the IEEE Computer Society Approved 17 March 2001 IEEE-SA Standards Board Abstract: The Verilog® Hardware Description Language (HDL) is defined in this standard. Verilog HDL is a formal notation intended for use in all phases of the creation of electronic systems. Because it is both machine readable and human readable, it supports the development, verification, synthesis, and testing of hardware designs; the communication of hardware design data; and the maintenance, modification, and procurement of hardware. The primary audiences for this standard are the implementors of tools supporting the language and advanced users of the language. Keywords: computer, computer languages, digital systems, electronic systems, hardware, hardware description languages, hardware design, HDL, PLI, programming language interface, Verilog HDL, Verilog PLI, Verilog® The Institute of Electrical and Electronics Engineers, Inc. 3 Park Avenue, New York, NY 10016-5997, USA Copyright © 2001 by the Institute of Electrical and Electronics Engineers, Inc. All rights reserved. Published 28 September 2001. Printed in the United States of America. Print: ISBN 0-7381-2826-0 SH94921 PDF: ISBN 0-7381-2827-9 SS94921 No part of this publication may be reproduced in any form, in an electronic retrieval system or otherwise, without the prior written permission of the publisher. IEEE Standards documents are developed within the IEEE Societies and the Standards Coordinating Committees of the IEEE Standards Association (IEEE-SA) Standards Board. The IEEE develops its standards through a consensus development process, approved by the American National Standards Institute, which brings together volunteers representing varied viewpoints and interests to achieve the final product. Volunteers are not necessarily members of the Institute and serve without compensation. While the IEEE administers the process and establishes rules to promote fairness in the consensus development process, the IEEE does not independently evaluate, test, or verify the accuracy of any of the information contained in its standards. Use of an IEEE Standard is wholly voluntary. The IEEE disclaims liability for any personal injury, property or other damage, of any nature whatsoever, whether special, indirect, consequential, or compensatory, directly or indirectly resulting from the publication, use of, or reliance upon this, or any other IEEE Standard document. The IEEE does not warrant or represent the accuracy or content of the material contained herein, and expressly disclaims any express or implied warranty, including any implied warranty of merchantability or fitness for a specific purpose, or that the use of the material contained herein is free from patent infringement. IEEE Standards documents are supplied “AS IS.” The existence of an IEEE Standard does not imply that there are no other ways to produce, test, measure, purchase, market, or provide other goods and services related to the scope of the IEEE Standard. Furthermore, the viewpoint expressed at the time a standard is approved and issued is subject to change brought about through developments in the state of the art and comments received from users of the standard. Every IEEE Standard is subjected to review at least every five years for revision or reaffirmation. When a document is more than five years old and has not been reaffirmed, it is reasonable to conclude that its contents, although still of some value, do not wholly reflect the present state of the art. Users are cautioned to check to determine that they have the latest edition of any IEEE Standard. In publishing and making this document available, the IEEE is not suggesting or rendering professional or other services for, or on behalf of, any person or entity. Nor is the IEEE undertaking to perform any duty owed by any other person or entity to another. Any person utilizing this, and any other IEEE Standards document, should rely upon the advice of a competent professional in determining the exercise of reasonable care in any given circumstances. Interpretations: Occasionally questions may arise regarding the meaning of portions of standards as they relate to specific applications. When the need for interpretations is brought to the attention of IEEE, the Institute will initiate action to prepare appropriate responses. Since IEEE Standards represent a consensus of concerned interests, it is important to ensure that any interpretation has also received the concurrence of a balance of interests. For this reason, IEEE and the members of its societies and Standards Coordinating Committees are not able to provide an instant response to interpretation requests except in those cases where the matter has previously received formal consideration. Comments for revision of IEEE Standards are welcome from any interested party, regardless of membership affiliation with IEEE. Suggestions for changes in documents should be in the form of a proposed change of text, together with appropriate supporting comments. Comments on standards and requests for interpretations should be addressed to: Secretary, IEEE-SA Standards Board 445 Hoes Lane P.O. Box 1331 Piscataway, NJ 08855-1331 USA Note: Attention is called to the possibility that implementation of this standard may require use of subject matter covered by patent rights. By publication of this standard, no position is taken with respect to the existence or validity of any patent rights in connection therewith. The IEEE shall not be responsible for identifying patents for which a license may be required by an IEEE standard or for conducting inquiries into the legal validity or scope of those patents that are brought to its attention. The IEEE and its designees are the sole entities that may authorize the use of IEEE-owned certification marks and/or trademarks to indicate compliance with the materials set forth herein. Authorization to photocopy portions of any individual standard for internal or personal use is granted by the Institute of Electrical and Electronics Engineers, Inc., provided that the appropriate fee is paid to Copyright Clearance Center. To arrange for payment of licensing fee, please contact Copyright Clearance Center, Customer Service, 222 Rosewood Drive, Danvers, MA 01923 USA; (978) 750-8400. Permission to photocopy portions of any individual standard for educational classroom use can also be obtained through the Copyright Clearance Center. About IEEE Std 1364-2001 Version C and the Errata During the past two years, the IEEE 1364 Working Group’s Errata Task Force has thoroughly reviewed the Standard and has identified and corrected a number of production and editorial errors that crept in between balloting and printing of the Standard. IEEE Std 1364-2001 Version C incorporates all of these corrections. In addition, during the IEEE 1364 Working Group’s review of the Standard, the Working Group and its Errata Task Force identified other areas where the standard had logical inconsistencies which were not the result of production problems. The Working Group developed an Errata document that identifies these and specifies the Working Group’s statement as to the correct interpretation of the Standard. Copyright © 2001 IEEE. All rights reserved. iii Participants—Version C and Errata At the time IEEE Std 1364-2001 Version C and the Errata were completed, the IEEE 1364 Working Group had the following membership: Michael T. Y. (Mac) McNamara, Chair Shalom Bresticker, Editor Stefen Boyd, Web Master Kurt Baty Dennis Brophy Clifford E. Cummings Charles Dawson Tom Fitzpatrick Krishna Garlapati Keith Gover Ennis Hawk Richard Ho Atsushi Kasuya Jay Lawrence Andrew Lynch James A. Markevitch Dennis Marsa Francoise Martinolle Mehdi Mohtashemi Anders Nordstrom Karen Pieper Brad Pierce Steven Sharp Alec Stanculescu Stuart Sutherland Chong Guan Tan Gordon Vreugdenhil The Errata Task Force had the following membership: Karen Pieper, Chair Stefen Boyd, Vice Chair Kurt Baty Shalom Bresticker Dennis Brophy Clifford E. Cummings Charles Dawson Ted Elkind Tom Fitzpatrick Jay Lawrence Andrew Lynch James A. Markevitch Dennis Marsa Francoise Martinolle Michael T. Y. (Mac) McNamara Elliot Mednick Don Mills Mehdi Mohtashemi Anders Nordstrom Brad Pierce David Roberts Steven Sharp David Smith Stuart Sutherland Gordon Vreugdenhil The Behavioral Task Force had the following membership: Steven Sharp, Chair Kurt Baty Stefen Boyd Dennis Brophy Clifford E. Cummings Tom Fitzpatrick Ennis Hawk Atsushi Kasuya Jay Lawrence Francoise Martinolle Michael T. Y. (Mac) McNamara Don Mills The PLI Task Force had the following membership: Charles Dawson, Co-Chair Stuart Sutherland, Co-Chair Steven Dovich Dennis Marsa Francoise Martinolle Mehdi Mohtashemi Karen Pieper Brad Pierce Alec Stanculescu Stuart Sutherland Gordon Vreugdenhil Nisa Parikh David Roberts iv Copyright © 2001 IEEE. All rights reserved. Introduction (This introduction is not part of IEEE Std 1364-2001, IEEE Standard Verilog® Hardware Description Language.) The Verilog® Hardware Description Language (Verilog HDL) became an IEEE standard in 1995 as IEEE Std 1364-1995. It was designed to be simple, intuitive, and effective at multiple levels of abstraction in a standard textual format for a variety of design tools, including verification simulation, timing analysis, test analysis, and synthesis. It is because of these rich features that Verilog has been accepted to be the language of choice by an overwhelming number of IC designers. Verilog contains a rich set of built-in primitives, including logic gates, user-definable primitives, switches, and wired logic. It also has device pin-to-pin delays and timing checks. The mixing of abstract levels is essentially provided by the semantics of two data types: nets and variables. Continuous assignments, in which expressions of both variables and nets can continuously drive values onto nets, provide the basic structural construct. Procedural assignments, in which the results of calculations involving variable and net values can be stored into variables, provide the basic behavioral construct. A design consists of a set of modules, each of which has an I/O interface, and a description of its function, which can be structural, behavioral, or a mix. These modules are formed into a hierarchy and are interconnected with nets. The Verilog language is extensible via the Programming Language Interface (PLI) and the Verilog Procedural Interface (VPI) routines. The PLI/VPI is a collection of routines that allows foreign functions to access information contained in a Verilog HDL description of the design and facilitates dynamic interaction with simulation. Applications of PLI/VPI include connecting to a Verilog HDL simulator with other simulation and CAD systems, customized debugging tasks, delay calculators, and annotators. The language that influenced Verilog HDL the most was HILO-2, which was developed at Brunel University in England under a contract to produce a test generation system for the British Ministry of Defense. HILO-2 successfully combined the gate and register transfer levels of abstraction and supported verification simulation, timing analysis, fault simulation, and test generation. In 1990, Cadence Design Systems placed the Verilog HDL into the public domain and the independent Open Verilog International (OVI) was formed to manage and promote Verilog HDL. In 1992, the Board of Directors of OVI began an effort to establish Verilog HDL as an IEEE standard. In 1993, the first IEEE Working Group was formed and after 18 months of focused efforts Verilog became an IEEE standard as IEEE Std 1364-1995. After the standardization process was complete the 1364 Working Group started looking for feedback from 1364 users worldwide so the standard could be enhanced and modified accordingly. This led to a five year effort to get a much better Verilog standard in IEEE Std 1364-2001. Objective of the IEEE Std 1364-2001 effort The starting point for the IEEE 1364 Working Group for this standard was the feedback received from the IEEE Std 1364-1995 users worldwide. It was clear from the feedback that users wanted improvements in all aspects of the language. Users at the higher levels wanted to expand and improve the language at the RTL and behavioral levels, while users at the lower levels wanted improved capability for ASIC designs and signoff. It was for this reason that the 1364 Working Group was organized into three task forces: Behavioral, ASIC, and PLI. Copyright © 2001 IEEE. All rights reserved. v The clear directive from the users for these three task forces was to start by solving some of the following problems: — Consolidate existing IEEE Std 1364-1995 — Verilog Generate statement — Multi-dimensional arrays — Enhanced Verilog file I/O — Re-entrant tasks — Standardize Verilog configurations — Enhance timing representation — Enhance the VPI routines Achievements Over a period of four years the 1364 Verilog Standards Group (VSG) has produced five drafts of the LRM. The three task forces went through the IEEE Std 1364-1995 LRM very thoroughly and in the process of consolidating the existing LRM have been able to provide nearly three hundred clarifications and errata for the Behavioral, ASIC, and PLI sections. In addition, the VSG has also been able to agree on all the enhancements that were requested (including the ones stated above). Three new sections have been added. Clause 13, “Configuring the contents of a design,” deals with configuration management and has been added to facilitate both the sharing of Verilog designs between designers and/or design groups and the repeatability of the exact contents of a given simulation session. Clause 15, “Timing checks,” has been broken out of Clause 17, “System tasks and functions,” and details more fully how timing checks are used in specify blocks. Clause 16, “Backannotation using the Standard Delay Format (SDF),” addresses using back annotation (IEEE Std 1497-1999) within IEEE Std 1364-2001. Extreme care has been taken to enhance the VPI routines to handle all the enhancements in the Behavioral and other areas of the LRM. Minimum work has been done on the PLI routines and most of the work has been concentrated on the VPI routines. Some of the enhancements in the VPI are the save and restart, simulation control, work area access, error handling, assign/deassign and support for array of instances, generate, and file I/O. Work on this standard would not have been possible without funding from the CAS society of the IEEE and Open Verilog International. The IEEE Std 1364-2001 Verilog Standards Group organization Many individuals from many different organizations participated directly or indirectly in the standardization process. The main body of the IEEE Std 1364-2001 working group is located in the United States, with a subgroup in Japan (EIAJ/1364HDL). The members of the IEEE Std 1364-2001 working group had voting privileges and all motions had to be approved by this group to be implemented. The three task forces focused on their specific areas and their recommendations were eventually voted on by the IEEE Std 1364-2001 working group. vi Copyright © 2001 IEEE. All rights reserved. At the time this document was approved, the IEEE Std 1364-2001 working group had the following membership: Maqsoodul (Maq) Mannan, Chair Kasumi Hamaguchi, Vice Chair (Japan) Alec G. Stanculescu, Vice Chair (USA) Lynn A. Horobin, Secretary Yatin Trivedi, Technical Editor The Behavioral Task Force consisted of the following members: Clifford E. Cummings, Leader Kurt Baty Stefen Boyd Shalom Bresticker Tom Fitzpatrick Adam Krolnik James A. Markevitch Michael McNamara Anders Nordstrom Karen Pieper Steven Sharp Chris Spear Stuart Sutherland The ASIC Task Force consisted of the following members: Steve Wadsworth, Leader Leigh Brady Paul Colwill Tom Dewey Ted Elkind Naveen Gupta Prabhakaran Krishnamurthy Marek Ryniejski Lukasz Senator The PLI Task Force consisted of the following members: Andrew T. Lynch, Leader Stuart Sutherland, Co-Leader and Editor Deborah J. Dalio Charles Dawson Steve Meyer Girish S. Rao David Roberts The IEEE 1364 Japan subgroup (EIAJ/1364HDL) consisted of the following members: Kasumi Hamaguchi, Vice Chair (Japan) Yokozeki Atsushi Yasuaki Hatta Makoto Makino Takashima Mitsuya Tatsuro Nakamura Hiroaki Nishi Tsutomu Someya Copyright © 2001 IEEE. All rights reserved. vii The following members of the balloting committee voted on this standard: Guy Adam Shigehiro Asano Peter J. Ashenden Victor Berman J Bhasker Stefan Boyd Dennis B. Brophy Keith Chow Clifford E. Cummings Brian A. Dalio Timothy R. Davis Charles Dawson Douglas D. Dunlop Ted Elkind Joerg-Oliver Fischer-Binder Peter Flake Robert A. Flatt Masahiro Fukui Kenji Goto Naveen Gupta Andrew Guyler Yoshiaki Hagiwara Anne C. Harris Lynn A. Horobin ChiLai Huang Takahiro Ichinomiya Masato Ikeda Mitsuaki Ishikawa Neil G. Jacobson Richard O. Jones Osamu Karatsu Jake Karrfalt Masayuki Katakura Kaoru Kawamura Masamichi Kawarabayashi Satoshi Kojima Masuyoshi Kurokawa Gunther Lehmann Andrew T. Lynch Serge Maginot Maqsoodul Mannan James A. Markevitch Francoise Martinolle Yoshio Masubuchi Paul J. Menchini Hiroshi Mizuno Egbert Molenkamp John T. Montague Akira Motohara Hiroaki Nishi Anders Nordstrom Ryosuke Okuda Yoichi Onishi Uma P. Parvathy William R. Paulsen Karen L. Pieper Girish S. Rao Jaideep Roy Francesco Sforza Charles F. Shelor Chris Spear Alec G. Stanculescu Steve Start Stuart Sutherland Masahiko Toyonaga Yatin K. Trivedi Cary Ussery Steven D. Wadsworth Sui-Ki Wan Ronald Waxman John M. Williams John Willis Takashi Yamada Lun Ye Hirokazu Yonezawa Tetsuo Yutani Mark Zwolinski When the IEEE-SA Standards Board approved this standard on 17 March 2001, it had the following membership: Donald N. Heirman, Chair James T. Carlo, Vice Chair Judith Gorman, Secretary Satish K. Aggarwal Mark D. Bowman Gary R. Engmann Harold E. Epstein H. Landis Floyd Jay Forster* Howard M. Frazier Ruben D. Garzon James H. Gurney Richard J. Holleman Lowell G. Johnson Robert J. Kennelly Joseph L. Koepfinger* Peter H. Lips L. Bruce McClung Daleep C. Mohla James W. Moore Robert F. Munzner Ronald C. Petersen Gerald H. Peterson John B. Posey Gary S. Robinson Akio Tojo Donald W. Zipse *Member Emeritus Also included is the following nonvoting IEEE-SA Standards Board liaison: Alan Cookson, NIST Representative Donald R. Volzka, TAB Representative Andrew Ickowicz IEEE Standards Project Editor Verilog is a registered trademark of Cadence Design Systems, Inc. viii Copyright © 2001 IEEE. All rights reserved. Contents 1. Overview.............................................................................................................................................. 1 1.1 Objectives of this standard......................................................................................................... 1 1.2 Conventions used in this standard.............................................................................................. 1 1.3 Syntactic description.................................................................................................................. 2 1.4 Contents of this standard............................................................................................................ 2 1.5 Header file listings ..................................................................................................................... 4 1.6 Examples.................................................................................................................................... 5 1.7 Prerequisites............................................................................................................................... 5 2. Lexical conventions ............................................................................................................................. 6 2.1 Lexical tokens ............................................................................................................................ 6 2.2 White space................................................................................................................................ 6 2.3 Comments .................................................................................................................................. 6 2.4 Operators.................................................................................................................................... 6 2.5 Numbers..................................................................................................................................... 6 2.5.1 Integer constants ........................................................................................................... 7 2.5.2 Real constants ............................................................................................................. 10 2.5.3 Conversion .................................................................................................................. 10 2.6 Strings ...................................................................................................................................... 10 2.6.1 String variable declaration .......................................................................................... 11 2.6.2 String manipulation..................................................................................................... 11 2.6.3 Special characters in strings........................................................................................ 11 2.7 Identifiers, keywords, and system names ................................................................................ 12 2.7.1 Escaped identifiers ...................................................................................................... 12 2.7.2 Generated identifiers................................................................................................... 13 2.7.3 Keywords .................................................................................................................... 13 2.7.4 System tasks and functions ......................................................................................... 13 2.7.5 Compiler directives..................................................................................................... 14 2.8 Attributes.................................................................................................................................. 14 2.8.1 Examples..................................................................................................................... 15 2.8.2 Syntax ......................................................................................................................... 16 3. Data types........................................................................................................................................... 20 3.1 Value set................................................................................................................................... 20 3.2 Nets and variables .................................................................................................................... 20 3.2.1 Net declarations .......................................................................................................... 20 3.2.2 Variable declarations .................................................................................................. 22 3.3 Vectors ..................................................................................................................................... 23 3.3.1 Specifying vectors....................................................................................................... 23 3.3.2 Vector net accessibility ............................................................................................... 24 3.4 Strengths .................................................................................................................................. 24 3.4.1 Charge strength ........................................................................................................... 24 3.4.2 Drive strength.............................................................................................................. 24 3.5 Implicit declarations................................................................................................................. 25 3.6 Net initialization....................................................................................................................... 25 3.7 Net types .................................................................................................................................. 25 3.7.1 Wire and tri nets.......................................................................................................... 25 3.7.2 Wired nets ................................................................................................................... 26 Copyright © 2001 IEEE. All rights reserved. ix 3.7.3 Trireg net..................................................................................................................... 26 3.7.4 Tri0 and tri1 nets......................................................................................................... 30 3.7.5 Supply nets.................................................................................................................. 31 3.8 Regs.......................................................................................................................................... 31 3.9 Integers, reals, times, and realtimes ......................................................................................... 31 3.9.1 Operators and real numbers ........................................................................................ 32 3.9.2 Conversion .................................................................................................................. 32 3.10 Arrays....................................................................................................................................... 33 3.10.1 Net arrays .................................................................................................................... 33 3.10.2 reg and variable arrays ................................................................................................ 33 3.10.3 Memories .................................................................................................................... 33 3.11 Parameters................................................................................................................................ 34 3.11.1 Module parameters...................................................................................................... 35 3.11.2 Local parameters - localparam.................................................................................... 36 3.11.3 Specify parameters...................................................................................................... 37 3.12 Name spaces............................................................................................................................. 38 4. Expressions ........................................................................................................................................ 40 4.1 Operators.................................................................................................................................. 40 4.1.1 Operators with real operands ...................................................................................... 41 4.1.2 Operator precedence ................................................................................................... 42 4.1.3 Using integer numbers in expressions ........................................................................ 43 4.1.4 Expression evaluation order........................................................................................ 43 4.1.5 Arithmetic operators ................................................................................................... 44 4.1.6 Arithmetic expressions with regs and integers ........................................................... 45 4.1.7 Relational operators .................................................................................................... 46 4.1.8 Equality operators ....................................................................................................... 46 4.1.9 Logical operators ........................................................................................................ 47 4.1.10 Bit-wise operators ....................................................................................................... 47 4.1.11 Reduction operators .................................................................................................... 48 4.1.12 Shift operators............................................................................................................. 49 4.1.13 Conditional operator ................................................................................................... 50 4.1.14 Concatenations............................................................................................................ 51 4.1.15 Event or....................................................................................................................... 52 4.2 Operands .................................................................................................................................. 52 4.2.1 Vector bit-select and part-select addressing ............................................................... 52 4.2.2 Array and memory addressing .................................................................................... 54 4.2.3 Strings ......................................................................................................................... 55 4.3 Minimum, typical, and maximum delay expressions .............................................................. 57 4.4 Expression bit lengths .............................................................................................................. 59 4.4.1 Rules for expression bit lengths.................................................................................. 59 4.4.2 An example of an expression bit-length problem ....................................................... 60 4.4.3 Example of self-determined expressions .................................................................... 61 4.5 Signed expressions................................................................................................................... 62 4.5.1 Rules for expression types .......................................................................................... 62 4.5.2 Steps for evaluating an expression.............................................................................. 62 4.5.3 Steps for evaluating an assignment............................................................................. 63 4.5.4 Handling X and Z in signed expressions .................................................................... 63 5. Scheduling semantics......................................................................................................................... 64 5.1 Execution of a model ............................................................................................................... 64 5.2 Event simulation ...................................................................................................................... 64 x Copyright © 2001 IEEE. All rights reserved. 5.3 The stratified event queue........................................................................................................ 64 5.4 The Verilog simulation reference model ................................................................................. 65 5.4.1 Determinism................................................................................................................ 66 5.4.2 Nondeterminism.......................................................................................................... 66 5.5 Race conditions........................................................................................................................ 66 5.6 Scheduling implication of assignments ................................................................................... 66 5.6.1 Continuous assignment ............................................................................................... 67 5.6.2 Procedural continuous assignment.............................................................................. 67 5.6.3 Blocking assignment................................................................................................... 67 5.6.4 Nonblocking assignment............................................................................................. 67 5.6.5 Switch (transistor) processing..................................................................................... 67 5.6.6 Port connections.......................................................................................................... 68 5.6.7 Functions and tasks ..................................................................................................... 68 6. Assignments....................................................................................................................................... 69 6.1 Continuous assignments .......................................................................................................... 69 6.1.1 The net declaration assignment................................................................................... 70 6.1.2 The continuous assignment statement ........................................................................ 70 6.1.3 Delays ......................................................................................................................... 72 6.1.4 Strength ....................................................................................................................... 72 6.2 Procedural assignments............................................................................................................ 73 6.2.1 Variable declaration assignment ................................................................................. 73 6.2.2 Variable declaration syntax......................................................................................... 74 7. Gate and switch level modeling......................................................................................................... 75 7.1 Gate and switch declaration syntax.......................................................................................... 75 7.1.1 The gate type specification ......................................................................................... 77 7.1.2 The drive strength specification.................................................................................. 77 7.1.3 The delay specification ............................................................................................... 78 7.1.4 The primitive instance identifier................................................................................. 78 7.1.5 The range specification ............................................................................................... 78 7.1.6 Primitive instance connection list ............................................................................... 79 7.2 and, nand, nor, or, xor, and xnor gates..................................................................................... 81 7.3 buf and not gates ...................................................................................................................... 82 7.4 bufif1, bufif0, notif1, and notif0 gates..................................................................................... 83 7.5 MOS switches .......................................................................................................................... 84 7.6 Bidirectional pass switches ...................................................................................................... 86 7.7 CMOS switches ....................................................................................................................... 86 7.8 pullup and pulldown sources ................................................................................................... 87 7.9 Logic strength modeling .......................................................................................................... 88 7.10 Strengths and values of combined signals ............................................................................... 89 7.10.1 Combined signals of unambiguous strength ............................................................... 89 7.10.2 Ambiguous strengths: sources and combinations ....................................................... 90 7.10.3 Ambiguous strength signals and unambiguous signals .............................................. 95 7.10.4 Wired logic net types .................................................................................................. 99 7.11 Strength reduction by nonresistive devices............................................................................ 102 7.12 Strength reduction by resistive devices.................................................................................. 102 7.13 Strengths of net types............................................................................................................. 102 7.13.1 tri0 and tri1 net strengths .......................................................................................... 102 7.13.2 trireg strength ............................................................................................................ 102 7.13.3 supply0 and supply1 net strengths ............................................................................ 102 Copyright © 2001 IEEE. All rights reserved. xi 7.14 Gate and net delays ................................................................................................................ 103 7.14.1 min:typ:max delays................................................................................................... 104 7.14.2 trireg net charge decay .............................................................................................. 105 8. User-defined primitives (UDPs) ...................................................................................................... 107 8.1 UDP definition ....................................................................................................................... 107 8.1.1 UDP header............................................................................................................... 109 8.1.2 UDP port declarations............................................................................................... 109 8.1.3 Sequential UDP initial statement .............................................................................. 109 8.1.4 UDP state table ......................................................................................................... 109 8.1.5 Z values in UDP........................................................................................................ 110 8.1.6 Summary of symbols ................................................................................................ 110 8.2 Combinational UDPs ............................................................................................................. 111 8.3 Level-sensitive sequential UDPs ........................................................................................... 112 8.4 Edge-sensitive sequential UDPs ............................................................................................ 112 8.5 Sequential UDP initialization ................................................................................................ 113 8.6 UDP instances........................................................................................................................ 115 8.7 Mixing level-sensitive and edge-sensitive descriptions......................................................... 116 8.8 Level-sensitive dominance..................................................................................................... 117 9. Behavioral modeling........................................................................................................................ 118 9.1 Behavioral model overview ................................................................................................... 118 9.2 Procedural assignments.......................................................................................................... 119 9.2.1 Blocking procedural assignments ............................................................................. 119 9.2.2 The nonblocking procedural assignment .................................................................. 121 9.3 Procedural continuous assignments ....................................................................................... 124 9.3.1 The assign and deassign procedural statements........................................................ 125 9.3.2 The force and release procedural statements ............................................................ 126 9.4 Conditional statement ............................................................................................................ 127 9.4.1 If-else-if construct..................................................................................................... 128 9.5 Case statement ....................................................................................................................... 130 9.5.1 Case statement with don’t-cares ............................................................................... 133 9.5.2 Constant expression in case statement...................................................................... 133 9.6 Looping statements ................................................................................................................ 134 9.7 Procedural timing controls..................................................................................................... 136 9.7.1 Delay control............................................................................................................. 137 9.7.2 Event control............................................................................................................. 138 9.7.3 Named events............................................................................................................ 138 9.7.4 Event or operator....................................................................................................... 139 9.7.5 Implicit event_expression list ................................................................................... 140 9.7.6 Level-sensitive event control .................................................................................... 141 9.7.7 Intra-assignment timing controls .............................................................................. 142 9.8 Block statements .................................................................................................................... 146 9.8.1 Sequential blocks ...................................................................................................... 146 9.8.2 Parallel blocks........................................................................................................... 147 9.8.3 Block names.............................................................................................................. 148 9.8.4 Start and finish times ................................................................................................ 148 9.9 Structured procedures ............................................................................................................ 149 9.9.1 Initial construct ......................................................................................................... 150 9.9.2 Always construct....................................................................................................... 150 xii Copyright © 2001 IEEE. All rights reserved. 10. Tasks and functions.......................................................................................................................... 152 10.1 Distinctions between tasks and functions .............................................................................. 152 10.2 Tasks and task enabling ......................................................................................................... 152 10.2.1 Task declarations ...................................................................................................... 153 10.2.2 Task enabling and argument passing ........................................................................ 154 10.2.3 Task memory usage and concurrent activation......................................................... 156 10.3 Functions and function calling............................................................................................... 157 10.3.1 Function declarations ................................................................................................ 158 10.3.2 Returning a value from a function ............................................................................ 159 10.3.3 Calling a function...................................................................................................... 160 10.3.4 Function rules............................................................................................................ 160 10.3.5 Use of constant functions.......................................................................................... 161 11. Disabling of named blocks and tasks............................................................................................... 163 12. Hierarchical structures ..................................................................................................................... 166 12.1 Modules.................................................................................................................................. 166 12.1.1 Top-level modules .................................................................................................... 168 12.1.2 Module instantiation ................................................................................................. 168 12.1.3 Generated instantiation ............................................................................................. 170 12.2 Overriding module parameter values..................................................................................... 180 12.2.1 defparam statement ................................................................................................... 182 12.2.2 Module instance parameter value assignment .......................................................... 183 12.2.3 Parameter dependence .............................................................................................. 185 12.3 Ports ....................................................................................................................................... 185 12.3.1 Port definition ........................................................................................................... 185 12.3.2 List of ports ............................................................................................................... 185 12.3.3 Port declarations........................................................................................................ 186 12.3.4 List of ports declarations........................................................................................... 188 12.3.5 Connecting module instance ports by ordered list.................................................... 188 12.3.6 Connecting module instance ports by name ............................................................. 189 12.3.7 Real numbers in port connections............................................................................. 190 12.3.8 Connecting dissimilar ports ...................................................................................... 191 12.3.9 Port connection rules................................................................................................. 191 12.3.10 Net types resulting from dissimilar port connections ............................................... 192 12.3.11 Connecting signed values via ports........................................................................... 193 12.4 Hierarchical names ................................................................................................................ 193 12.5 Upwards name referencing .................................................................................................... 196 12.6 Scope rules ............................................................................................................................ 198 13. Configuring the contents of a design ............................................................................................... 200 13.1 Introduction............................................................................................................................ 200 13.1.1 Library notation ........................................................................................................ 200 13.1.2 Basic configuration elements.................................................................................... 201 13.2 Libraries ................................................................................................................................. 201 13.2.1 Specifying libraries - the library map file ................................................................. 201 13.2.2 Using multiple library mapping files ........................................................................ 203 13.2.3 Mapping source files to libraries............................................................................... 203 13.3 Configurations........................................................................................................................ 203 13.3.1 Basic configuration syntax........................................................................................ 203 13.3.2 Hierarchical configurations....................................................................................... 206 Copyright © 2001 IEEE. All rights reserved. xiii 13.4 Using libraries and configs .................................................................................................... 207 13.4.1 Precompiling in a single-pass use-model.................................................................. 207 13.4.2 Elaboration-time compiling in a single-pass use-model........................................... 207 13.4.3 Precompiling using a separate compilation tool ....................................................... 207 13.4.4 Command line considerations................................................................................... 207 13.5 Configuration examples ......................................................................................................... 208 13.5.1 Default configuration from library map file ............................................................. 208 13.5.2 Using the default clause ............................................................................................ 208 13.5.3 Using the cell clause ................................................................................................. 209 13.5.4 Using the instance clause .......................................................................................... 209 13.5.5 Using a hierarchical config ....................................................................................... 209 13.6 Displaying library binding information ................................................................................. 210 13.7 Library mapping examples .................................................................................................... 210 13.7.1 Using the command line to control library searching............................................... 210 13.7.2 File path specification examples............................................................................... 210 13.7.3 Resolving multiple path specifications ..................................................................... 211 14. Specify blocks.................................................................................................................................. 212 14.1 Specify block declaration....................................................................................................... 212 14.2 Module path declarations....................................................................................................... 213 14.2.1 Module path restrictions ........................................................................................... 214 14.2.2 Simple module paths................................................................................................. 214 14.2.3 Edge-sensitive paths.................................................................................................. 215 14.2.4 State-dependent paths ............................................................................................... 216 14.2.5 Full connection and parallel connection paths.......................................................... 220 14.2.6 Declaring multiple module paths in a single statement ............................................ 221 14.2.7 Module path polarity................................................................................................. 222 14.3 Assigning delays to module paths.......................................................................................... 223 14.3.1 Specifying transition delays on module paths .......................................................... 224 14.3.2 Specifying x transition delays................................................................................... 225 14.3.3 Delay selection.......................................................................................................... 226 14.4 Mixing module path delays and distributed delays................................................................ 227 14.5 Driving wired logic ................................................................................................................ 228 14.6 Detailed control of pulse filtering behavior ........................................................................... 229 14.6.1 Specify block control of pulse limit values............................................................... 230 14.6.2 Global control of pulse limit values.......................................................................... 231 14.6.3 SDF annotation of pulse limit values........................................................................ 231 14.6.4 Detailed pulse control capabilities ............................................................................ 232 15. Timing checks.................................................................................................................................. 238 15.1 Overview................................................................................................................................ 238 15.2 Timing checks using a stability window................................................................................ 241 15.2.1 $setup ........................................................................................................................ 242 15.2.2 $hold ......................................................................................................................... 242 15.2.3 $setuphold ................................................................................................................. 243 15.2.4 $removal ................................................................................................................... 245 15.2.5 $recovery................................................................................................................... 246 15.2.6 $recrem ..................................................................................................................... 247 15.3 Timing checks for clock and control signals ......................................................................... 249 15.3.1 $skew ........................................................................................................................ 250 15.3.2 $timeskew ................................................................................................................. 251 15.3.3 $fullskew................................................................................................................... 253 xiv Copyright © 2001 IEEE. All rights reserved. 15.3.4 $width ....................................................................................................................... 255 15.3.5 $period ...................................................................................................................... 256 15.3.6 $nochange ................................................................................................................. 257 15.4 Edge-control specifiers .......................................................................................................... 258 15.5 Notifiers: user-defined responses to timing violations .......................................................... 260 15.5.1 Requirements for accurate simulation ...................................................................... 262 15.5.2 Conditions in negative timing checks ....................................................................... 264 15.5.3 Notifiers in negative timing checks .......................................................................... 266 15.5.4 Option behavior ........................................................................................................ 266 15.6 Enabling timing checks with conditioned events................................................................... 266 15.7 Vector signals in timing checks ............................................................................................. 267 15.8 Negative timing checks.......................................................................................................... 268 16. Backannotation using the Standard Delay Format (SDF)................................................................ 270 16.1 The SDF annotator................................................................................................................. 270 16.2 Mapping of SDF constructs to Verilog.................................................................................. 270 16.2.1 Mapping of SDF delay constructs to Verilog declarations....................................... 270 16.2.2 Mapping of SDF timing check constructs to Verilog ............................................... 272 16.2.3 SDF annotation of specparams ................................................................................. 273 16.2.4 SDF annotation of interconnect delays ..................................................................... 274 16.3 Multiple annotations .............................................................................................................. 275 16.4 Multiple SDF files ................................................................................................................. 276 16.5 Pulse limit annotation ............................................................................................................ 276 16.6 SDF to Verilog delay value mapping..................................................................................... 277 17. System tasks and functions .............................................................................................................. 278 17.1 Display system tasks .............................................................................................................. 278 17.1.1 The display and write tasks....................................................................................... 279 17.1.2 Strobed monitoring ................................................................................................... 286 17.1.3 Continuous monitoring ............................................................................................. 287 17.2 File input-output system tasks and functions......................................................................... 287 17.2.1 Opening and closing files.......................................................................................... 287 17.2.2 File output system tasks ............................................................................................ 289 17.2.3 Formatting data to a string ........................................................................................ 290 17.2.4 Reading data from a file............................................................................................ 291 17.2.5 File positioning ......................................................................................................... 295 17.2.6 Flushing output ......................................................................................................... 295 17.2.7 I/O error status .......................................................................................................... 295 17.2.8 Loading memory data from a file ............................................................................. 296 17.2.9 Loading timing data from an SDF file...................................................................... 297 17.3 Timescale system tasks .......................................................................................................... 298 17.3.1 $printtimescale.......................................................................................................... 298 17.3.2 $timeformat............................................................................................................... 299 17.4 Simulation control system tasks ............................................................................................ 302 17.4.1 $finish ....................................................................................................................... 302 17.4.2 $stop.......................................................................................................................... 302 17.5 PLA modeling system tasks................................................................................................... 303 17.5.1 Array types................................................................................................................ 303 17.5.2 Array logic types....................................................................................................... 304 17.5.3 Logic array personality declaration and loading....................................................... 304 17.5.4 Logic array personality formats ................................................................................ 304 Copyright © 2001 IEEE. All rights reserved. xv 17.6 Stochastic analysis tasks ........................................................................................................ 307 17.6.1 $q_initialize............................................................................................................... 307 17.6.2 $q_add....................................................................................................................... 308 17.6.3 $q_remove................................................................................................................. 308 17.6.4 $q_full ....................................................................................................................... 308 17.6.5 $q_exam.................................................................................................................... 308 17.6.6 Status codes............................................................................................................... 309 17.7 Simulation time system functions.......................................................................................... 309 17.7.1 $time ......................................................................................................................... 309 17.7.2 $stime........................................................................................................................ 310 17.7.3 $realtime ................................................................................................................... 310 17.8 Conversion functions ............................................................................................................. 311 17.9 Probabilistic distribution functions ........................................................................................ 312 17.9.1 $random function ...................................................................................................... 312 17.9.2 $dist_ functions......................................................................................................... 313 17.9.3 Algorithm for probabilistic distribution functions.................................................... 314 17.10 Command line input............................................................................................................... 321 17.10.1 $test$plusargs (string)............................................................................................... 322 17.10.2 $value$plusargs (user_string, variable) .................................................................... 322 18. Value change dump (VCD) files...................................................................................................... 325 18.1 Creating the four state value change dump file ..................................................................... 325 18.1.1 Specifying the name of the dump file ($dumpfile)................................................... 325 18.1.2 Specifying the variables to be dumped ($dumpvars)................................................ 326 18.1.3 Stopping and resuming the dump ($dumpoff/$dumpon).......................................... 327 18.1.4 Generating a checkpoint ($dumpall)......................................................................... 328 18.1.5 Limiting the size of the dump file ($dumplimit) ...................................................... 328 18.1.6 Reading the dump file during simulation ($dumpflush)........................................... 329 18.2 Format of the four state VCD file .......................................................................................... 330 18.2.1 Syntax of the four state VCD file ............................................................................. 330 18.2.2 Formats of variable values ........................................................................................ 332 18.2.3 Description of keyword commands .......................................................................... 333 18.2.4 Four state VCD file format example......................................................................... 339 18.3 Creating the extended value change dump file ...................................................................... 340 18.3.1 Specifying the dumpfile name and the ports to be dumped ($dumpports) ............... 340 18.3.2 Stopping and resuming the dump ($dumpportsoff/$dumpportson).......................... 341 18.3.3 Generating a checkpoint ($dumpportsall)................................................................. 342 18.3.4 Limiting the size of the dump file ($dumpportslimit) .............................................. 342 18.3.5 Reading the dump file during simulation ($dumpportsflush)................................... 343 18.3.6 Description of keyword commands .......................................................................... 343 18.3.7 General rules for extended VCD system tasks ......................................................... 344 18.4 Format of the extended VCD file........................................................................................... 344 18.4.1 Syntax of the extended VCD file .............................................................................. 344 18.4.2 Extended VCD node information ............................................................................. 346 18.4.3 Value changes ........................................................................................................... 348 18.4.4 Extended VCD file format example ......................................................................... 349 19. Compiler directives.......................................................................................................................... 351 19.1 `celldefine and `endcelldefine................................................................................................ 351 19.2 `default_nettype ..................................................................................................................... 351 xvi Copyright © 2001 IEEE. All rights reserved. 19.3 `define and `undef .................................................................................................................. 352 19.3.1 `define ....................................................................................................................... 352 19.3.2 `undef ........................................................................................................................ 354 19.4 `ifdef, `else, `elsif, `endif, `ifndef .......................................................................................... 354 19.5 `include .................................................................................................................................. 358 19.6 `resetall................................................................................................................................... 358 19.7 `line ........................................................................................................................................ 358 19.8 `timescale ............................................................................................................................... 359 19.9 `unconnected_drive and `nounconnected_drive .................................................................... 361 20. PLI overview.................................................................................................................................... 362 20.1 PLI purpose and history (informative)................................................................................... 362 20.2 User-defined system task or function names ......................................................................... 362 20.3 User-defined system task or function types ........................................................................... 363 20.4 Overriding built-in system task and function names ............................................................. 363 20.5 User-supplied PLI applications.............................................................................................. 363 20.6 PLI interface mechanism ....................................................................................................... 363 20.7 User-defined system task and function arguments ................................................................ 364 20.8 PLI include files..................................................................................................................... 364 20.9 PLI Memory Restrictions....................................................................................................... 364 21. PLI TF and ACC interface mechanism............................................................................................ 365 21.1 User-supplied PLI applications.............................................................................................. 365 21.1.1 The sizetf class of PLI applications .......................................................................... 365 21.1.2 The checktf class of PLI applications ....................................................................... 365 21.1.3 The calltf class of PLI applications........................................................................... 366 21.1.4 The misctf class of PLI applications......................................................................... 366 21.1.5 The consumer class of PLI applications ................................................................... 366 21.2 Associating PLI applications to a class and system task/function name ............................... 366 21.3 PLI application arguments ..................................................................................................... 367 21.3.1 The data C argument................................................................................................. 367 21.3.2 The reason C argument ............................................................................................. 367 21.3.3 The paramvc C argument.......................................................................................... 368 22. Using ACC routines......................................................................................................................... 369 22.1 ACC routine definition .......................................................................................................... 369 22.2 The handle data type .............................................................................................................. 369 22.3 Using ACC routines............................................................................................................... 370 22.3.1 Header files ............................................................................................................... 370 22.3.2 Initializing ACC routines.......................................................................................... 370 22.3.3 Exiting ACC routines................................................................................................ 370 22.4 List of ACC routines by major category................................................................................ 370 22.4.1 Fetch routines............................................................................................................ 371 22.4.2 Handle routines ......................................................................................................... 372 22.4.3 Next routines............................................................................................................. 373 22.4.4 Modify routines......................................................................................................... 375 22.4.5 Miscellaneous routines.............................................................................................. 375 22.4.6 VCL routines............................................................................................................. 376 22.5 Accessible objects.................................................................................................................. 376 22.5.1 ACC routines that operate on module instances ...................................................... 378 22.5.2 ACC routines that operate on module ports ............................................................. 378 Copyright © 2001 IEEE. All rights reserved. xvii 22.5.3 ACC routines that operate on bits of a port ............................................................. 379 22.5.4 ACC routines that operate on module paths or data paths ....................................... 379 22.5.5 ACC routines that operate on intermodule paths ..................................................... 380 22.5.6 ACC routines that operate on top-level modules...................................................... 380 22.5.7 ACC routines that operate on primitive instances .................................................... 380 22.5.8 ACC routines that operate on primitive terminals .................................................... 381 22.5.9 ACC routines that operate on nets ............................................................................ 381 22.5.10 ACC routines that operate on reg types .................................................................... 382 22.5.11 ACC routines that operate on integer, real, and time variables ................................ 382 22.5.12 ACC routines that operate on named events............................................................. 382 22.5.13 ACC routines that operate on parameters and specparams....................................... 383 22.5.14 ACC routines that operate on timing checks ............................................................ 383 22.5.15 ACC routines that operate on timing check terminals .............................................. 383 22.5.16 ACC routines that operate on user-defined system task/function arguments ........... 384 22.6 ACC routine types and fulltypes............................................................................................ 384 22.7 Error handling ........................................................................................................................ 387 22.7.1 Suppressing error messages ...................................................................................... 388 22.7.2 Enabling warnings .................................................................................................... 388 22.7.3 Testing for errors....................................................................................................... 388 22.7.4 Example .................................................................................................................... 388 22.7.5 Exception values ....................................................................................................... 389 22.8 Reading and writing delay values .......................................................................................... 389 22.8.1 Number of delays for Verilog HDL objects ............................................................. 390 22.8.2 ACC routine configuration ....................................................................................... 390 22.8.3 Determining the number of arguments for ACC delay routines............................... 391 22.9 String handling....................................................................................................................... 395 22.9.1 ACC routines share an internal string buffer ............................................................ 395 22.9.2 String buffer reset ..................................................................................................... 396 22.9.3 Preserving string values ............................................................................................ 397 22.9.4 Example of preserving string values......................................................................... 397 22.10 Using VCL ACC routines...................................................................................................... 397 22.10.1 VCL objects .............................................................................................................. 398 22.10.2 The VCL record definition........................................................................................ 398 22.10.3 Effects of acc_initialize() and acc_close() on VCL consumer routines ................... 401 22.10.4 An example of using VCL ACC routines ................................................................. 401 23. ACC routine definitions................................................................................................................... 404 24. Using TF routines ............................................................................................................................ 579 24.1 TF routine definition .............................................................................................................. 579 24.2 TF routine system task/function arguments........................................................................... 579 24.3 Reading and writing system task/function argument values.................................................. 579 24.3.1 Reading and writing 2-state parameter argument values.......................................... 579 24.3.2 Reading and writing 4-state values ........................................................................... 579 24.3.3 Reading and writing strength values......................................................................... 580 24.3.4 Reading and writing to memories ............................................................................. 580 24.3.5 Reading and writing string values............................................................................. 580 24.3.6 Writing return values of user-defined functions ....................................................... 580 24.3.7 Writing the correct C data types ............................................................................... 580 24.4 Value change detection .......................................................................................................... 581 24.5 Simulation time...................................................................................................................... 581 24.6 Simulation synchronization ................................................................................................... 581 24.7 Instances of user-defined tasks or functions .......................................................................... 582 xviii Copyright © 2001 IEEE. All rights reserved. 24.8 Module and scope instance names......................................................................................... 582 24.9 Saving information from one system TF call to the next....................................................... 582 24.10 Displaying output messages................................................................................................... 582 24.11 Stopping and finishing ........................................................................................................... 582 25. TF routine definitions ...................................................................................................................... 583 26. Using VPI routines........................................................................................................................... 659 26.1 VPI system tasks and functions ............................................................................................. 659 26.2 The VPI interface................................................................................................................... 659 26.2.1 VPI callbacks ............................................................................................................ 659 26.2.2 VPI access to Verilog HDL objects and simulation objects ..................................... 660 26.2.3 Error handling ........................................................................................................... 660 26.2.4 Function availability ................................................................................................. 660 26.2.5 Traversing expressions.............................................................................................. 660 26.3 VPI object classifications....................................................................................................... 661 26.3.1 Accessing object relationships and properties .......................................................... 662 26.3.2 Object type properties ............................................................................................... 663 26.3.3 Object file and line properties................................................................................... 663 26.3.4 Delays and values ..................................................................................................... 664 26.4 List of VPI routines by functional category........................................................................... 664 26.5 Key to data model diagrams .................................................................................................. 666 26.5.1 Diagram key for objects and classes ........................................................................ 667 26.5.2 Diagram key for accessing properties....................................................................... 667 26.5.3 Diagram key for traversing relationships ................................................................. 668 26.6 Object data model diagrams................................................................................................... 669 27. VPI routine definitions..................................................................................................................... 700 Annex A (normative) Formal syntax definition........................................................................................... 761 A.1 Source text ............................................................................................................................... 761 A.1.1 Library source text ........................................................................................................ 761 A.1.2 Configuration source text.............................................................................................. 761 A.1.3 Module and primitive source text ................................................................................. 762 A.1.4 Module parameters and ports ........................................................................................ 762 A.1.5 Module items................................................................................................................. 762 A.2 Declarations ............................................................................................................................. 763 A.2.1 Declaration types........................................................................................................... 763 A.2.2 Declaration data types ................................................................................................... 765 A.2.3 Declaration lists............................................................................................................. 765 A.2.4 Declaration assignments ............................................................................................... 766 A.2.5 Declaration ranges......................................................................................................... 766 A.2.6 Function declarations .................................................................................................... 766 A.2.7 Task declarations........................................................................................................... 766 A.2.8 Block item declarations................................................................................................. 767 A.3 Primitive instances ................................................................................................................... 768 A.3.1 Primitive instantiation and instances............................................................................. 768 A.3.2 Primitive strengths ........................................................................................................ 768 A.3.3 Primitive terminals ........................................................................................................ 769 A.3.4 Primitive gate and switch types .................................................................................... 769 Copyright © 2001 IEEE. All rights reserved. xix A.4 Module and generated instantiation ......................................................................................... 769 A.4.1 Module instantiation ..................................................................................................... 769 A.4.2 Generated instantiation ................................................................................................. 769 A.5 UDP declaration and instantiation ........................................................................................... 770 A.5.1 UDP declaration ............................................................................................................ 770 A.5.2 UDP ports...................................................................................................................... 770 A.5.3 UDP body...................................................................................................................... 771 A.5.4 UDP instantiation .......................................................................................................... 771 A.6 Behavioral statements .............................................................................................................. 771 A.6.1 Continuous assignment statements ............................................................................... 771 A.6.2 Procedural blocks and assignments............................................................................... 771 A.6.3 Parallel and sequential blocks ....................................................................................... 772 A.6.4 Statements ..................................................................................................................... 772 A.6.5 Timing control statements............................................................................................. 772 A.6.6 Conditional statements .................................................................................................. 773 A.6.7 Case statements ............................................................................................................. 774 A.6.8 Looping statements ....................................................................................................... 774 A.6.9 Task enable statements.................................................................................................. 774 A.7 Specify section ......................................................................................................................... 774 A.7.1 Specify block declaration.............................................................................................. 774 A.7.2 Specify path declarations .............................................................................................. 775 A.7.3 Specify block terminals................................................................................................. 775 A.7.4 Specify path delays ....................................................................................................... 775 A.7.5 System timing checks.................................................................................................... 777 A.8 Expressions .............................................................................................................................. 779 A.8.1 Concatenations .............................................................................................................. 779 A.8.2 Function calls ................................................................................................................ 779 A.8.3 Expressions ................................................................................................................... 780 A.8.4 Primaries ....................................................................................................................... 781 A.8.5 Expression left-side values............................................................................................ 781 A.8.6 Operators ....................................................................................................................... 782 A.8.7 Numbers ........................................................................................................................ 782 A.8.8 Strings ........................................................................................................................... 783 A.9 General..................................................................................................................................... 783 A.9.1 Attributes....................................................................................................................... 783 A.9.2 Comments ..................................................................................................................... 783 A.9.3 Identifiers ...................................................................................................................... 783 A.9.4 Identifier branches......................................................................................................... 784 A.9.5 White space ................................................................................................................... 785 Annex B (normative) List of keywords ....................................................................................................... 786 Annex C (informative) System tasks and functions .................................................................................... 788 C.1 $countdrivers............................................................................................................................ 788 C.2 $getpattern................................................................................................................................ 789 C.3 $input ....................................................................................................................................... 790 C.4 $key and $nokey ...................................................................................................................... 790 C.5 $list........................................................................................................................................... 791 C.6 $log and $nolog........................................................................................................................ 791 C.7 $reset, $reset_count, and $reset_value..................................................................................... 791 C.8 $save, $restart, and $incsave.................................................................................................... 792 C.9 $scale........................................................................................................................................ 793 C.10$scope ...................................................................................................................................... 793 xx Copyright © 2001 IEEE. All rights reserved. C.11$showscopes ............................................................................................................................ 793 C.12$showvars ................................................................................................................................ 794 C.13$sreadmemb and $sreadmemh................................................................................................. 794 Annex D (informative) Compiler directives ................................................................................................ 795 D.1 `default_decay_time................................................................................................................. 795 D.2 `default_trireg_strength............................................................................................................ 795 D.3 `delay_mode_distributed.......................................................................................................... 796 D.4 `delay_mode_path.................................................................................................................... 796 D.5 `delay_mode_unit .................................................................................................................... 796 D.6 `delay_mode_zero.................................................................................................................... 796 Annex E (normative) acc_user.h.................................................................................................................. 797 Annex F (normative) veriuser.h................................................................................................................... 806 Annex G (normative) vpi_user.h ................................................................................................................. 814 Annex H (informative) Bibliography........................................................................................................... 828 Copyright © 2001 IEEE. All rights reserved. xxi IEEE Standard Verilog® Hardware Description Language 1. Overview 1.1 Objectives of this standard The intent of this standard is to serve as a complete specification of the Verilog® Hardware Description Language (HDL). This document contains — The formal syntax and semantics of all Verilog HDL constructs — The formal syntax and semantics of Standard Delay Format (SDF) constructs — Simulation system tasks and functions, such as text output display commands — Compiler directives, such as text substitution macros and simulation time scaling — The Programming Language Interface (PLI) binding mechanism — The formal syntax and semantics of access routines, task/function routines, and Verilog procedural interface routines — Informative usage examples — Informative delay model for SDF — Listings of header files for PLI 1.2 Conventions used in this standard This standard is organized into clauses, each of which focuses on a specific area of the language. There are subclauses within each clause to discuss individual constructs and concepts. The discussion begins with an introduction and an optional rationale for the construct or the concept, followed by syntax and semantic descriptions, followed by some examples and notes. The term shall is used throughout this standard to indicate mandatory requirements, whereas the term can is used to indicate optional features. These terms denote different meanings to different readers of this standard: a) To the developers of tools that process the Verilog HDL, the term shall denotes a requirement that the standard imposes. The resulting implementation is required to enforce the requirements and to issue an error if the requirement is not met by the input. b) To the Verilog HDL model developer, the term shall denotes that the characteristics of the Verilog HDL are natural consequences of the language definition. The model developer is required to adhere to the constraint implied by the characteristic. The term can denotes optional features that the model Copyright © 2001 IEEE. All rights reserved. 1 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® developer can exercise at discretion. If used, however, the model developer is required to follow the requirements set forth by the language definition. c) To the Verilog HDL model user, the term shall denotes that the characteristics of the models are natural consequences of the language definition. The model user can depend on the characteristics of the model implied by its Verilog HDL source text. 1.3 Syntactic description The formal syntax of the Verilog HDL is described using Backus-Naur Form (BNF). The following conventions are used: a) Lowercase words, some containing embedded underscores, are used to denote syntactic categories. For example: module_declaration b) Boldface words are used to denote reserved keywords, operators, and punctuation marks as a required part of the syntax. These words appear in a larger font for distinction. For example: module => ; c) A vertical bar separates alternative items unless it appears in boldface, in which case it stands for itself. For example: unary_operator ::= + | - | ! | ~ | & | ~& | | | ~| | ^ | ~^ | ^~ d) Square brackets enclose optional items. For example: input_declaration ::= input [range] list_of_variables ; e) Braces enclose a repeated item unless it appears in boldface, in which case it stands for itself. The item may appear zero or more times; the repetitions occur from left to right as with an equivalent left-recursive rule. Thus, the following two rules are equivalent: list_of_param_assignments ::= param_assignment { , param_assignment } list_of_param_assignments ::= param_assignment | list_of_param_assignment , param_assignment f) If the name of any category starts with an italicized part, it is equivalent to the category name without the italicized part. The italicized part is intended to convey some semantic information. For example, msb_constant_expression and lsb_constant_expression are equivalent to constant_expression. The main text uses italicized font when a term is being defined, and constant-width font for examples, file names, and while referring to constants, especially 0, 1, x, and z values. 1.4 Contents of this standard A synopsis of the clauses and annexes is presented as a quick reference. There are 27 clauses and 8 annexes. All clauses, as well as Annex A, Annex B, Annex E, Annex F, and Annex G, are normative parts of this standard. Annex C, Annex D, and Annex H are included for informative purposes only. Clause 1—Overview: This clause discusses the conventions used in this standard and its contents. 2 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Clause 2—Lexical conventions: This clause describes the lexical tokens used in Verilog HDL source text and their conventions. It describes how to specify and interpret the lexical tokens. Clause 3—Data types: This clause describes net and variable data types. This clause also discusses the parameter data type for constant values and describes drive and charge strength of the values on nets. Clause 4—Expressions: This clause describes the operators and operands that can be used in expressions. Clause 5—Scheduling semantics: This clause describes the scheduling semantics of the Verilog HDL. Clause 6—Assignments: This clause compares the two main types of assignment statements in the Verilog HDL—continuous assignments and procedural assignments. It describes the continuous assignment statement that drives values onto nets. Clause 7—Gate and switch level modeling: This clause describes the gate and switch level primitives and logic strength modeling. Clause 8—User-defined primitives (UDPs): This clause describes how a primitive can be defined in the Verilog HDL and how these primitives are included in Verilog HDL models. Clause 9—Behavioral modeling: This clause describes procedural assignments, procedural continuous assignments, and behavioral language statements. Clause 10—Tasks and functions: This clause describes tasks and functions—procedures that can be called from more than one place in a behavioral model. It describes how tasks can be used like subroutines and how functions can be used to define new operators. Clause 11—Disabling of named blocks and tasks: This clause describes how to disable the execution of a task and a block of statements that has a specified name. Clause 12—Hierarchical structures: This clause describes how hierarchies are created in the Verilog HDL and how parameter values declared in a module can be overridden. It describes how generated instantiations can be used to do conditional or multiple instantiations in a design. Clause 13—Configuring the contents of a design: This clause describes how to configure the contents of a design. Clause 14—Specify blocks: This clause describes how to specify timing relationships between input and output ports of a module. Clause 15—Timing checks: This clause describes how timing checks are used in specify blocks to determine if signals obey the timing constraints. Clause 16—Backannotation using the Standard Delay Format (SDF): This clause describes syntax and semantics of Standard Delay Format (SDF) constructs. Clause 17—System tasks and functions: This clause describes the system tasks and functions. Clause 18—Value change dump (VCD) files: This clause describes the system tasks associated with Value Change Dump (VCD) file, and the format of the file. Clause 19—Compiler directives: This clause describes the compiler directives. Copyright © 2001 IEEE. All rights reserved. 3 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Clause 20—PLI overview: This clause previews the C language procedural interface standard (Programming Language Interface or PLI) and interface mechanisms that are part of the Verilog HDL. Clause 21—PLI TF and ACC interface mechanism: This clause describes the interface mechanism that provides a means for users to link PLI task/function (TF) routine and access (ACC) routine applications to Verilog software tools. Clause 22—Using ACC routines: This clause describes the ACC routines in general, including how and why to use them. Clause 23—ACC routine definitions: This clause describes the specific ACC routines, explaining their function, syntax, and usage. Clause 24—Using TF routines: This clause provides an overview of the types of operations that are done with the TF routines. Clause 25—TF routine definitions: This clause describes the specific TF routines, explaining their function, syntax, and usage. Clause 26—Using VPI routines: This clause provides an overview of the types of operations that are done with the Verilog Programming Interface (VPI) routines. Clause 27—VPI routine definitions: This clause describes the VPI routines. Annex A—Formal syntax definition: This normative annex describes, using BNF, the syntax of the Verilog HDL. Annex B—List of keywords: This normative annex lists the Verilog HDL keywords. Annex C—System tasks and functions: This informative annex describes system tasks and functions that are frequently used, but that are not part of the standard. Annex D—Compiler directives: This informative annex describes compiler directives that are frequently used, but that are not part of the standard. Annex E—acc_user.h: This normative annex provides a listing of the contents of the acc_user.h file. Annex F—veriuser.h: This normative annex provides a listing of the contents of the veriuser.h file. Annex G—vpi_user.h: This normative annex provides a listing of the contents of the vpi_user.h file. Annex H—Bibliography: This informative annex contains bibliographic entries pertaining to this standard. 1.5 Header file listings The header file listings included in the annexes E, F, and G for acc_user.h, veriuser.h, and vpi_user.h are a normative part of this standard. All compliant software tools should use the same function declarations, constant definitions, and structure definitions contained in these header file listings. 4 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 1.6 Examples Several small examples in the Verilog HDL and the C programming language are shown throughout this standard. These examples are informative—they are intended to illustrate the usage of Verilog HDL constructs and PLI functions in a simple context and do not define the full syntax. 1.7 Prerequisites Clauses 20 through 27 and Annexes E through G presuppose a working knowledge of the C programming language. Copyright © 2001 IEEE. All rights reserved. 5 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 2. Lexical conventions This clause describes the lexical tokens used in Verilog HDL source text and their conventions. 2.1 Lexical tokens Verilog HDL source text files shall be a stream of lexical tokens. A lexical token shall consist of one or more characters. The layout of tokens in a source file shall be free format—that is, spaces and newlines shall not be syntactically significant other than being token separators, except for escaped identifiers (see 2.7.1). The types of lexical tokens in the language are as follows: — White space — Comment — Operator — Number — String — Identifier — Keyword 2.2 White space White space shall contain the characters for spaces, tabs, newlines, and formfeeds. These characters shall be ignored except when they serve to separate other lexical tokens. However, blanks and tabs shall be considered significant characters in strings (see 2.6). 2.3 Comments The Verilog HDL has two forms to introduce comments. A one-line comment shall start with the two characters // and end with a new line. A block comment shall start with /* and end with */. Block comments shall not be nested. The one-line comment token // shall not have any special meaning in a block comment. 2.4 Operators Operators are single-, double-, or triple-character sequences and are used in expressions. Clause 4 discusses the use of operators in expressions. Unary operators shall appear to the left of their operand. Binary operators shall appear between their operands. A conditional operator shall have two operator characters that separate three operands. 2.5 Numbers Constant numbers can be specified as integer constants (defined in 2.5.1) or real constants. 6 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C number ::= (From Annex A - A.8.7) decimal_number | octal_number | binary_number | hex_number | real_number real_numbera ::= unsigned_number . unsigned_number | unsigned_number [ . unsigned_number ] exp [ sign ] unsigned_number exp ::= e | E decimal_number ::= unsigned_number | [ size ] decimal_base unsigned_number | [ size ] decimal_base x_digit { _ } | [ size ] decimal_base z_digit { _ } binary_number ::= [ size ] binary_base binary_value octal_number ::= [ size ] octal_base octal_value hex_number ::= [ size ] hex_base hex_value sign ::= + | size ::= non_zero_unsigned_number non_zero_unsigned_numbera ::= non_zero_decimal_digit { _ | decimal_digit} unsigned_numbera ::= decimal_digit { _ | decimal_digit } binary_valuea ::= binary_digit { _ | binary_digit } octal_valuea ::= octal_digit { _ | octal_digit } hex_valuea ::= hex_digit { _ | hex_digit } decimal_basea ::= '[s|S]d | '[s|S]D binary_basea ::= '[s|S]b | '[s|S]B octal_basea::= '[s|S]o | '[s|S]O hex_basea ::= '[s|S]h | '[s|S]H non_zero_decimal_digit ::= 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 decimal_digit ::= 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 binary_digit ::= x_digit | z_digit | 0 | 1 octal_digit ::= x_digit | z_digit | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 hex_digit ::= x_digit | z_digit | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 |a|b|c|d|e|f|A|B|C|D|E|F x_digit ::= x | X z_digit ::= z | Z | ? aEmbedded spaces are illegal. Syntax 2-1—Syntax for integer and real numbers 2.5.1 Integer constants Integer constants can be specified in decimal, hexadecimal, octal, or binary format. Copyright © 2001 IEEE. All rights reserved. 7 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® There are two forms to express integer constants. The first form is a simple decimal number, which shall be specified as a sequence of digits 0 through 9, optionally starting with a plus or minus unary operator. The second form specifies a size constant, which shall be composed of up to three tokens—an optional size constant, a single quote followed by a base format character, and the digits representing the value of the number. The first token, a size constant, shall specify the size of the constant in terms of its exact number of bits. It shall be specified as a non-zero unsigned decimal number. For example, the size specification for two hexadecimal digits is 8, because one hexadecimal digit requires 4 bits. Unsized unsigned constants where the high order bit is unknown (X or x) or three-state (Z or z) are extended to the size of the expression containing the constant. NOTE—In IEEE Std 1364-1995, unsized constants where the high order bit is unknown or three-state, the x or z was only extended to 32 bits. The second token, a base_format, shall consist of a case-insensitive letter specifying the base for the number, optionally preceded by the single character s (or S) to indicate a signed quantity, preceded by the single quote character (’). Legal base specifications are d, D, h, H, o, O, b, or B, for the bases decimal, hexadecimal, octal, and binary respectively. The use of x and z in defining the value of a number is case insensitive. The single quote and the base format character shall not be separated by any white space. The third token, an unsigned number, shall consist of digits that are legal for the specified base format. The unsigned number token shall immediately follow the base format, optionally preceded by white space. The hexadecimal digits a to f shall be case insensitive. Simple decimal numbers without the size and the base format shall be treated as signed integers, whereas the numbers specified with the base format shall be treated as signed integers if the s designator is included or as unsigned integers if the base format only is used. The s designator does not affect the bit pattern specified, only its interpretation. A plus or minus operator preceding the size constant is a unary plus or minus operator. A plus or minus operator between the base format and the number is an illegal syntax. Negative numbers shall be represented in 2’s complement form. An x represents the unknown value in hexadecimal, octal, and binary constants. A z represents the highimpedance value. See 3.1 for a discussion of the Verilog HDL value set. An x shall set 4 bits to unknown in the hexadecimal base, 3 bits in the octal base, and 1 bit in the binary base. Similarly, a z shall set 4 bits, 3 bits, and 1 bit, respectively, to the high-impedance value. If the size of the unsigned number is smaller than the size specified for the constant, the unsigned number shall be padded to the left with zeros. If the leftmost bit in the unsigned number is an x or a z, then an x or a z shall be used to pad to the left respectively. When used in a number, the question-mark (?) character is a Verilog HDL alternative for the z character. It sets 4 bits to the high-impedance value in hexadecimal numbers, 3 bits in octal, and 1 bit in binary. The question mark can be used to enhance readability in cases where the high-impedance value is a don’t-care condition. See the discussion of casez and casex in 9.5.1. The question-mark character is also used in user-defined primitive state tables. See 8.1.6, Table 40. The underscore character (_) shall be legal anywhere in a number except as the first character. The underscore character is ignored. This feature can be used to break up long numbers for readability purposes. 8 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Examples: Example 1—Unsized constant numbers 659 // is a decimal number ’h 837FF // is a hexadecimal number ’o7460 // is an octal number 4af // is illegal (hexadecimal format requires ’h) Example 2—Sized constant numbers 4’b1001 5 ’D 3 3’b01x 12’hx 16’hz // is a 4-bit binary number // is a 5-bit decimal number // is a 3-bit number with the least // significant bit unknown // is a 12-bit unknown number // is a 16-bit high-impedance number Example 3—Using sign with constant numbers 8 ’d -6 -8 ’d 6 4 ’shf -4 ’sd15 // this is illegal syntax // this defines the two’s complement of 6, // held in 8 bits—equivalent to -(8’d 6) // this denotes the 4-bit number ‘1111’, to // be interpreted as a 2’s complement number, // or ‘-1’. This is equivalent to -4’h 1 // this is equivalent to -(-4’d 1), or ‘0001’. Example 4—Automatic left padding reg [11:0] a, b, c, d; initial begin a = ’h x; // yields xxx b = ’h 3x; // yields 03x c = ’h z3; // yields zz3 d = ’h 0z3; // yields 0z3 end reg [84:0] e, f, g; e = 'h5; f = 'hx; g = 'hz; // yields {82{1'b0},3'b101} // yields {85{1'hx}} // yields {85{1'hz}} Example 5—Using underscore character in numbers 2277__119955__000000 1166’’bb00001111__00110011__00000011__11111111 3322 ’’hh 1122aabb__ff000011 Copyright © 2001 IEEE. All rights reserved. 9 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® NOTES: 1) Sized negative constant numbers and sized signed constant numbers are sign-extended when assigned to a reg data type, regardless of whether the reg itself is signed or not. 2) Each of the three tokens for specifying a number may be macro substituted. 3) The number of bits that make up an unsized number (which is a simple decimal number or a number without the size specification) shall be at least 32. 2.5.2 Real constants The real constant numbers shall be represented as described by IEEE Std 754-1985 [B1],1 an IEEE standard for double-precision floating-point numbers. Real numbers can be specified in either decimal notation (for example, 14.72) or in scientific notation (for example, 39e8, which indicates 39 multiplied by 10 to the eighth power). Real numbers expressed with a decimal point shall have at least one digit on each side of the decimal point. Examples: 1.2 0.1 2394.26331 1.2E12 (the exponent symbol can be e or E) 1.30e-2 0.1e-0 23E10 29E-2 236.123_763_e-12 (underscores are ignored) The following are invalid forms of real numbers because they do not have at least one digit on each side of the decimal point: .12 9. 4.E3 .2e-7 2.5.3 Conversion Real numbers shall be converted to integers by rounding the real number to the nearest integer, rather than by truncating it. Implicit conversion shall take place when a real number is assigned to an integer. The ties shall be rounded away from zero. For example: — The real numbers 35.7 and 35.5 both become 36 when converted to an integer and 35.2 becomes 35. — Converting -1.5 to integer yields -2, converting 1.5 to integer yields 2. 2.6 Strings A string is a sequence of characters enclosed by double quotes ("") and contained on a single line. Strings used as operands in expressions and assignments shall be treated as unsigned integer constants represented by a sequence of 8-bit ASCII values, with one 8-bit ASCII value representing one character. 1The numbers in brackets correspond to those of the bibliography in Annex H. 10 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 2.6.1 String variable declaration String variables are variables of reg type (see 3.2) with width equal to the number of characters in the string multiplied by 8. Example: To store the twelve-character string "Hello world!" requires a reg 8 * 12, or 96 bits wide reg [8*12:1] stringvar; initial begin stringvar = "Hello world!"; end 2.6.2 String manipulation Strings can be manipulated using the Verilog HDL operators. The value being manipulated by the operator is the sequence of 8-bit ASCII values. Example: module string_test; reg [8*14:1] stringvar; initial begin stringvar = "Hello world"; $display("%s is stored as %h", stringvar,stringvar); stringvar = {stringvar,"!!!"}; $display("%s is stored as %h", stringvar,stringvar); end endmodule The output is: Hello world is stored as 00000048656c6c6f20776f726c64 Hello world!!! is stored as 48656c6c6f20776f726c64212121 NOTE—When a variable is larger than required to hold a value being assigned, the contents on the left are padded with zeros after the assignment. This is consistent with the padding that occurs during assignment of nonstring values. If a string is larger than the destination string variable, the string is truncated to the left, and the leftmost characters will be lost. 2.6.3 Special characters in strings Certain characters can only be used in strings when preceded by an introductory character called an escape character. Table 1 lists these characters in the right-hand column, with the escape sequence that represents the character in the left-hand column. Copyright © 2001 IEEE. All rights reserved. 11 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Table 1—Specifying special characters in string Escape string Character produced by escape string \n New line character \t Tab character \\ \ character \" " character \ddd A character specified in 1–3 octal digits (0 ≤ d ≤ 7) 2.7 Identifiers, keywords, and system names An identifier is used to give an object a unique name so it can be referenced. An identifier is either a simple identifier or an escaped identifier (see 2.7.1). A simple identifier shall be any sequence of letters, digits, dollar signs ($), and underscore characters (_). The first character of a simple identifier shall not be a digit or $; it can be a letter or an underscore. Identifiers shall be case sensitive. Example: shiftreg_a busa_index error_condition merge_ab _bus3 n$657 NOTE—Implementations may set a limit on the maximum length of identifiers, but they shall at least be 1024 characters. If an identifier exceeds the implementation-specified length limit, an error shall be reported. 2.7.1 Escaped identifiers Escaped identifiers shall start with the backslash character (\) and end with white space (space, tab, newline). They provide a means of including any of the printable ASCII characters in an identifier (the decimal values 33 through 126, or 21 through 7E in hexadecimal). Neither the leading backslash character nor the terminating white space is considered to be part of the identifier. Therefore, an escaped identifier \cpu3 is treated the same as a nonescaped identifier cpu3. Example: \busa+index \-clock \***error-condition*** \net1/\net2 \{a,b} \a*(b+c) 12 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 2.7.2 Generated identifiers Generated identifiers are created by generate loops (see 12.1.3.2); and are a special case of identifiers in that they can be used in hierarchical names (see 12.4). A generated identifier is the named generate block identifier terminated with a ([digit(s)]) string. This identifier is used as a node name in hierarchical names (see 12.4). 2.7.3 Keywords Keywords are predefined nonescaped identifiers that are used to define the language constructs. A Verilog HDL keyword preceded by an escape character is not interpreted as a keyword. All keywords are defined in lowercase only. Annex B gives a list of all defined keywords. 2.7.4 System tasks and functions The $ character introduces a language construct that enables development of user-defined tasks and functions. System constructs are not design semantics, but refer to simulator functionality. A name following the $ is interpreted as a system task or a system function. The syntax for a system task or function is given in Syntax 2-2. system_task_enable ::= (From Annex A - A.6.9) system_task_identifier [ ( expression { , expression } ) ] ; system_function_call ::= (From Annex A - A.8.2) system_function_identifier [ ( expression { , expression } ) ] system_function_identifiera ::= (From Annex A - A.9.3) $[ a-zA-Z0-9_$ ]{ [ a-zA-Z0-9_$ ] } system_task_identifiera ::= $[ a-zA-Z0-9_$ ]{ [ a-zA-Z0-9_$ ] } aThe $ character in a system_function_identifier or system_task_identifier shall not be followed by white space. A system_function_identifier or system_task_identifier shall not be escaped. Syntax 2-2—Syntax for system tasks and functions The $identifier system task or function can be defined in three places — A standard set of $identifier system tasks and functions, as defined in Clauses 17 and 18. — Additional $identifier system tasks and functions defined using the PLI, as described in Clause 20. — Additional $identifier system tasks and functions defined by software implementations. Any valid identifier, including keywords already in use in contexts other than this construct, can be used as a system task or function name. The system tasks and functions described in Clause 17 and Clause 18 are part of this standard. Additional system tasks and functions with the $identifier construct are not part of this standard. Example: $display ("display a message"); $finish; Copyright © 2001 IEEE. All rights reserved. 13 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 2.7.5 Compiler directives The ` character (the ASCII value 60, called open quote or accent grave) introduces a language construct used to implement compiler directives. The compiler behavior dictated by a compiler directive shall take effect as soon as the compiler reads the directive. The directive shall remain in effect for the rest of the compilation unless a different compiler directive specifies otherwise. A compiler directive in one description file can therefore control compilation behavior in multiple description files. The `identifier compiler directive construct can be defined in two places — A standard set of `identifier compiler directives defined in Clause 19. — Additional `identifier compiler directives defined by software implementations. Any valid identifier, including keywords already in use in contexts other than this construct, can be used as a compiler directive name. The compiler directives described in Clause 19 are part of this standard. Additional compiler directives with the `identifier construct are not part of this standard. Example: `define wordsize 8 2.8 Attributes With the proliferation of tools other than simulators that use Verilog HDL as their source, a mechanism is included for specifying properties about objects, statements and groups of statements in the HDL source that may be used by various tools, including simulators, to control the operation or behavior of the tool. These properties shall be referred to as "attributes". This subclause specifies the syntactic mechanism that shall be used for specifying attributes, without standardizing on any particular attributes. The syntax for specifying an attribute is shown in Syntax 2-3. attribute_instance ::= (From Annex A - A.9.1) (* attr_spec { , attr_spec } *) attr_spec ::= attr_name = constant_expression | attr_name attr_name ::= identifier Syntax 2-3—Syntax for attributes An attribute_instance can appear in the Verilog description as a prefix attached to a declaration, a module item, a statement, or a port connection. It can appear as a suffix to an operator or a Verilog function name in an expression. If a value is not specifically assigned to the attribute, then its value shall be 1. If the same attribute name is defined more than once for the same language element, the last attribute value shall be used and a tool can give a warning that a duplicate attribute specification has occurred. 14 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 2.8.1 Examples Example 1—The following example shows how to attach attributes to a case statement: (* full_case, parallel_case *) case (foo) or (* full_case=1, parallel_case=1 *) case (foo) or (* full_case, // no value assigned parallel_case=1 *) case (foo) Example 2—To attach the full_case attribute, but NOT the parallel_case attribute: (* full_case *) // parallel_case not specified case (foo) or (* full_case=1, parallel_case = 0 *) case (foo) Example 3—To attach an attribute to a module definition: (* optimize_power *) module mod1 (); or (* optimize_power=1 *) module mod1 (); Example 4—To attach an attribute to a module instantiation: (* optimize_power=0 *) mod1 synth1 (); Example 5—To attach an attribute to a reg declaration: (* fsm_state *) reg [7:0] state1; (* fsm_state=1 *) reg [3:0] state2, state3; reg [3:0] reg1; // this reg does NOT have fsm_state set (* fsm_state=0 *) reg [3:0] reg2; // nor does this one Copyright © 2001 IEEE. All rights reserved. 15 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Example 6—To attach an attribute to an operator: a = b + (* mode = "cla" *) c; This sets the value for the attribute mode to be the string cla. Example 7—To attach an attribute to a Verilog function call: a = add (* mode = "cla" *) (b, c); Example 8—To attach an attribute to a conditional operator: a = b ? (* no_glitch *) c : d; 2.8.2 Syntax The syntax for legal statements with attributes is shown in Syntax 2-4 — Syntax 2-9. The syntax for module declaration attributes is given in Syntax 2-4. module_declaration ::= (From Annex A - A.1.3) { attribute_instance } module_keyword module_identifier [ module_parameter_port_list ] [ list_of_ports ] ; { module_item } endmodule | { attribute_instance } module_keyword module_identifier [ module_parameter_port_list ] [ list_of_port_declarations ] ; { non_port_module_item } endmodule Syntax 2-4—Syntax for module declaration attributes The syntax for port declaration attributes is given in Syntax 2-5. port_declaration ::= (From Annex A - A.1.4) {attribute_instance} inout_declaration | {attribute_instance} input_declaration | {attribute_instance} output_declaration Syntax 2-5—Syntax for port declaration attributes 16 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The syntax for module item attributes is given in Syntax 2-6. module_item ::= (From Annex A - A.1.5) module_or_generate_item | port_declaration ; | { attribute_instance } generated_instantiation | { attribute_instance } local_parameter_declaration | { attribute_instance } parameter_declaration | { attribute_instance } specify_block | { attribute_instance } specparam_declaration module_or_generate_item ::= { attribute_instance } module_or_generate_item_declaration | { attribute_instance } parameter_override | { attribute_instance } continuous_assign | { attribute_instance } gate_instantiation | { attribute_instance } udp_instantiation | { attribute_instance } module_instantiation | { attribute_instance } initial_construct | { attribute_instance } always_construct non_port_module_item ::= { attribute_instance } generated_instantiation | { attribute_instance } local_parameter_declaration | { attribute_instance } module_or_generate_item | { attribute_instance } parameter_declaration | { attribute_instance } specify_block | { attribute_instance } specparam_declaration Syntax 2-6—Syntax for module item attributes Copyright © 2001 IEEE. All rights reserved. 17 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® The syntax for function port, task, and block attributes is given in Syntax 2-7. function_port_list ::= (From Annex A - A.2.6) {attribute_instance} input_declaration { , {attribute_instance } input_declaration} task_item_declaration ::= (From Annex A - A.2.7) block_item_declaration | { attribute_instance } input_declaration ; | { attribute_instance } output_declaration ; | { attribute_instance } inout_declaration ; task_port_item ::= { attribute_instance } input_declaration | { attribute_instance } output_declaration | { attribute_instance } inout_declaration block_item_declaration ::= (From Annex A - A.2.8) { attribute_instance } block_reg_declaration | { attribute_instance } event_declaration | { attribute_instance } integer_declaration | { attribute_instance } local_parameter_declaration | { attribute_instance } parameter_declaration | { attribute_instance } real_declaration | { attribute_instance } realtime_declaration | { attribute_instance } time_declaration Syntax 2-7—Syntax for function port, task, and block attributes The syntax for port connection attributes is given in Syntax 2-8. ordered_port_connection ::= (From Annex A - A.4.1) { attribute_instance } [ expression ] named_port_connection ::= { attribute_instance } .port_identifier ( [ expression ] ) Syntax 2-8—Syntax for port connection attributes 18 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The syntax for udp attributes is given in Syntax 2-9. udp_declaration ::= (From Annex A - A.5.1) { attribute_instance } primitive udp_identifier ( udp_port_list ) ; udp_port_declaration { udp_port_declaration } udp_body endprimitive | { attribute_instance } primitive udp_identifier ( udp_declaration_port_list ) ; udp_body endprimitive udp_output_declaration ::= (From Annex A - A.5.2) { attribute_instance } output port_identifier | { attribute_instance } output reg port_identifier [ = constant_expression ] udp_input_declaration ::= { attribute_instance } input list_of_port_identifiers udp_reg_declaration ::= { attribute_instance } reg variable_identifier Syntax 2-9—Syntax for udp attributes Copyright © 2001 IEEE. All rights reserved. 19 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 3. Data types The set of Verilog HDL data types is designed to represent the data storage and transmission elements found in digital hardware. 3.1 Value set The Verilog HDL value set consists of four basic values: 0 - represents a logic zero, or a false condition 1 - represents a logic one, or a true condition x - represents an unknown logic value z - represents a high-impedance state The values 0 and 1 are logical complements of one another. When the z value is present at the input of a gate, or when it is encountered in an expression, the effect is usually the same as an x value. Notable exceptions are the metal-oxide semiconductor (MOS) primitives, which can pass the z value. Almost all of the data types in the Verilog HDL store all four basic values. The exception is the event type (see 9.7.3), which has no storage. All bits of vectors can be independently set to one of the four basic values. The language includes strength information in addition to the basic value information for net variables. This is described in detail in Clause 7. 3.2 Nets and variables There are two main groups of data types: the variable data types and the net data types. These two groups differ in the way that they are assigned and hold values. They also represent different hardware structures. 3.2.1 Net declarations The net data types shall represent physical connections between structural entities, such as gates. A net shall not store a value (except for the trireg net). Instead, its value shall be determined by the values of its drivers, such as a continuous assignment or a gate. See Clause 6 and Clause 7 for definitions of these constructs. If no driver is connected to a net, its value shall be high-impedance (z) unless the net is a trireg, in which case it shall hold the previously driven value. It is illegal to redeclare a name already declared by a net, parameter, or variable declaration (see 3.12). 20 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The syntax for net declarations is given in Syntax 3-1. net_declaration ::= (From Annex A - A.2.1.3) net_type [ signed ] [ delay3 ] list_of_net_identifiers ; | net_type [ drive_strength ] [ signed ] [ delay3 ] list_of_net_decl_assignments ; | net_type [ vectored | scalared ] [ signed ] range [ delay3 ] list_of_net_identifiers ; | net_type [ drive_strength ] [ vectored | scalared ] [ signed ] range [ delay3 ] list_of_net_decl_assignments ; | trireg [ charge_strength ] [ signed ] [ delay3 ] list_of_net_identifiers ; | trireg [ drive_strength ] [ signed ] [ delay3 ] list_of_net_decl_assignments ; | trireg [ charge_strength ] [ vectored | scalared ] [ signed ] range [ delay3 ] list_of_net_identifiers ; | trireg [ drive_strength ] [ vectored | scalared ] [ signed ] range [ delay3 ] list_of_net_decl_assignments ; net_type ::= (From Annex A - A.2.2.1) supply0 | supply1 | tri | triand | trior | tri0 | tri1 | wire | wand | wor drive_strength ::= (From Annex A - A.2.2.2) ( strength0 , strength1 ) | ( strength1 , strength0 ) | ( strength0 , highz1 ) | ( strength1 , highz0 ) | ( highz0 , strength1 ) | ( highz1 , strength0 ) strength0 ::= supply0 | strong0 | pull0 | weak0 strength1 ::= supply1 | strong1 | pull1 | weak1 charge_strength ::= ( small ) | ( medium ) | ( large ) delay3 ::= (From Annex A - A.2.2.3) # delay_value | # ( delay_value [ , delay_value [ , delay_value ] ] ) delay2 ::= # delay_value | # ( delay_value [ , delay_value ] ) delay_value ::= unsigned_number | parameter_identifier | specparam_identifier | mintypmax_expression list_of_net_decl_assignments ::= (From Annex A - A.2.3) net_decl_assignment { , net_decl_assignment } list_of_net_identifiers ::= net_identifier [ dimension { dimension }] { , net_identifier [ dimension { dimension }] } net_decl_assignment ::= (From Annex A - A.2.4) net_identifier = expression dimension ::= (From Annex A -A.2.5) [ dimension_constant_expression : dimension_constant_expression ] range ::= [ msb_constant_expression : lsb_constant_expression ] Syntax 3-1—Syntax for net declaration Copyright © 2001 IEEE. All rights reserved. 21 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® The first two forms of net declaration are described in this section. The third form, called net assignment, is described in Clause 6. 3.2.2 Variable declarations A variable is an abstraction of a data storage element. A variable shall store a value from one assignment to the next. An assignment statement in a procedure acts as a trigger that changes the value in the data storage element. The initialization value for reg, time, and integer data types shall be the unknown value, x. The default initialization value for real and realtime variable datatypes shall be 0.0. If a variable declaration assignment is used (see 6.2.1), the variable shall take this value as if the assignment occurred in a blocking assignment in an initial construct. It is illegal to redeclare a name already declared by a net, parameter, or variable declaration. NOTE—In previous versions of the Verilog standard, the term register was used to encompass both the reg, integer, time, real and realtime types, but that term is no longer used as a Verilog data type. The syntax for variable declarations is given in Syntax 3-2. integer_declaration ::= (From Annex A - A.2.1.3) integer list_of_variable_identifiers ; real_declaration ::= real list_of_real_identifiers ; realtime_declaration ::= realtime list_of_real_identifiers ; reg_declaration ::= reg [ signed ] [ range ] list_of_variable_identifiers ; time_declaration ::= time list_of_variable_identifiers ; real_type ::= (From Annex A - A.2.2.1) real_identifier [ = constant_expression ] | real_identifier dimension { dimension } variable_type ::= variable_identifier [ = constant_expression ] | variable_identifier dimension { dimension } list_of_real_identifiers ::= (From Annex A - A.2.3) real_type { , real_type } list_of_variable_identifiers ::= variable_type { , variable_type } dimension ::= (From Annex A - A.2.5) [ dimension_constant_expression : dimension_constant_expression ] range ::= [ msb_constant_expression : lsb_constant_expression ] Syntax 3-2—Syntax for variable declaration If a set of nets or variables share the same characteristics, they can be declared in the same declaration statement. 22 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C CAUTION Variables can be assigned negative values, but only signed regs, integer, real, and realtime variables shall retain the significance of the sign. The unsigned reg and time variables shall treat the value assigned to them as an unsigned value. Refer to 4.1.6 for a description of how signed and unsigned variables are treated by certain Verilog operators. 3.3 Vectors A net or reg declaration without a range specification shall be considered 1 bit wide and is known as a scalar. Multiple bit net and reg data types shall be declared by specifying a range, which is known as a vector. 3.3.1 Specifying vectors The range specification gives addresses to the individual bits in a multibit net or reg. The most significant bit specified by the msb constant expression is the left-hand value in the range and the least significant bit specified by the lsb constant expression is the right-hand value in the range. Both msb constant expression and lsb constant expression shall be constant expressions. The msb and lsb constant expressions can be any value—positive, negative, or zero. The lsb constant expression can be a greater, equal, or lesser value than msb constant expression. Vector nets and regs shall obey laws of arithmetic modulo 2 to the power n (2n), where n is the number of bits in the vector. Vector nets and regs shall be treated as unsigned quantities, unless the net or reg is declared to be signed or is connected to a port that is declared to be signed (see 12.2.3). Examples: wand w; // a scalar net of type “wand” tri [15:0] busa; // a three-state 16-bit bus trireg (small) storeit; // a charge storage node of strength small reg a; // a scalar reg reg[3:0] v; // a 4-bit vector reg made up of (from most to // least significant) v[3], v[2], v[1], and v[0] reg signed [3:0] signed_reg; // a 4-bit vector in range -8 to 7 reg [-1:4] b; // a 6-bit vector reg wire w1, w2; // declares two wires reg [4:0] x, y, z; // declares three 5-bit regs NOTES: 1) Implementations may set a limit on the maximum length of a vector, but they will at least be 65536 (216) bits. 2) Implementations do not have to detect overflow of integer operations. Copyright © 2001 IEEE. All rights reserved. 23 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 3.3.2 Vector net accessibility Vectored and scalared shall be optional advisory keywords to be used in vector net or reg declaration. If these keywords are implemented, certain operations on vectors may be restricted. If the keyword vectored is used, bit and part selects and strength specifications may not be permitted, and the PLI may consider the object unexpanded. If the keyword scalared is used, bit and part selects of the object shall be permitted, and the PLI shall consider the object expanded. Examples: tri1 scalared [63:0] bus64; //a bus that will be expanded tri vectored [31:0] data; //a bus that may or may not be expanded 3.4 Strengths There are two types of strengths that can be specified in a net declaration. They are as follows: charge strength shall only be used when declaring a net of type trireg drive strength shall only be used when placing a continuous assignment on a net in the same statement that declares the net Gate declarations can also specify a drive strength. See Clause 7 for more information on gates and for information on strengths. 3.4.1 Charge strength The charge strength specification shall be used only with trireg nets. A trireg net shall be used to model charge storage; charge strength shall specify the relative size of the capacitance indicated by one of the following keywords: — small — medium — large The default charge strength of a trireg net shall be medium. A trireg net can model a charge storage node whose charge decays over time. The simulation time of a charge decay shall be specified in the delay specification for the trireg net (see 7.14.2). Examples: trireg a; // a trireg net of charge strength medium trireg (large) #(0,0,50) cap1 ; // a trireg net of charge strength large //with charge decay time 50 time units trireg (small)signed [3:0] cap2 ; // a signed 4-bit trireg vector of // charge strength small 3.4.2 Drive strength The drive strength specification allows a continuous assignment to be placed on a net in the same statement that declares that net. See Clause 6 for more details. Net strength properties are described in detail in Clause 7. 24 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 3.5 Implicit declarations The syntax shown in 3.2 shall be used to declare nets and variables explicitly. In the absence of an explicit declaration, an implicit net of default net type shall be assumed in the following circumstances: — If an identifier is used in a port expression declaration, then an implicit net of type wire shall be assumed, with the vector width of the port expression declaration. See 12.3.3 for a discussion of port expression declarations. — If an identifier is used in the terminal list of a primitive instance or a module instance, and that identifier has not been explicitly declared previously in one of the declaration statements of the instantiating module, then an implicit scalar net of default net type shall be assumed. — If an identifier appears on the left-hand side of a continuous assignment statement, and that identifier has not been declared previously, then an implicit scalar net of default net type shall be assumed. See 6.1.2 for a discussion of continuous assignment statements. See 19.2 for a discussion of control of the type for implicitly declared nets with the `default_nettype compiler directive. 3.6 Net initialization The default initialization value for a net shall be the value z. Nets with drivers shall assume the output value of their drivers. The trireg net is an exception. The trireg net shall default to the value x, with the strength specified in the net declaration (small, medium, or large). 3.7 Net types There are several distinct types of nets, as shown in Table 2. wire wand wor Table 2—Net types tri triand trior tri0 tri1 trireg supply0 supply1 3.7.1 Wire and tri nets The wire and tri nets connect elements. The net types wire and tri shall be identical in their syntax and functions; two names are provided so that the name of a net can indicate the purpose of the net in that model. A wire net can be used for nets that are driven by a single gate or continuous assignment. The tri net type can be used where multiple drivers drive a net. Logical conflicts from multiple sources of the same strength on a wire or a tri net result in x (unknown) values. Table 3 is a truth table for resolving multiple drivers on wire and tri nets. Note that it assumes equal strengths for both drivers. Please refer to 7.9 for a discussion of logic strength modeling. Table 3—Truth table for wire and tri nets wire/ tri 0 1 x 01xz 0xx0 x1x1 xxxx Copyright © 2001 IEEE. All rights reserved. 25 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Table 3—Truth table for wire and tri nets wire/ tri z 01xz 01xz 3.7.2 Wired nets Wired nets are of type wor, wand, trior, and triand, and are used to model wired logic configurations. Wired nets use different truth tables to resolve the conflicts that result when multiple drivers drive the same net. The wor and trior nets shall create wired or configurations, such that when any of the drivers is 1, the resulting value of the net is 1. The wand and triand nets shall create wired and configurations, such that if any driver is 0, the value of the net is 0. The net types wor and trior shall be identical in their syntax and functionality. The net types wand and triand shall be identical in their syntax and functionality. Table 4 and Table 5 give the truth tables for wired nets. Note that they assume equal strengths for both drivers. See 7.9 for a discussion of logic strength modeling. Table 4—Truth table for wand and triand nets wand/ triand 01xz 0 0000 1 01x1 x 0xxx z 01xz Table 5—Truth table for wor and trior nets wor/ trior 01xz 0 01x0 1 1111 x x1xx z 01xz 3.7.3 Trireg net The trireg net stores a value and is used to model charge storage nodes. A trireg net can be in one of two states: driven state When at least one driver of a trireg net has a value of 1, 0, or x, the resolved value propagates into the trireg net and is the driven value of the trireg net. 26 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C capacitive state When all the drivers of a trireg net are at the high-impedance value (z), the trireg net retains its last driven value; the high-impedance value does not propagate from the driver to the trireg. The strength of the value on the trireg net in the capacitive state can be small, medium, or large, depending on the size specified in the declaration of the trireg net. The strength of a trireg net in the driven state can be supply, strong, pull, or weak, depending on the strength of the driver. Examples: Figure 1 shows a schematic that includes a trireg net whose size is medium, its driver, and the simulation results. wire a wire b wire c nmos1 nmos2 trireg d simulation time 0 10 wire a wire b wire c 1 1 strong 1 0 1 HiZ trireg d strong 1 medium 1 Figure 1—Simulation values of a trireg and its driver a) At simulation time 0, wire a and wire b have a value of 1. A value of 1 with a strong strength propagates from the and gate through the nmos switches connected to each other by wire c into trireg net d. b) At simulation time 10, wire a changes value to 0, disconnecting wire c from the and gate. When wire c is no longer connected to the and gate, the value of wire c changes to HiZ. The value of wire b remains 1 so wire c remains connected to trireg net d through the nmos2 switch. The HiZ value does not propagate from wire c into trireg net d. Instead, trireg net d enters the capacitive state, storing its last driven value of 1. It stores the 1 with a medium strength. 3.7.3.1 Capacitive networks A capacitive network is a connection between two or more trireg nets. In a capacitive network whose trireg nets are in the capacitive state, logic and strength values can propagate between trireg nets. Examples: Figure 2 shows a capacitive network in which the logic value of some trireg nets change the logic value of other trireg nets of equal or smaller size. Copyright © 2001 IEEE. All rights reserved. 27 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® wire a wire b wire c nmos_1 tranif1_1 wire d nmos_2 trireg_la trireg_sm tranif1_2 trireg_me1 trireg_me2 simulation time wire a wire b wire c wire d trireg_la trireg_sm trireg_me1 trireg_me2 0 1 11 1 1 1 1 1 10 1 01 1 1 1 1 1 20 1 00 1 0 1 1 1 30 1 00 0 0 1 0 1 40 0 00 0 0 1 0 1 50 0 10 0 0 0 x x Figure 2—Simulation results of a capacitive network In Figure 2, the capacitive strength of trireg_la net is large, trireg_me1 and trireg_me2 are medium, and trireg_sm is small. Simulation reports the following sequence of events: a) At simulation time 0, wire a and wire b have a value of 1. The wire c drives a value of 1 into trireg_la and trireg_sm; wire d drives a value of 1 into trireg_me1 and trireg_me2. b) At simulation time 10, the value of wire b changes to 0, disconnecting trireg_sm and trireg_me2 from their drivers. These trireg nets enter the capacitive state and store the value 1, their last driven value. c) At simulation time 20, wire c drives a value of 0 into trireg_la. d) At simulation time 30, wire d drives a value of 0 into trireg_me1. 28 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C e) At simulation time 40, the value of wire a changes to 0, disconnecting trireg_la and trireg_me1 from their drivers. These trireg nets enter the capacitive state and store the value 0. f) At simulation time 50, the value of wire b changes to 1. This change of value in wire b connects trireg_sm to trireg_la; these trireg nets have different sizes and stored different values. This connection causes the smaller trireg net to store the value of the larger trireg net, and trireg_sm now stores a value of 0. This change of value in wire b also connects trireg_me1 to trireg_me2; these trireg nets have the same size and stored different values. The connection causes both trireg_me1 and trireg_me2 to change value to x. In a capacitive network, charge strengths propagate from a larger trireg net to a smaller trireg net. Figure 3 shows a capacitive network and its simulation results. wire b wire c wire a tranif1_1 tranif1_2 trireg_la trireg_sm simulation time wire a wire b wire c trireg_la trireg_sm 0 strong 1 1 1 strong 1 strong 1 10 strong 1 0 1 large 1 large 1 20 strong 1 0 0 large 1 small 1 30 strong 1 0 1 large 1 large 1 40 strong 1 0 0 large 1 small 1 Figure 3—Simulation results of charge sharing In Figure 3, the capacitive strength of trireg_la is large and the capacitive strength of trireg_sm is small. Simulation reports the following results: a) At simulation time 0, the values of wire a, wire b, and wire c are 1, and wire a drives a strong 1 into trireg_la and trireg_sm. Copyright © 2001 IEEE. All rights reserved. 29 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® b) At simulation time 10, the value of wire b changes to 0, disconnecting trireg_la and trireg_sm from wire a. The trireg_la and trireg_sm nets enter the capacitive state. Both trireg nets share the large charge of trireg_la because they remain connected through tranif1_2. c) At simulation time 20, the value of wire c changes to 0, disconnecting trireg_sm from trireg_la. The trireg_sm no longer shares large charge of trireg_la and now stores a small charge. d) At simulation time 30, the value of wire c changes to 1, connecting the two trireg nets. These trireg nets now share the same charge. e) At simulation time 40, the value of wire c changes again to 0, disconnecting trireg_sm from trireg_la. Once again, trireg_sm no longer shares the large charge of trireg_la and now stores a small charge. 3.7.3.2 Ideal capacitive state and charge decay A trireg net can retain its value indefinitely or its charge can decay over time. The simulation time of charge decay is specified in the delay specification of the trireg net. See 7.14.2 for charge decay explanation. 3.7.4 Tri0 and tri1 nets The tri0 and tri1 nets model nets with resistive pulldown and resistive pullup devices on them. When no driver drives a tri0 net, its value is 0. When no driver drives a tri1 net, its value is 1. The strength of this value is pull. See Clause 7 for a description of strength modeling. A tri0 net is equivalent to a wire net with a continuous 0 value of pull strength driving it. A tri1 net is equivalent to a wire net with a continuous 1 value of pull strength driving it. A truth table for tri0 is shown in Table 6. A truth table for tri1 is shown in Table 7. Table 6—Truth table for tri0 net tri0 01xz 0 0xx0 1 x1x1 x xxxx z 01x0 Table 7—Truth table for tri1 net tri1 01xz 0 0xx0 1 x1x1 x xxxx z 01x1 30 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 3.7.5 Supply nets The supply0 and supply1 nets may be used to model the power supplies in a circuit. These nets shall have supply strengths. 3.8 Regs Assignments to a reg are made by procedural assignments (see 6.2 and 9.2). Since the reg holds a value between assignments, it can be used to model hardware registers. Edge-sensitive (i.e., flip-flops) and level sensitive (i.e., RS and transparent latches) storage elements can be modeled. A reg need not represent a hardware storage element since it can also be used to represent combinatorial logic. 3.9 Integers, reals, times, and realtimes In addition to modeling hardware, there are other uses for variables in an HDL model. Although reg variables can be used for general purposes such as counting the number of times a particular net changes value, the integer and time variable data types are provided for convenience and to make the description more selfdocumenting. The syntax for declaring integer, time, real, and realtime variables is given in Syntax 3-3 (from Syntax 3-2). integer_declaration ::= (From Annex A - A.2.1.3) integer list_of_variable_identifiers ; real_declaration ::= real list_of_real_identifiers ; realtime_declaration ::= realtime list_of_real_identifiers ; time_declaration ::= time list_of_variable_identifiers ; real_type ::= (From Annex A - A.2.2.1) real_identifier [ = constant_expression ] | real_identifier dimension { dimension } variable_type ::= variable_identifier [ = constant_expression ] | variable_identifier dimension { dimension } list_of_real_identifiers ::= (From Annex A- A.2.3) real_type { , real_type } list_of_variable_identifiers ::= variable_type { , variable_type } dimension ::= (From Annex A - A.2.5) [ dimension_constant_expression : dimension_constant_expression ] Syntax 3-3—Syntax for integer, time, real, and realtime declarations The syntax for a list of reg variables is defined in 3.2.2. An integer is a general-purpose variable used for manipulating quantities that are not regarded as hardware registers. Copyright © 2001 IEEE. All rights reserved. 31 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® A time variable is used for storing and manipulating simulation time quantities in situations where timing checks are required and for diagnostics and debugging purposes. This data type is typically used in conjunction with the $time system function (see 17.7.1). The integer and time variables shall be assigned values in the same manner as reg. Procedural assignments shall be used to trigger their value changes. The time variables shall behave the same as a reg of at least 64 bits, with the least significant bit being bit 0. They shall be unsigned quantities, and unsigned arithmetic shall be performed on them. In contrast, integer variables shall be treated as signed regs with the least significant bit being zero. Arithmetic operations performed on integer variables shall produce 2’s complement results. NOTE—Implementations may limit the maximum size of an integer variable, but they shall at least be 32 bits. The Verilog HDL supports real number constants and real variable data types in addition to integer and time variable data types. Except for the following restrictions, variables declared as real can be used in the same places that integer and time variables are used: — Not all Verilog HDL operators can be used with real number values. See Table 10 and Table 11 for lists of valid and invalid operators for real numbers and real variables. — Real variables shall not use range in the declaration. — Real variables shall default to an initial value of zero. The realtime declarations shall be treated synonymously with real declarations and can be used interchangeably. Examples: integer a; time last_chng; real float ; realtime rtime ; // integer value // time value // a variable to store real value // a variable to store time as a real value 3.9.1 Operators and real numbers The result of using logical or relational operators on real numbers and real variables is a single-bit scalar value. Not all Verilog HDL operators can be used with expressions involving real numbers and real variables. Table 10 lists the valid operators for use with real numbers and real variables. Real number constants and real variables are also prohibited in the following cases: — Edge descriptors (posedge, negedge) applied to real variables — Bit-select or part-select references of variables declared as real — Real number index expressions of bit-select or part-select references of vectors 3.9.2 Conversion Real numbers shall be converted to integers by rounding the real number to the nearest integer, rather than by truncating it. Implicit conversion shall take place when a real number is assigned to an integer. If the fractional part of the real number is exactly 0.5, it shall be rounded away from zero. Implicit conversion shall take place when an expression is assigned to a real. Individual bits that are x or z in the net or the variable shall be treated as zero upon conversion. See 17.8 for a discussion of system tasks that perform explicit conversion. 32 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 3.10 Arrays An array declaration for a net or a variable declares an element type which is either scalar or vector (see 3.3). For example: Declaration reg x[11:0]; wire [0:7] y[5:0]; reg [31:0] x [127:0]; Element Type scalar reg eight-bit-wide vector wire indexed from 0 to 7 thirty-two-bit-wide reg NOTE—Array size does not affect the element size.. Arrays can be used to group elements of the declared element type into multi-dimensional objects. Arrays shall be declared by specifying the element address range(s) after the declared identifier. Each dimension shall be represented by an address range. See 3.2.1 and 3.2.2 for net and variable declarations. The expression(s) that specify the indices of the array shall be constant expressions. The value of the constant expression can be a positive integer, a negative integer, or zero. One declaration statement can be used for declaring both arrays and elements of the declared data type. This ability makes it convenient to declare both arrays and elements that match the element vector width in the same declaration statement. An element can be assigned a value in a single assignment, but complete or partial array dimensions cannot. Nor can complete or partial array dimensions be used to provide a value to an expression. To assign a value to an element of an array, an index for every dimension shall be specified. The index can be an expression. This option provides a mechanism to reference different array elements depending on the value of other variables and nets in the circuit. For example, a program counter reg can be used to index into a RAM. Implementations may limit the maximum size of an array, but they shall at least be 16777216 (224). 3.10.1 Net arrays Arrays of nets can be used to connect ports of generated instances. Each element of the array can be used in the same fashion as a scalar or vector net. 3.10.2 reg and variable arrays Arrays for all variables types (reg, integer, time, real, realtime) shall be possible. 3.10.3 Memories A one dimensional array with elements of type reg is also called a memory. These memories can be used to model read-only memories (ROMs), random access memories (RAMs), and reg files. Each reg in the array is known as an element or word and is addressed by a single array index. An n-bit reg can be assigned a value in a single assignment, but a complete memory cannot. To assign a value to a memory word, an index shall be specified. The index can be an expression. This option provides a mechanism to reference different memory words, depending on the value of other variables and nets in the circuit. For example, a program counter reg could be used to index into a RAM. Copyright © 2001 IEEE. All rights reserved. 33 IEEE Std 1364-2001 Version C 3.10.3.1 Array examples 3.10.3.1.1 Array declarations IEEE STANDARD VERILOG® reg [7:0] mema[0:255]; // declares a memory mema of 256 8-bit // registers. The indices are 0 to 255 reg arrayb[7:0][0:255]; wire w_array[7:0][5:0]; integer inta[1:64]; time chng_hist[1:1000] integer t_index; // declare a two dimensional array of // one bit registers // declare array of wires // an array of 64 integer values // an array of 1000 time values 3.10.3.1.2 Assignment to array elements The assignment statements in this section assume the presence of the declarations in 3.10.3.1.1. mema = 0; // Illegal syntax- Attempt to write to entire array arrayb[1] = 0; // Illegal Syntax - Attempt to write to elements // [1][0]..[1][255] arrayb[1][12:31] = 0; // Illegal Syntax - Attempt to write to // elements [1][12]..[1][31] mema[1] = 0; //Assigns 0 to the second element of mema arrayb[1][0] = 0; // Assigns 0 to the bit referenced by indices // [1][0] inta[4] = 33559; // Assign decimal number to integer in array chng_hist[t_index] = $time; // Assign current simulation time to // element addressed by integer index 3.10.3.1.3 Memory differences A memory of n 1-bit regs is different from an n-bit vector reg reg [1:n] rega; // An n-bit register is not the same reg mema [1:n]; // as a memory of n 1-bit registers 3.11 Parameters Verilog HDL parameters do not belong to either the variable or the net group. Parameters are not variables, they are constants. There are two types of parameters: module parameters and specify parameters. It is illegal to redeclare a name already declared by a net, parameter or variable declaration. Both types of parameters accept a range specification. By default, parameters and specparams shall be as wide as necessary to contain the value of the constant, except when a range specification is present. 34 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE 3.11.1 Module parameters The syntax for module parameter declarations is given in Syntax 3-4. IEEE Std 1364-2001 Version C local_parameter_declaration ::= (From Annex A - A.2.1.1) localparam [ signed ] [ range ] list_of_param_assignments ; | localparam integer list_of_param_assignments ; | localparam real list_of_param_assignments ; | localparam realtime list_of_param_assignments ; | localparam time list_of_param_assignments ; parameter_declaration ::= parameter [ signed ] [ range ] list_of_param_assignments ; | parameter integer list_of_param_assignments ; | parameter real list_of_param_assignments ; | parameter realtime list_of_param_assignments ; | parameter time list_of_param_assignments ; list_of_param_assignments ::= (From Annex A - A.2.3) param_assignment { , param_assignment } param_assignment ::= (From Annex A - A.2.4) parameter_identifier = constant_expression range ::= (From Annex A - A.2.5) [ msb_constant_expression : lsb_constant_expression ] Syntax 3-4—Syntax for module parameter declaration The list_of_param_assignments shall be a comma-separated list of assignments, where the right-hand side of the assignment shall be a constant expression; that is, an expression containing only constant numbers and previously defined parameters. (See Clause 4) The list_of_param_assignments can appear in a module as a set of module_items or in the module declaration in the module_parameter_port_list. (See 12.1). If any param_assignments appear in a module_parameter_port_list, then any param_assignments that appear in the module become local parameters and shall not be overridden by any method. Parameters represent constants; hence, it is illegal to modify their value at runtime. However, module parameters can be modified at compilation time to have values that are different from those specified in the declaration assignment. This allows customization of module instances. A parameter can be modified with the defparam statement or in the module instance statement. Typical uses of parameters are to specify delays and width of variables. See Clause 12 for details on parameter value assignment. A module parameter can have a type specification and a range specification. The type and range of module parameters shall be in accordance with the following rules: — A parameter declaration with no type or range specification shall default to the type and range of the final value assigned to the parameter, after any value overrides have been applied. — A parameter with a range specification, but with no type specification, shall be the range of the parameter declaration and shall be unsigned. The sign and range shall not be affected by value overrides. — A parameter with a type specification, but with no range specification, shall be of the type specified. A signed parameter shall default to the range of the final value assigned to the parameter, after any value overrides have been applied. Copyright © 2001 IEEE. All rights reserved. 35 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® — A parameter with a signed type specification and with a range specification shall be signed, and shall be the range of its declaration. The sign and range shall not be affected by value overrides. — A parameter with no range specification, and with either a signed type specification or no type specification, shall have an implied range with an lsb equal to 0 and an msb equal to one less than the size of the final value assigned to the parameter. — A parameter with no range specification, and with either a signed type specification or no type specification, and for which the final value assigned to it is unsized, shall have an implied range with an lsb equal to 0 and an msb equal to an implementation-dependent value of at least 31. The conversion rules between real and integer values described in 3.9.2 apply to parameters as well. Examples: parameter parameter parameter parameter parameter msb = 7; // defines msb as a constant value 7 e = 25, f = 9; // defines two constant numbers r = 5.7; // declares r as a real parameter byte_size = 8, byte_mask = byte_size - 1; average_delay = (r + f) / 2; parameter signed [3:0] mux_selector = 0; parameter real r1 = 3.5e17; parameter p1 = 13’h7e; parameter [31:0] dec_const = 1’b1; // value converted to 32 bits parameter newconst = 3’h4; // implied range of [2:0] parameter newconst = 4; // implied range of at least [31:0] 3.11.2 Local parameters - localparam Verilog HDL localparam - local parameter(s) are identical to parameters except that they can not directly be modified by defparam statements (see 12.2.1) or module instance parameter value assignments (see 12.2.2). Local parameters can be assigned constant expressions containing parameters, which can be modified with defparam statements or module instance parameter value assignments. The syntax for local parameter declarations is given in Syntax 3-4. 36 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE 3.11.3 Specify parameters The syntax for declaring specify parameters is shown in Syntax 3-5. IEEE Std 1364-2001 Version C specparam_declaration ::= (From Annex A - A.2.1.1) specparam [ range ] list_of_specparam_assignments ; list_of_specparam_assignments ::= (From Annex A- A.2.3) specparam_assignment { , specparam_assignment } specparam_assignment ::= (From Annex A - A.2.4) specparam_identifier = constant_mintypmax_expression | pulse_control_specparam pulse_control_specparam ::= PATHPULSE$ = ( reject_limit_value [ , error_limit_value ] ) ; | PATHPULSE$specify_input_terminal_descriptor$specify_output_terminal_descriptor = ( reject_limit_value [ , error_limit_value ] ) ; error_limit_value ::= limit_value reject_limit_value ::= limit_value limit_value ::= constant_mintypmax_expression range ::= (From Annex A - A.2.5) [ msb_constant_expression : lsb_constant_expression ] Syntax 3-5—Syntax of the specparam declaration The keyword specparam declares a special type of parameter which is intended only for providing timing and delay values, but can appear in any expression that is not assigned to a parameter and is not part of the range specification of a declaration. Originally permitted only in specify blocks (see Clause 14), with this revision specify parameters (also called specparams) are now permitted both within the specify block and in the main module body. A specify parameter declared outside a specify block shall be declared before it is referenced. The value assigned to a specify parameter can be any constant expression. A specify parameter can be used as part of a constant expression for a subsequent specify parameter declaration. Unlike a module parameter, a specify parameter cannot be modified from within the language, but it may be modified through SDF annotation (see Clause 16). The specify parameters and module parameters shall not be interchangeable. In addition, module parameters shall not be assigned a constant expression that includes any specify parameters. Table 8 summarizes the differences between the two types of parameter declarations. Table 8—Differences between specparams and parameters Specparams (specify parameter) Use keyword specparam Shall be declared inside a module or specify block May only be used inside a module or specify block May be assigned specparams and parameters Use SDF annotation to override values Parameters (module parameter) Use keyword parameter Shall be declared outside specify blocks May not be used inside specify blocks May not be assigned specparams Use defparam or instance declaration parameter value passing to override values Copyright © 2001 IEEE. All rights reserved. 37 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® A specify parameter can have a range specification. The range of specify parameters shall be in accordance with the following rules: — A specparam declaration with no range specification shall default to the range of the final value assigned to the parameter, after any value overrides have been applied. — A specparam with a range specification shall be the range of the parameter declaration. The range shall not be affected by value overrides. Examples: specify specparam tRise_clk_q = 150, tFall_clk_q = 200; specparam tRise_control = 40, tFall_control = 50; endspecify The lines between the keywords specify and endspecify declare four specify parameters. The first line declares specify parameters called tRise_clk_q and tFall_clk_q with values 150 and 200 respectively; the second line declares tRise_control and tFall_control specify parameters with values 40 and 50 respectively. Examples: module RAM16GEN (output [7:0] DOUT, input [7:0] DIN, input [5:0] ADR, input WE, CE); specparam dhold = 1.0; specparam ddly = 1.0; parameter width = 1; parameter regsize = dhold + 1.0; // Illegal - can’t assign // specparams to parameters endmodule 3.12 Name spaces In Verilog HDL, there are seven name spaces; two are global and five are local. The global name spaces are definitions and text macros. The definitions name space unifies all the module (see 12.1), macromodule (see 12.1), and primitive (see 8.1) definitions. Once a name is used to define a module, macromodule, or primitive, the name shall not be used again to declare another module, macromodule, or primitive. The text macro name space is global. Since text macro names are introduced and used with a leading ` character, they remain unambiguous with any other name space (see 19.3). The text macro names are defined in the linear order of appearance in the set of input files that make up the description of the design unit. Subsequent definitions of the same name override the previous definitions for the balance of the input files. There are five local name spaces: block, module, port, specify block, and attribute. Once a name is defined within one of the five name spaces, it shall not be defined again in that space (with the same or a different type). The block name space is introduced by the named block (see 9.8), function (see 10.3), and task (see 10.2) constructs. It unifies the definitions of the named blocks, functions, tasks, parameters, named events and the variable type of declaration (see 3.2.2). The variable type of declaration includes the reg, integer, time, real, and realtime declarations. The module name space is introduced by the module, macromodule, and primitive constructs. It unifies the definition of functions, tasks, named blocks, instance names, parameters, named events, net type of declaration, and variable type of declaration. The net type of declaration includes wire, wor, wand, tri, trior, triand, tri0, tri1, trireg, supply0, and supply1 (see 3.7). 38 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The port name space is introduced by the module, macromodule, primitive, function, and task constructs. It provides a means of structurally defining connections between two objects that are in two different name spaces. The connection can be unidirectional (either input or output) or bidirectional (inout). The port name space overlaps the module and the block name spaces. Essentially, the port name space specifies the type of connection between names in different name spaces. The port type of declarations include input, output, and inout (see 12.3). A port name introduced in the port name space may be reintroduced in the module name space by declaring a variable or a wire with the same name as the port name. The specify block name space is introduced by the specify construct (see 14.2). The attribute name space is enclosed by the (* and *) constructs attached to a language element (see 2.8). An attribute name can be defined and used only in the attribute name space. Any other type of name cannot be defined in this name space. Copyright © 2001 IEEE. All rights reserved. 39 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 4. Expressions This clause describes the operators and operands available in the Verilog HDL and how to use them to form expressions. An expression is a construct that combines operands with operators to produce a result that is a function of the values of the operands and the semantic meaning of the operator. Any legal operand, such as a net bitselect, without any operator is considered an expression. Wherever a value is needed in a Verilog HDL statement, an expression can be used. Some statement constructs require an expression to be a constant expression. The operands of a constant expression consist of constant numbers, parameter names, constant bit-selects of parameters, constant partselects of parameters, and constant function calls (see 10.3.5) only, but they can use any of the operators defined in Table 9. A scalar expression is an expression that evaluates to a scalar (single-bit) result. If the expression evaluates to a vector (multibit) result, then the least significant bit of the result is used as the scalar result. The data types reg, integer, time, real, and realtime are all variable data types. Descriptions pertaining to variable usage apply to all of these data types. An operand can be one of the following: — Constant number (including real) — Net — Variables of type reg, integer, time, real, and realtime — Net bit-select — Bit-select of type reg, integer, and time — Net part-select — Part-select of type reg, integer, and time — Array element — A call to a user-defined function or system-defined function that returns any of the above 4.1 Operators The symbols for the Verilog HDL operators are similar to those in the C programming language. Table 9 lists these operators. Table 9—Operators in the Verilog HDL {} {{}} + - * / ** % > >= < <= ! && || == != Concatenation, replication Arithmetic Modulus Relational Logical negation Logical and Logical or Logical equality Logical inequality 40 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Table 9—Operators in the Verilog HDL (continued) === !== ~ & | ^ ^~ or ~^ & ~& | ~| ^ ~^ or ^~ << >> <<< >>> ?: or Case equality Case inequality Bit-wise negation Bit-wise and Bit-wise inclusive or Bit-wise exclusive or Bit-wise equivalence Reduction and Reduction nand Reduction or Reduction nor Reduction xor Reduction xnor Logical left shift Logical right shift Arithmetic left shift Arithmetic right shift Conditional Event or 4.1.1 Operators with real operands The operators shown in Table 10 shall be legal when applied to real operands. All other operators shall be considered illegal when used with real operands. Table 10—Legal operators for use in real expressions unary + unary - Unary operators + - * / ** Arithmetic > >= < <= Relational ! && || Logical == != Logical equality ?: Conditional or Event or The result of using logical or relational operators on real numbers is a single-bit scalar value. Copyright © 2001 IEEE. All rights reserved. 41 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Table 11 lists operators that shall not be used to operate on real numbers. Table 11—Operators not allowed for real expressions {} {{}} % === !== ~&| ^ ^~ ~^ ^ ^~ ~^ & ~& | ~| << >> <<< >>> Concatenate, replicate Modulus Case equality Bit-wise Reduction Shift See 3.9.1 for more information on use of real numbers. 4.1.2 Operator precedence The precedence order of the Verilog operators is shown in Table 12. Table 12—Precedence rules for operators + - ! ~ (unary) ** */% + - (binary) << >> <<< >>> < <= > >= == != === !== & ~& ^ ^~ ~^ | ~| && || ?: (conditional operator) Highest precedence Lowest precedence Operators shown on the same row in Table 12 shall have the same precedence. Rows are arranged in order of decreasing precedence for the operators. For example, *, /, and % all have the same precedence, which is higher than that of the binary + and - operators. All operators shall associate left to right with the exception of the conditional operator, which shall associate right to left. Associativity refers to the order in which the operators having the same precedence are evaluated. Thus, in the following example B is added to A and then C is subtracted from the result of A+B. A+B-C 42 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C When operators differ in precedence, the operators with higher precedence shall associate first. In the following example, B is divided by C (division has higher precedence than addition) and then the result is added to A. A+B/C Parentheses can be used to change the operator precedence. (A + B) / C // not the same as A + B / C 4.1.3 Using integer numbers in expressions Integer numbers can be used as operands in expressions. An integer number can be expressed as — An unsized, unbased integer (e.g., 12) — An unsized, based integer (e.g., ’d12, ’sd12) — A sized, based integer (e.g., 16’d12, 16’sd12) A negative value for an integer with no base specifier shall be interpreted differently than for an integer with a base specifier. An integer with no base specifier shall be interpreted as a signed value in 2’s complement form. An integer with an unsigned base specifier shall be interpreted as an unsigned value. Example: This example shows four ways to write the expression “minus 12 divided by 3.” Note that -12 and -’d12 both evaluate to the same 2’s complement bit pattern, but, in an expression, the -’d12 loses its identity as a signed negative number. integer IntA; IntA = -12 / 3; // The result is -4. IntA = -’d 12 / 3; // The result is 1431655761. IntA = -’sd 12 / 3; // The result is -4. IntA = -4'sd 12 / 3; // -4'sd12 is the negative of the 4-bit // quantity 1100, which is -4. -(-4) = 4. // The result is 1. 4.1.4 Expression evaluation order The operators shall follow the associativity rules while evaluating an expression as described in 4.1.2. However, if the final result of an expression can be determined early, the entire expression need not be evaluated. This is called short-circuiting an expression evaluation. Example: reg regA, regB, regC, result ; result = regA & (regB | regC) ; If regA is known to be zero, the result of the expression can be determined as zero without evaluating the sub-expression regB | regC. Copyright © 2001 IEEE. All rights reserved. 43 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 4.1.5 Arithmetic operators The binary arithmetic operators are given in Table 13. Table 13—Arithmetic operators defined a+b a-b a*b a/b a%b a ** b a plus b a minus b a multiplied by b (or a times b) a divided by b a modulo b a to the power of b The integer division shall truncate any fractional part toward zero. For the division or modulus operators, if the second operand is a zero, then the entire result value shall be x. The modulus operator, for example y % z, gives the remainder when the first operand is divided by the second, and thus is zero when z divides y exactly. The result of a modulus operation shall take the sign of the first operand. The result of the power operator shall be real if either operand is a real, integer, or signed. If both operands are unsigned then the result shall be unsigned. The result of the power operator is unspecified if the first operand is zero and the second operand is non-positive, or if the first operand is negative and the second operand is not an integral value. The unary arithmetic operators shall take precedence over the binary operators. The unary operators are given in Table 14. Table 14—Unary operators defined +m Unary plus m (same as m) -m Unary minus m For the arithmetic operators, if any operand bit value is the unknown value x or the high-impedance value z, then the entire result value shall be x. Example: Table 15 gives examples of modulus operations. Table 15—Examples of modulus operators Modulus expression 10 % 3 11 % 3 12 % 3 -10 % 3 11 % -3 -4’d12 % 3 Result 1 2 0 -1 2 1 Comments 10/3 yields a remainder of 1 11/3 yields a remainder of 2 12/3 yields no remainder The result takes the sign of the first operand The result takes the sign of the first operand -4’d12 is seen as a large, positive number that leaves a remainder of 1 when divided by 3 44 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 4.1.6 Arithmetic expressions with regs and integers A reg data type shall be treated as an unsigned value unless explicitly declared to be signed. An integer variable shall be treated as signed. Signed values shall use a 2's complement representation. Conversions between signed and unsigned values shall keep the same bit representation; only the interpretation changes. Table 16 lists how arithmetic operators interpret each data type. Table 16—Data type interpretation by arithmetic operators Data type Interpretation unsigned net Unsigned signed net Signed, 2’s complement unsigned reg Unsigned signed reg Signed, 2’s complement integer Signed, 2’s complement time Unsigned real, realtime Signed, floating point Example: The following example shows various ways to divide “minus twelve by three”— using integer and reg data types in expressions. integer intA; reg [15:0] regA; reg signed [15:0] regS; intA = -4’d12; regA = intA / 3; // expression result is -4, // intA is an integer data type, regA is 65532 regA = -4’d12; // regA is 65524 intA = regA / 3; // expression result is 21841, // regA is a reg data type intA = -4’d12 / 3; // expression result is 1431655761. // -4’d12 is effectively a 32-bit reg data type regA = -12 / 3; // expression result is -4, -12 is effectively // an integer data type. regA is 65532 regS = -12 / 3; // expression result is -4. regS is a signed // reg regS = -4’sd12 / 3;// expression result is 1. -4’sd12 is actually // 4. The rules for integer division yield 4/3==1 Copyright © 2001 IEEE. All rights reserved. 45 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 4.1.7 Relational operators Table 17 lists and defines the relational operators. Table 17—Definitions of the relational operators ab a <= b a >= b a less than b a greater than b a less than or equal to b a greater than or equal to b An expression using these relational operators shall yield the scalar value 0 if the specified relation is false or the value 1 if it is true. If either operand of a relational operator contains an unknown (x) or high impedance (z) value, then the result shall be a 1-bit unknown value (x). When two operands of unequal bit lengths are used and one or both of the operands is unsigned, the smaller operand shall be zero filled on the most significant bit side to extend to the size of the larger operand. All the relational operators shall have the same precedence. Relational operators shall have lower precedence than arithmetic operators. Examples: The following examples illustrate the implications of this precedence rule: a < foo - 1 a < (foo - 1) foo - (1 < a) foo - 1 < a // this expression is the same as // this expression, but . . . // this one is not the same as // this expression When foo - (1 < a) evaluates, the relational expression evaluates first and then either zero or one is subtracted from foo. When foo - 1 < a evaluates, the value of foo operand is reduced by one and then compared with a. When both operands of a relational expression are signed integral operands (an integer, a signed reg data type, or an unsized, unbased integer) then the expression shall be interpreted as a comparison between signed values. When either operand of a relational expression is a real operand then the other operand shall be converted to an equivalent real value, and the expression shall be interpreted as a comparison between two real values. Otherwise the expression shall be interpreted as a comparison between unsigned values. 4.1.8 Equality operators The equality operators shall rank lower in precedence than the relational operators. Table 18 lists and defines the equality operators. Table 18—Definitions of the equality operators a ===b a !==b a ==b a !=b a equal to b, including x and z a not equal to b, including x and z a equal to b, result may be unknown a not equal to b, result may be unknown 46 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C All four equality operators shall have the same precedence. These four operators compare operands bit for bit, with zero filling if the two operands are of unequal bit length. As with the relational operators, the result shall be 0 if comparison fails, 1 if it succeeds. For the logical equality and logical inequality operators (== and !=), if, due to unknown or high-impedance bits in the operands, the relation is ambiguous, then the result shall be a one bit unknown value (x). For the case equality and case inequality operators (=== and !==), the comparison shall be done just as it is in the procedural case statement (see 9.5). Bits that are x or z shall be included in the comparison and shall match for the result to be considered equal. The result of these operators shall always be a known value, either 1 or 0. 4.1.9 Logical operators The operators logical and (&&) and logical or (||) are logical connectives. The result of the evaluation of a logical comparison shall be 1 (defined as true), 0 (defined as false), or, if the result is ambiguous, the unknown value (x). The precedence of && is greater than that of ||, and both are lower than relational and equality operators. A third logical operator is the unary logical negation operator (!). The negation operator converts a non- zero or true operand into 0 and a zero or false operand into 1. An ambiguous truth value remains as x. Examples: Example 1—If reg alpha holds the integer value 237 and beta holds the value zero, then the following examples perform as described: regA = alpha && beta; // regA is set to 0 regB = alpha || beta; // regB is set to 1 Example 2—The following expression performs a logical and of three subexpressions without needing any parentheses: a < size-1 && b != c && index != lastone However, it is recommended for readability purposes that parentheses be used to show very clearly the precedence intended, as in the following rewrite of this example: (a < size-1) && (b != c) && (index != lastone) Example 3—A common use of ! is in constructions like the following: if (!inword) In some cases, the preceding construct makes more sense to someone reading the code than this equivalent construct: if (inword == 0) 4.1.10 Bit-wise operators The bit-wise operators shall perform bit-wise manipulations on the operands—that is, the operator shall combine a bit in one operand with its corresponding bit in the other operand to calculate one bit for the result. Logic Tables 19 through 23 show the results for each possible calculation. Copyright © 2001 IEEE. All rights reserved. 47 IEEE Std 1364-2001 Version C Table 19—Bit-wise binary and operator & 01xz 0 0000 1 01xx x 0xxx z 0xxx Table 20—Bit-wise binary or operator | 01xz 0 01xx 1 1111 x x1xx z x1xx IEEE STANDARD VERILOG® Table 21—Bit-wise binary exclusive or operator ^ 01xz 0 01xx 1 10xx x xxxx z xxxx Table 22—Bit-wise binary exclusive nor operator ^~ ~^ 01xz 0 10xx 1 01xx x xxxx z xxxx Table 23—Bit-wise unary negation operator ~ 0 1 1 0 x x z x When the operands are of unequal bit length, the shorter operand is zero-filled in the most significant bit positions. 4.1.11 Reduction operators The unary reduction operators shall perform a bit-wise operation on a single operand to produce a single bit result. For reduction and, reduction or, and reduction xor operators, the first step of the operation shall apply the operator between the first bit of the operand and the second using logic Tables 24 through 26. The second and subsequent steps shall apply the operator between the 1-bit result of the prior step and the next bit of the operand using the same logic table. For reduction nand, reduction nor, and reduction xnor operators, the result shall be computed by inverting the result of the reduction and, reduction or, and reduction xor operation respectively. 48 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE Table 24—Reduction unary and operator & 01xz 0 0000 1 01xx x 0xxx z 0xxx IEEE Std 1364-2001 Version C Table 25—Reduction unary or operator | 01xz 0 01xx 1 1111 x x1xx z x1xx Table 26—Reduction unary exclusive or operator ^ 01xz 0 01xx 1 10xx x xxxx z xxxx Example: Table 27 shows the results of applying reduction operators on different operands. Table 27—Results of unary reduction operations Operand & ~& | 4’b0000 0 1 0 4’b1111 1 0 1 4’b0110 0 1 1 4’b1000 0 1 1 ~| ^ ~^ Comments 1 0 1 No bits set 0 0 1 All bits set 0 0 1 Even number of bits set 0 1 0 Odd number of bits set 4.1.12 Shift operators There are two types of shift operators, the logical shift operators, << and >>, and the arithmetic shift operators, <<< and >>>. The left shift operators, << and <<<, shall shift their left operand to the left by the number by the number of bit positions given by the right operand. In both cases, the vacated bit positions shall be filled with zeroes. The right shift operators, >> and >>>, shall shift their left operand to the right by the number of bit positions given by the right operand. The logical right shift shall fill the vacated bit positions with Copyright © 2001 IEEE. All rights reserved. 49 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® zeroes. The arithmetic right shift shall fill the vacated bit positions with zeroes if the result type is unsigned. It shall fill the vacated bit positions with the value of the most-significant (i.e., sign) bit of the left operand if the result type is signed. If the right operand has an unknown or high impedance value, then the result shall be unknown. The right operand is always treated as an unsigned number and has no effect on the signedness of the result. The result signedness is determined by the left-hand operand and the remainder of the expression, as outlined in 4.5.1. Examples: Example 1—In this example, the reg result is assigned the binary value 0100, which is 0001 shifted to the left two positions and zero-filled. module shift; reg [3:0] start, result; initial begin start = 1; result = (start << 2); end endmodule Example 2—In this example, the reg result is assigned the binary value 1110, which is 1000 shifted to the right two positions and sign-filled. module ashift; reg signed [3:0] start, result; initial begin start = 4’b1000; result = (start >>> 2); end endmodule 4.1.13 Conditional operator The conditional operator, also known as ternary operator, shall be right associative and shall be constructed using three operands separated by two operators in the format given in Syntax 4-1. conditional_expression ::= (From Annex A - A.8.3) expression1 ? { attribute_instance } expression2 : expression3 expression1 ::= expression expression2 ::= expression expression3 ::= expression Syntax 4-1—Syntax for conditional operator The evaluation of a conditional operator shall begin with the evaluation of expression1. If expression1 evaluates to false (0), then expression3 shall be evaluated and used as the result of the conditional expression. If expression1 evaluates to true (known value other than 0), then expression2 is evaluated and used as the result. If expression1 evaluates to ambiguous value (x or z), then both expression2 and expression3 shall be evaluated and their results shall be combined, bit by bit, using Table 28 to calculate the final result unless 50 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C expression2 or expression3 is real, in which case the result shall be 0. If the lengths of expression2 and expression3 are different, the shorter operand shall be lengthened to match the longer and zero-filled from the left (the high-order end). Table 28—Ambiguous condition results for conditional operator ?: 01xz 0 0xxx 1 x1xx x xxxx z xxxx Example: The following example of a three-state output bus illustrates a common use of the conditional operator. wire [15:0] busa = drive_busa ? data : 16’bz; The bus called data is driven onto busa when drive_busa is 1. If drive_busa is unknown, then an unknown value is driven onto busa. Otherwise, busa is not driven. 4.1.14 Concatenations A concatenation is the joining together of bits resulting from two or more expressions. The concatenation shall be expressed using the brace characters { and }, with commas separating the expressions within. Unsized constant numbers shall not be allowed in concatenations. This is because the size of each operand in the concatenation is needed to calculate the complete size of the concatenation. Examples: This example concatenates four expressions: {a, b[3:0], w, 3’b101} and it is equivalent to the following example: {a, b[3], b[2], b[1], b[0], w, 1’b1, 1’b0, 1’b1} Another form of concatenation is the replication operation. The first expression shall be a non-zero, non-X and non-Z constant expression, the second expression follows the rules for concatenations. This example replicates "w" 4 times. {4{w}} // This is equivalent to {w, w, w, w} a[31:0] = {1’b1, {0{1’b0}} }; //illegal. RHS becomes {1’b1,; a[31:0] = {1’b1, {1’bz{1’b0}} }; //illegal. RHS becomes {1’b1,; a[31:0] = {1’b1, {1’bx{1’b0}} }; //illegal. RHS becomes {1’b1,; Copyright © 2001 IEEE. All rights reserved. 51 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® If the replication operator is used on a function call operand, the function need not be evaluated multiple times. For example: result = {4{func(w)}} may be computed as result = {func(w), func(w), func(w), func(w)} or y = func(w) ; result = {y, y, y, y} This is another form of expression evaluation short-circuiting. The next example illustrates nested concatenations: {b, {3{a, b}}} // This is equivalent to {b, a, b, a, b, a, b} 4.1.15 Event or The event or operator shall perform an or of events. The , operator does the same thing. See 9.7 for events and triggering of events. Example: The following example shows both ways to make an assignment to rega when an event (change) occurs on trig or enable. @(trig or enable) rega = regb ; @(trig , enable) rega = regb ; 4.2 Operands There are several types of operands that can be specified in expressions. The simplest type is a reference to a net or variable in its complete form—that is, just the name of the net or variable is given. In this case, all of the bits making up the net or variable value shall be used as the operand. If a single bit of a vector net, reg variable, integer variable, or time variable is required, then a bit-select operand shall be used. A part-select operand shall be used to reference a group of adjacent bits in a vector net, vector reg, integer variable, or time variable. A memory word can be referenced as an operand. A concatenation of other operands (including nested concatenations) can be specified as an operand. A function call is an operand. 4.2.1 Vector bit-select and part-select addressing Bit-selects extract a particular bit from a vector net, vector reg, integer variable, or time variable. The bit can be addressed using an expression. If the bit-select is out of the address bounds or the bit-select is x or z, then the value returned by the reference shall be x. The bit-select or part-select of a variable declared as real or realtime shall be considered illegal. Several contiguous bits in a vector net, vector reg, integer variable, or time variable can be addressed and are known as part-selects. There are two types of part-selects, a constant part-select and an indexed part-select. 52 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C A constant part-select of a vector reg or net is given with the following syntax: vect[msb_expr:lsb_expr] Both expressions shall be constant expressions. The first expression has to address a more significant bit than the second expression. If the part-select is out of the address bounds or the part-select is x or z, then the value returned by the reference shall be x. An indexed part select of a vector net, vector reg, integer variable, or time variable is given with the following syntax: reg [15:0] big_vect; reg [0:15] little_vect; big_vect[lsb_base_expr +: width_expr] little_vect[msb_base_expr +: width_expr] big_vect[msb_base_expr -: width_expr] little_vect[lsb_base_expr -: width_expr] The width_expr shall be a constant expression. It also shall not be affected by run-time parameter assignments. The lsb_base_expr and msb_base_expr can vary at run-time. The first two examples select bits starting at the base and ascending the bit range. The number of bits selected is equal to the width expression. The second two examples select bits starting at the base and descending the bit range. Part-selects that address a range of bits that are completely out of the address bounds of the net, reg, integer, or time, or when the part-select is x or z, shall yield the value x when read, and shall have no effect on the data stored when written. Part-selects that are partially out of range shall when read return x for the bits that are out of range, and when written shall only affect the bits that are in range. Examples: reg [31:0] big_vect; reg [0:31] little_vect; reg [63:0] dword; integer sel; The first four if statements show the identity between the two part select constructs. The last one shows an indexable nature. initial begin if ( big_vect[0 +:8] == big_vect[7 : 0]) begin end if (little_vect[0 +:8] == little_vect[0 : 7]) begin end if ( big_vect[15 -:8] == big_vect[15 : 8]) begin end if (little_vect[15 -:8] == little_vect[8 :15]) begin end if (sel >0 && sel < 8) dword[8*sel +:8] = big_vect[7:0]; // Replace the byte selected. Examples: Example 1—The following example specifies the single bit of acc vector that is addressed by the operand index. acc[index] Copyright © 2001 IEEE. All rights reserved. 53 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® The actual bit that is accessed by an address is, in part, determined by the declaration of acc. For instance, each of the declarations of acc shown in the next example causes a particular value of index to access a different bit: reg [15:0] acc; reg [2:17] acc Example 2—The next example and the bullet items that follow it illustrate the principles of bit addressing. The code declares an 8-bit reg called vect and initializes it to a value of 4. The list describes how the separate bits of that vector can be addressed. reg [7:0] vect; vect = 4; // fills vect with the pattern 00000100 // msb is bit 7, lsb is bit 0 — If the value of addr is 2, then vect[addr] returns 1. — If the value of addr is out of bounds, then vect[addr] returns x. — If addr is 0, 1, or 3 through 7, vect[addr] returns 0. — vect[3:0] returns the bits 0100. — vect[5:1] returns the bits 00010. — vect[expression that returns x] returns x. — vect[expression that returns z] returns x. — If any bit of addr is x or z, then the value of addr is x. NOTES: 1) Part-select indices that evaluate to x or z may be flagged as a compile time error. 2) Bit-select or part-select indices that are outside of the declared range may be flagged as a compile time error. 4.2.2 Array and memory addressing Declaration of arrays and memories (one dimensional arrays of reg) are discussed in 3.10. This subclause discusses array addressing. Examples: The next example declares a memory of 1024 8-bit words: reg [7:0] mem_name[0:1023]; The syntax for a memory address shall consist of the name of the memory and an expression for the address, specified with the following format: mem_name[addr_expr] The addr_expr can be any expression; therefore, memory indirections can be specified in a single expression. The next example illustrates memory indirection: mem_name[mem_name[3]] In this example, mem_name[3]addresses word three of the memory called mem_name. The value at word three is the index into mem_name that is used by the memory address mem_name[mem_name[3]]. As with bit-selects, the address bounds given in the declaration of the memory determine the effect of the address expression. If the index is out of the address bounds or if any bit in the address is x or z, then the value of the reference shall be x. 54 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Examples: The next example declares an array of 256 by 256 8-bit elements and an array 256 by 256 by 8 1-bit elements: reg [7:0] twod_array[0:255][0:255]; wire threed_array[0:255][0:255][0:7]; The syntax for access to the array shall consist of the name of the memory or array and an expression for each addressed dimension: twod_array[addr_expr][addr_expr] threed_array[addr_expr][addr_expr][addr_expr] As before, the addr_expr can be any expression. The array twod_array accesses a whole 8-bit vector, while the array threed_array accesses a single bit of the three dimensional array. To express bit selects or part selects of array elements, the desired word shall first be selected by supplying an address for each dimension. Once selected, bit and part selects shall be addressed in the same manner as net and reg bit and part selects (see 4.2.1). Examples: twod_array[14][1][3:0] twod_array[1][3][6] twod_array[1][3][sel] threed_array[14][1][3:0] // access lower 4 bits of word // access bit 6 of word // use variable bit select // Illegal 4.2.3 Strings String operands shall be treated as constant numbers consisting of a sequence of 8-bit ASCII codes, one per character. Any Verilog HDL operator can manipulate string operands. The operator shall behave as though the entire string were a single numeric value. When a variable is larger than required to hold the value being assigned, the contents after the assignment shall be padded on the left with zeros. This is consistent with the padding that occurs during assignment of nonstring values. Example: The following example declares a string variable large enough to hold 14 characters and assigns a value to it. The example then manipulates the string using the concatenation operator. module string_test; reg [8*14:1] stringvar; initial begin stringvar = "Hello world"; $display("%s is stored as %h", stringvar, stringvar); stringvar = {stringvar,"!!!"}; $display("%s is stored as %h", stringvar, stringvar); end endmodule Copyright © 2001 IEEE. All rights reserved. 55 IEEE Std 1364-2001 Version C The result of simulating the above description is IEEE STANDARD VERILOG® Hello world is stored as 00000048656c6c6f20776f726c64 Hello world!!! is stored as 48656c6c6f20776f726c64212121 4.2.3.1 String operations The common string operations copy, concatenate, and compare are supported by Verilog HDL operators. Copy is provided by simple assignment. Concatenation is provided by the concatenation operator. Comparison is provided by the equality operators. When manipulating string values in vector regs, the regs should be at least 8*n bits (where n is the number of ASCII characters) in order to preserve the 8-bit ASCII code. 4.2.3.2 String value padding and potential problems When strings are assigned to variables, the values stored shall be padded on the left with zeros. Padding can affect the results of comparison and concatenation operations. The comparison and concatenation operators shall not distinguish between zeros resulting from padding and the original string characters (\0, ASCII NULL). Examples: The following example illustrates the potential problem. reg [8*10:1] s1, s2; initial begin s1 = "Hello"; s2 = " world!"; if ({s1,s2} == "Hello world!") $display("strings are equal"); end The comparison in this example fails because during the assignment the string variables are padded as illustrated in the next example: s1 = 000000000048656c6c6f s2 = 00000020776f726c6421 The concatenation of s1 and s2 includes the zero padding, resulting in the following value: 000000000048656c6c6f00000020776f726c6421 56 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Since the string “Hello world!” contains no zero padding, the comparison fails, as shown in the following example: s1 s2 000000000048656c6c6f00000020776f726c6421 48656c6c6f20776f726c6421 "Hello" " world!" This comparison yields a result of zero, which is equivalent to false. 4.2.3.3 Null string handling The null string ("") shall be considered equivalent to the ASCII NULL ("\0") which has a value zero (0), which is different from a string "0". 4.3 Minimum, typical, and maximum delay expressions Verilog HDL delay expressions can be specified as three expressions separated by colons and enclosed by parentheses. This is intended to represent minimum, typical, and maximum values—in that order. The syntax is given in Syntax 4-2. Copyright © 2001 IEEE. All rights reserved. 57 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® constant_expression ::= (From Annex A - A.8.3) constant_primary | unary_operator { attribute_instance } constant_primary | constant_expression binary_operator { attribute_instance } constant_expression | constant_expression ? { attribute_instance } constant_expression constant_expression | string constant_mintypmax_expression ::= constant_expression | constant_expression : constant_expression : constant_expression expression ::= primary | unary_operator { attribute_instance } primary | expression binary_operator { attribute_instance } expression | conditional_expression | string mintypmax_expression ::= expression | expression : expression : expression constant_primary ::= (From Annex A - A.8.4) constant_concatenation | constant_function_call | ( constant_mintypmax_expression ) | constant_multiple_concatenation | genvar_identifier | number | parameter_identifier | specparam_identifier primary ::= number | hierarchical_identifier | hierarchical_identifier [ expression ] { [ expression ] } | hierarchical_identifier [ expression ] { [ expression ] } [ range_expression ] | hierarchical_identifier [ range_expression ] | concatenation | multiple_concatenation | function_call | system_function_call | constant_function_call | ( mintypmax_expression ) Syntax 4-2—Syntax for mintypmax expression Verilog HDL models typically specify three values for delay expressions. The three values allow a design to be tested with minimum, typical, or maximum delay values. Values expressed in min:typ:max format can be used in expressions. The min:typ:max format can be used wherever expressions can appear. 58 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Examples: Example 1—This example shows an expression that defines a single triplet of delay values. The minimum value is the sum of a+d; the typical value is b+e; the maximum value is c+f, as follows: (a:b:c) + (d:e:f) Example 2—The next example shows a typical expression that is used to specify min:typ:max format values: val - (32’d 50: 32’d 75: 32’d 100) 4.4 Expression bit lengths Controlling the number of bits that are used in expression evaluations is important if consistent results are to be achieved. Some situations have a simple solution; for example, if a bit-wise and operation is specified on two 16-bit regs, then the result is a 16-bit value. However, in some situations it is not obvious how many bits are used to evaluate an expression, or what size the result should be. For example, should an arithmetic add of two 16-bit values perform the evaluation using 16 bits, or should the evaluation use 17 bits in order to allow for a possible carry overflow? The answer depends on the type of device being modeled, and whether that device handles carry overflow. The Verilog HDL uses the bit length of the operands to determine how many bits to use while evaluating an expression. The bit length rules are given in 4.4.1. In the case of the addition operator, the bit length of the largest operand, including the lefthand side of an assignment, shall be used. Examples: reg [15:0] a, b; // 16-bit regs reg [15:0] sumA; // 16-bit reg reg [16:0] sumB; // 17-bit reg sumA = a + b; // expression evaluates using 16 bits sumB = a + b; // expression evaluates using 17 bits 4.4.1 Rules for expression bit lengths The rules governing the expression bit lengths have been formulated so that most practical situations have a natural solution. The number of bits of an expression (known as the size of the expression) shall be determined by the operands involved in the expression and the context in which the expression is given. A self-determined expression is one where the bit length of the expression is solely determined by the expression itself—for example, an expression representing a delay value. A context-determined expression is one where the bit length of the expression is determined by the bit length of the expression and by the fact that it is part of another expression. For example, the bit size of the righthand side expression of an assignment depends on itself and the size of the left-hand side. Table 29 shows how the form of an expression shall determine the bit lengths of the results of the expression. In Table 29, i, j, and k represent expressions of an operand, and L(i) represents the bit length of the operand represented by i. Copyright © 2001 IEEE. All rights reserved. 59 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Table 29—Bit lengths resulting from self-determined expressions Expression Bit length Comments Unsized constant numbera Same as integer Sized constant number As given i op j, where op is: + - * / % & | ^ ^~ ~^ max(L(i),L(j)) op i, where op is: L(i) +-~ i op j, where op is: 1 bit === !== == != && || > >= < <= Operands are sized to max(L(i),L(j)) op i, where op is: 1 bit & ~& | ~| ^ ~^ ^~ ! All operands are self-determined i op j, where op is: L(i) >> << ** >>> <<< j is self-determined i?j:k max(L(j),L(k)) i is self-determined {i,...,j} L(i)+..+L(j) All operands are self-determined {i{j,..,k}} i * (L(j)+..+L(k)) All operands are self-determined aIf an unsized constant is part of an expression that is longer than 32 bits, then if the most significant bit is unknown (X or x) or three-state (Z or z) the most significant bit is extended up to the size of the expression, otherwise signed constants are sign extended and unsigned constants are zero extended. NOTE—Multiplication without losing any overflow bits is still possible simply by assigning the result to something wide enough to hold it. 4.4.2 An example of an expression bit-length problem During the evaluation of an expression, interim results shall take the size of the largest operand (in case of an assignment, this also includes the left-hand side). Care has to be taken to prevent loss of a significant bit during expression evaluation. The example below describes how the bit lengths of the operands could result in the loss of a significant bit. Given the following declarations reg [15:0] a, b, answer; // 16-bit regs The intent is to evaluate the expression answer = (a + b) >> 1; //will not work properly where a and b are to be added, which may result in an overflow, and then shifted right by 1 bit to preserve the carry bit in the 16-bit answer. A problem arises, however, because all operands in the expression are of a 16-bit width. Therefore, the expression (a + b) produces an interim result that is only 16 bits wide, thus losing the carry bit before the evaluation performs the 1-bit right shift operation. 60 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The solution is to force the expression (a + b) to evaluate using at least 17 bits. For example, adding an integer value of 0 to the expression will cause the evaluation to be performed using the bit size of integers. The following example will produce the intended result: answer = (a + b + 0) >> 1; //will work correctly In the following example: module bitlength(); reg [3:0] a,b,c; reg [4:0] d; initial begin a = 9; b = 8; c = 1; $display("answer = %b", c ? (a&b) : d); end endmodule the $display statement will display: answer = 01000 By itself, the expression a&b would have the bit length 4, but since it is in the context of the conditional expression, which uses the maximum bit-length, the expression a&b actually has length 5, the length of d. 4.4.3 Example of self-determined expressions reg [3:0] a; reg [5:0] b; reg [15:0] c; initial begin a = 4’hF; b = 6’hA; $display("a*b=%h", a*b);// expression size is self-determined c = {a**b}; // expression a**b is self-determined // due to concatenation operator {} $display("a**b=%h", c); c = a**b; // expression size is determined by c $display("c=%h", c); end Simulator output for this example: a*b=16 // ’h96 was truncated to ’h16 since expression size is 6 a**b=1 // expression size is 4 bits (size of a) c=ac61 // expression size is 16 bits (size of c) Copyright © 2001 IEEE. All rights reserved. 61 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 4.5 Signed expressions Controlling the sign of an expression is important if consistent results are to be achieved. In addition to the rules outlined in the following sections, two system functions shall be used to handle type casting on expressions: $signed() and $unsigned(). These functions shall evaluate the input expression and return a value with the same size and value of the input expression and the type defined by the function: $signed - returned value is signed $unsigned - returned value is unsigned Example: reg [7:0] regA, regB; reg signed [7:0] regS; regA = $unsigned(-4); // regA = 8'b11111100 regB = $unsigned(-4'sd4); // regB = 8'b00001100 regS = $signed (4'b1100); // regS = -4 4.5.1 Rules for expression types The following are the rules for determining the resulting type of an expression: — Expression type depends only on the operands. It does not depend on the LHS (if any). — Decimal numbers are signed. — Based_numbers are unsigned, except where the s notation is used in the base specifier (as in "4'sd12"). — Bit-select results are unsigned, regardless of the operands. — Part-select results are unsigned, regardless of the operands. NOTE—This is true even if the part-select specifies the entire vector. reg [15:0] a; reg signed [7:0] b; initial a = b[7:0]; // b[7:0] is unsigned and therefore zero-extended — Concatenate results are unsigned, regardless of the operands. — Comparison results (1, 0) are unsigned, regardless of the operands. — Reals converted to integers by type coercion are signed — The sign and size of any self-determined operand is determined by the operand itself and indepen- dent of the remainder of the expression. — For non-self-determined operands the following rules apply: if any operand is real, the result is real; if any operand is unsigned, the result is unsigned, regardless of the operator; if all operands are signed, the result will be signed, regardless of operator, except as noted. 4.5.2 Steps for evaluating an expression — Determine the expression size based upon the standard rules of expression size determination. — Determine the sign of the expression using the rules outlined in 4.5.1. — Coerce the type of each operand of the expression (excepting those which are self-determined) to the type of the expression. 62 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C — Extend the size of each operand (excepting those which are self-determined) to the size of the expression. Perform sign extension if and only if the operand type (after type coercion) is signed. 4.5.3 Steps for evaluating an assignment — Determine the size of the RHS by the standard assignment size determination rules (see 4.4) — If needed, extend the size of the RHS, performing sign extension if and only if the type of the RHS is signed. 4.5.4 Handling X and Z in signed expressions If a signed operand is to be resized to a larger signed width and the value of the sign bit is X, the resulting value shall be bit-filled with Xs. If the sign bit of the value is Z, then the resulting value shall be bit-filled with Zs. If any bit of a signed value is X or Z, then any non logical operation involving the value shall result in the entire resultant value being an X and the type consistent with the expression's type. Copyright © 2001 IEEE. All rights reserved. 63 IEEE Std 1364-2001 Version C 5. Scheduling semantics IEEE STANDARD VERILOG® 5.1 Execution of a model The balance of the sections of this standard describe the behavior of each of the elements of the language. This section gives an overview of the interactions between these elements, especially with respect to the scheduling and execution of events. The elements that make up the Verilog HDL can be used to describe the behavior, at varying levels of abstraction, of electronic hardware. An HDL has to be a parallel programming language. The execution of certain language constructs is defined by parallel execution of blocks or processes. It is important to understand what execution order is guaranteed to the user, and what execution order is indeterminate. Although the Verilog HDL is used for more than simulation, the semantics of the language are defined for simulation, and everything else is abstracted from this base definition. 5.2 Event simulation The Verilog HDL is defined in terms of a discrete event execution model. The discrete event simulation is described in more detail in this section to provide a context to describe the meaning and valid interpretation of Verilog HDL constructs. These resulting definitions provide the standard Verilog reference model for simulation, which all compliant simulators shall implement. Note, though, that there is a great deal of choice in the definitions that follow, and differences in some details of execution are to be expected between different simulators. In addition, Verilog HDL simulators are free to use different algorithms than those described in this section, provided the user-visible effect is consistent with the reference model. A design consists of connected threads of execution or processes. Processes are objects that can be evaluated, that may have state, and that can respond to changes on their inputs to produce outputs. Processes include primitives, modules, initial and always procedural blocks, continuous assignments, asynchronous tasks, and procedural assignment statements. Every change in value of a net or variable in the circuit being simulated, as well as the named event, is considered an update event. Processes are sensitive to update events. When an update event is executed, all the processes that are sensitive to that event are evaluated in an arbitrary order. The evaluation of a process is also an event, known as an evaluation event. In addition to events, another key aspect of a simulator is time. The term simulation time is used to refer to the time value maintained by the simulator to model the actual time it would take for the circuit being simulated. The term time is used interchangeably with simulation time in this section. Events can occur at different times. In order to keep track of the events and to make sure they are processed in the correct order, the events are kept on an event queue, ordered by simulation time. Putting an event on the queue is called scheduling an event. 5.3 The stratified event queue The Verilog event queue is logically segmented into five different regions. Events are added to any of the five regions but are only removed from the active region. 1) Events that occur at the current simulation time and can be processed in any order. These are the active events. 2) Events that occur at the current simulation time, but that shall be processed after all the active events are processed. These are the inactive events. 64 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 3) Events that have been evaluated during some previous simulation time, but that shall be assigned at this simulation time after all the active and inactive events are processed. These are the nonblocking assign update events. 4) Events that shall be processed after all the active, inactive, and nonblocking assign update events are processed. These are the monitor events. 5) Events that occur at some future simulation time. These are the future events. Future events are divided into future inactive events, and future nonblocking assignment update events. The processing of all the active events is called a simulation cycle. The freedom to choose any active event for immediate processing is an essential source of nondeterminism in the Verilog HDL. An explicit zero delay (#0) requires that the process be suspended and added as an inactive event for the current time so that the process is resumed in the next simulation cycle in the current time. A nonblocking assignment (see 9.2.2) creates a nonblocking assign update event, scheduled for current or a later simulation time. The $monitor and $strobe system tasks (see 17.1) create monitor events for their arguments. These events are continuously re-enabled in every successive time step. The monitor events are unique in that they cannot create any other events. The call back procedures scheduled with PLI routines such as tf_synchronize() (see 25.58) or vpi_register_cb(cb_readwrite) (see 27.33) shall be treated as inactive events. 5.4 The Verilog simulation reference model In all the examples that follow, T refers to the current simulation time, and all events are held in the event queue, ordered by simulation time. while (there are events) { if (no active events) { if (there are inactive events) { activate all inactive events; } else if (there are nonblocking assign update events) { activate all nonblocking assign update events; } else if (there are monitor events) { activate all monitor events; } else { advance T to the next event time; activate all inactive events for time T; } } E = any active event; if (E is an update event) { update the modified object; add evaluation events for sensitive processes to event queue; } else { /* shall be an evaluation event */ evaluate the process; add update events to the event queue; } } Copyright © 2001 IEEE. All rights reserved. 65 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 5.4.1 Determinism This standard guarantees a certain scheduling order. 1) Statements within a begin-end block shall be executed in the order in which they appear in that begin-end block. Execution of statements in a particular begin-end block can be suspended in favor of other processes in the model; however, in no case shall the statements in a begin-end block be executed in any order other than that in which they appear in the source. 2) Nonblocking assignments shall be performed in the order the statements were executed. Consider the following example: initial begin a <= 0; a <= 1; end When this block is executed, there will be two events added to the nonblocking assign update queue. The previous rule requires that they be entered on the queue in source order; this rule requires that they be taken from the queue and performed in source order as well. Hence, at the end of time step 1, the variable a will be assigned 0 and then 1. 5.4.2 Nondeterminism One source of nondeterminism is the fact that active events can be taken off the queue and processed in any order. Another source of nondeterminism is that statements without time-control constructs in behavioral blocks do not have to be executed as one event. Time control statements are the # expression and @ expression constructs (see 9.7). At any time while evaluating a behavioral statement, the simulator may suspend execution and place the partially completed event as a pending active event on the event queue. The effect of this is to allow the interleaving of process execution. Note that the order of interleaved execution is nondeterministic and not under control of the user. 5.5 Race conditions Because the execution of expression evaluation and net update events may be intermingled, race conditions are possible: assign p = q; initial begin q = 1; #1 q = 0; $display(p); end The simulator is correct in displaying either a 1 or a 0. The assignment of 0 to q enables an update event for p. The simulator may either continue and execute the $display task or execute the update for p, followed by the $display task. 5.6 Scheduling implication of assignments Assignments are translated into processes and events as follows. 66 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 5.6.1 Continuous assignment A continuous assignment statement (Clause 6) corresponds to a process, sensitive to the source elements in the expression. When the value of the expression changes, it causes an active update event to be added to the event queue, using current values to determine the target. 5.6.2 Procedural continuous assignment A procedural continuous assignment (which are the assign or force statement; see 9.3) corresponds to a process that is sensitive to the source elements in the expression. When the value of the expression changes, it causes an active update event to be added to the event queue, using current values to determine the target. A deassign or a release statement deactivates any corresponding assign or force statement(s). 5.6.3 Blocking assignment A blocking assignment statement (see 9.2.1) with a delay computes the right-hand side value using the current values, then causes the executing process to be suspended and scheduled as a future event. If the delay is 0, the process is scheduled as an inactive event for the current time. When the process is returned (or if it returns immediately if no delay is specified), the process performs the assignment to the left-hand side and enables any events based upon the update of the left-hand side. The values at the time the process resumes are used to determine the target(s). Execution may then continue with the next sequential statement or with other active events. 5.6.4 Nonblocking assignment A nonblocking assignment statement (see 9.2.2) always computes the updated value and schedules the update as a nonblocking assign update event, either in this time step if the delay is zero or as a future event if the delay is nonzero. The values in effect when the update is placed on the event queue are used to compute both the right-hand value and the left-hand target. 5.6.5 Switch (transistor) processing The event-driven simulation algorithm described in 5.4 depends on unidirectional signal flow and can process each event independently. The inputs are read, the result is computed, and the update is scheduled. The Verilog HDL provides switch-level modeling in addition to behavioral and gate-level modeling. Switches provide bi-directional signal flow and require coordinated processing of nodes connected by switches. The Verilog HDL source elements that model switches are various forms of transistors, called tran, tranif0, tranif1, rtran, rtranif0, and rtranif1. Switch processing shall consider all the devices in a bidirectional switch-connected net before it can determine the appropriate value for any node on the net, because the inputs and outputs interact. A simulator can do this using a relaxation technique. The simulator can process tran at any time. It can process a subset of tran-connected events at a particular time, intermingled with the execution of other active events. Further refinement is required when some transistors have gate value x. A conceptually simple technique is to solve the network repeatedly with these transistors set to all possible combinations of fully conducting and nonconducting transistors. Any node that has a unique logic level in all cases has steady-state response equal to this level. All other nodes have steady-state response x. Copyright © 2001 IEEE. All rights reserved. 67 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 5.6.6 Port connections Ports connect processes through implicit continuous assignment statements or implicit bidirectional connections. Bidirectional connections are analogous to an always-enabled tran connection between the two nets, but without any strength reduction. Port connection rules require that a value receiver be a net or a structural net expression. Ports can always be represented as declared objects connected as follows: — If an input port, then a continuous assignment from an outside expression to a local (input) net — If an output port, then a continuous assignment from a local output expression to an outside net — If an inout, then a nonstrength-reducing transistor connecting the local net to an outside net 5.6.7 Functions and tasks Task and function parameter passing is by value, and it copies in on invocation and copies out on return. The copy out on the return function behaves in the same manner as does any blocking assignment. 68 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 6. Assignments The assignment is the basic mechanism for placing values into nets and variables. There are two basic forms of assignments: — The continuous assignment, which assigns values to nets — The procedural assignment, which assigns values to variables There are two additional forms of assignments, assign / deassign and force / release which are called procedural continuous assignments, described in 9.3. An assignment consists of two parts, a left-hand side and a right-hand side, separated by the equals ( = ) character; or, in the case of nonblocking procedural assignment, the less-than-equals ( <= ) character pair. The right-hand side can be any expression that evaluates to a value. The left-hand side indicates the variable to which the right-hand side value is to be assigned. The left-hand side can take one of the forms given in Table 30, depending on whether the assignment is a continuous assignment or a procedural assignment. Table 30—Legal left-hand side forms in assignment statements Statement type Left-hand side (LHS) Continuous assignment Procedural assignment Net (vector or scalar) Constant bit select of a vector net Constant part select of a vector net Constant indexed part select of a vector net Concatenation of any of the above four LHS Variables (vector or scalar) Bit-select of a vector reg, integer, or time variable Constant part select of a vector reg, integer, or time variable Memory word Indexed part select of a vector reg, integer, or time variable Concatenation of regs; bit or part selects of regs 6.1 Continuous assignments Continuous assignments shall drive values onto nets, both vector and scalar. This assignment shall occur whenever the value of the right-hand side changes. Continuous assignments provide a way to model combinational logic without specifying an interconnection of gates. Instead, the model specifies the logical expression that drives the net. Copyright © 2001 IEEE. All rights reserved. 69 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® The syntax for continuous assignments is given in Syntax 6-1. net_declaration ::= (From Annex A - A.2.1.3) net_type [ signed ] [ delay3 ] list_of_net_identifiers ; | net_type [ drive_strength ] [ signed ] [ delay3 ] list_of_net_decl_assignments ; | net_type [ vectored | scalared ] [ signed ] range [ delay3 ] list_of_net_identifiers ; | net_type [ drive_strength ] [ vectored | scalared ] [ signed ] range [ delay3 ] list_of_net_decl_assignments ; | trireg [ charge_strength ] [ signed ] [ delay3 ] list_of_net_identifiers ; | trireg [ drive_strength ] [ signed ] [ delay3 ] list_of_net_decl_assignments ; | trireg [ charge_strength ] [ vectored | scalared ] [ signed ] range [ delay3 ] list_of_net_identifiers ; | trireg [ drive_strength ] [ vectored | scalared ] [ signed ] range [ delay3 ] list_of_net_decl_assignments ; list_of_net_decl_assignments ::= (From Annex A - A.2.3) net_decl_assignment { , net_decl_assignment } net_decl_assignment ::= (From Annex A - A.2.4) net_identifier = expression continuous_assign ::= (From Annex A - A.6.1) assign [ drive_strength ] [ delay3 ] list_of_net_assignments ; list_of_net_assignments ::= net_assignment { , net_assignment } net_assignment ::= net_lvalue = expression Syntax 6-1—Syntax for continuous assignment 6.1.1 The net declaration assignment The first two alternatives in the net declaration are discussed in see 3.2. The third alternative, the net declaration assignment, allows a continuous assignment to be placed on a net in the same statement that declares the net. Example: The following is an example of the net declaration form of a continuous assignment: wire (strong1, pull0) mynet = enable ; NOTE—Because a net can be declared only once, only one net declaration assignment can be made for a particular net. This contrasts with the continuous assignment statement; one net can receive multiple assignments of the continuous assignment form. 6.1.2 The continuous assignment statement The continuous assignment statement shall place a continuous assignment on a net data type. The net may be explicitly declared, or may inherit an implicit declaration in accordance with the implicit declarations rules defined in 3.5. 70 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Assignments on nets shall be continuous and automatic. This means that whenever an operand in the righthand side expression changes value, the whole right-hand side shall be evaluated and if the new value is different from the previous value, then the new value shall be assigned to the left-hand side. Examples: Example 1—The following is an example of a continuous assignment to a net that has been previously declared: wire mynet ; assign (strong1, pull0) mynet = enable ; Example 2—The following is an example of the use of a continuous assignment to model a 4-bit adder with carry. The assignment could not be specified directly in the declaration of the nets because it requires a concatenation on the left-hand side. module adder (sum_out, carry_out, carry_in, ina, inb); output [3:0] sum_out; output carry_out; input [3:0] ina, inb; input carry_in; wire carry_out, carry_in; wire [3:0] sum_out, ina, inb; assign {carry_out, sum_out} = ina + inb + carry_in; endmodule Example 3—The following example describes a module with one 16-bit output bus. It selects between one of four input busses and connects the selected bus to the output bus. module select_bus(busout, bus0, bus1, bus2, bus3, enable, s); parameter n = 16; parameter Zee = 16’bz; output [1:n] busout; input [1:n] bus0, bus1, bus2, bus3; input enable; input [1:2] s; tri [1:n] data; // net declaration // net declaration with continuous assignment tri [1:n] busout = enable ? data : Zee; // assignment statement with four continuous assignments assign data = (s == 0) ? bus0 : Zee, data = (s == 1) ? bus1 : Zee, data = (s == 2) ? bus2 : Zee, data = (s == 3) ? bus3 : Zee; endmodule The following sequence of events is experienced during simulation of this example: a) The value of s, a bus selector input variable, is checked in the assign statement. Based on the value of s, the net data receives the data from one of the four input buses. b) The setting of data net triggers the continuous assignment in the net declaration for busout. If enable is set, the contents of data are assigned to busout; if enable is 0, the contents of Zee are assigned to busout. Copyright © 2001 IEEE. All rights reserved. 71 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 6.1.3 Delays A delay given to a continuous assignment shall specify the time duration between a right-hand side operand value change and the assignment made to the left-hand side. If the left-hand side references a scalar net, then the delay shall be treated in the same way as for gate delays—that is, different delays can be given for the output rising, falling, and changing to high impedance (see Clause 7). If the left-hand side references a vector net, then up to three delays can be applied. The following rules determine which delay controls the assignment: — If the right-hand side makes a transition from nonzero to zero, then the falling delay shall be used. — If the right-hand side makes a transition to z, then the turn-off delay shall be used. — For all other cases, the rising delay shall be used. Specifying the delay in a continuous assignment that is part of the net declaration shall be treated differently from specifying a net delay and then making a continuous assignment to the net. A delay value can be applied to a net in a net declaration, as in the following example: wire #10 wireA; This syntax, called a net delay, means that any value change that is to be applied to wireA by some other statement shall be delayed for ten time units before it takes effect. When there is a continuous assignment in a declaration, the delay is part of the continuous assignment and is not a net delay. Thus, it shall not be added to the delay of other drivers on the net. Furthermore, if the assignment is to a vector net, then the rising and falling delays shall not be applied to the individual bits if the assignment is included in the declaration. In situations where a right-hand side operand changes before a previous change has had time to propagate to the left-hand side, then the following steps are taken: a) The value of the right-hand side expression is evaluated. b) If this RHS value differs from the value currently scheduled to propagate to the left-hand side, then the currently scheduled propagation event is descheduled. c) If the new RHS value equals the current left-hand side value, no event is scheduled. d) If the new RHS value differs from the current LHS value, a delay is calculated in the standard way using the current value of the left-hand side, the newly calculated value of the right-hand side, and the delays indicated on the statement; a new propagation event is then scheduled to occur delay time units in the future. 6.1.4 Strength The driving strength of a continuous assignment can be specified by the user. This applies only to assignments to scalar nets of the following types: wire wand wor tri triand trior trireg tri0 tri1 Continuous assignments driving strengths can be specified in either a net declaration or in a stand-alone assignment, using the assign keyword. The strength specification, if provided, shall immediately follow the keyword (either the keyword for the net type or assign) and precede any delay specified. Whenever the continuous assignment drives the net, the strength of the value shall be simulated as specified. 72 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C A drive strength specification shall contain one strength value that applies when the value being assigned to the net is 1 and a second strength value that applies when the assigned value is 0. The following keywords shall specify the strength value for an assignment of 1: supply1 strong1 pull1 weak1 highz1 The following keywords shall specify the strength value for an assignment of 0: supply0 strong0 pull0 weak0 highz0 The order of the two strength specifications shall be arbitrary. The following two rules shall constrain the use of drive strength specifications: — The strength specifications (highz1, highz0) and (highz0, highz1) shall be treated as illegal constructs. — If drive strength is not specified, it shall default to (strong1, strong0). 6.2 Procedural assignments The primary discussion of procedural assignments is in 9.2. However, a description of the basic ideas in this clause highlights the differences between continuous assignments and procedural assignments. As stated in 6.1, continuous assignments drive nets in a manner similar to the way gates drive nets. The expression on the right-hand side can be thought of as a combinatorial circuit that drives the net continuously. In contrast, procedural assignments put values in variables. The assignment does not have duration; instead, the variable holds the value of the assignment until the next procedural assignment to that variable. Procedural assignments occur within procedures such as always, initial (see 9.9), task, and function (see Clause 10) and can be thought of as “triggered” assignments. The trigger occurs when the flow of execution in the simulation reaches an assignment within a procedure. Reaching the assignment can be controlled by conditional statements. Event controls, delay controls, if statements, case statements, and looping statements can all be used to control whether assignments are evaluated. Clause 9 gives details and examples. 6.2.1 Variable declaration assignment The variable declaration assignment is a special case of procedural assignment as it assigns a value to a variable. It allows an initial value to be placed in a variable in the same statement that declares the variable. The assignment shall be to a constant expression. The assignment does not have duration; instead, the variable holds the value until the next assignment to that variable. Variable declaration assignments to an array are not allowed. Variable declaration assignments are only allowed at the module level. Examples: Example 1—Declare a 4 bit reg and assign it the value 4. reg[3:0] a = 4'h4; This is equivalent to writing: reg[3:0] a; initial a = 4'h4; Example 2—The following example is not legal. reg [3:0] array [3:0] = 0; Copyright © 2001 IEEE. All rights reserved. 73 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Example 3—Declare two integers, the first is assigned the value of 0. integer i = 0, j; Example 4—Declare two real variables, assigned to the values 2.5 and 300,000. real r1 = 2.5, n300k = 3E6; Example 5—Declare a time variable and realtime variable with initial values. time t1 = 25; realtime rt1 = 2.5; NOTE—If the same variable is assigned different values both in an initial block and in a variable declaration assignment, the order of the evaluation is undefined. 6.2.2 Variable declaration syntax The syntax for variable declaration assignments is given in Syntax 6-2. integer_declaration ::= (From Annex A - A.2.1.3) integer list_of_variable_identifiers ; real_declaration ::= real list_of_real_identifiers ; realtime_declaration ::= realtime list_of_real_identifiers ; reg_declaration ::= reg [ signed ] [ range ] list_of_variable_identifiers ; time_declaration ::= time list_of_variable_identifiers ; real_type ::= (From Annex A - A.2.2.1) real_identifier [ = constant_expression ] | real_identifier dimension { dimension } variable_type ::= variable_identifier [ = constant_expression ] | variable_identifier dimension { dimension } list_of_real_identifiers ::= (From Annex A - A.2.3) real_type { , real_type } list_of_variable_identifiers ::= variable_type { , variable_type } Syntax 6-2—Syntax for reg declaration assignment 74 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 7. Gate and switch level modeling This clause describes the syntax and semantics of these built-in primitives and how a hardware design can be described using these primitives. There are 14 logic gates and 12 switches predefined in the Verilog HDL to provide the gate and switch level modeling facility. Modeling with logic gates and switches has the following advantages: — Gates provide a much closer one-to-one mapping between the actual circuit and the model. — There is no continuous assignment equivalent to the bidirectional transfer gate. 7.1 Gate and switch declaration syntax Syntax 7-1 shows the gate and switch declaration syntax. A gate or a switch instance declaration shall have the following specifications: — The keyword that names the type of gate or switch primitive — An optional drive strength — An optional propagation delay — An optional identifier that names each gate or switch instance — An optional range for array of instances — The terminal connection list Multiple instances of the one type of gate or switch primitive can be declared as a comma-separated list. All such instances shall have the same drive strength and delay specification. Copyright © 2001 IEEE. All rights reserved. 75 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® gate_instantiation ::= (From Annex A - A.3.1) cmos_switchtype [delay3] cmos_switch_instance { , cmos_switch_instance } ; | enable_gatetype [drive_strength] [delay3] enable_gate_instance { , enable_gate_instance } ; | mos_switchtype [delay3] mos_switch_instance { , mos_switch_instance } ; | n_input_gatetype [drive_strength] [delay2] n_input_gate_instance {, n_input_gate_instance }; | n_output_gatetype [drive_strength] [delay2] n_output_gate_instance { , n_output_gate_instance } ; | pass_en_switchtype [delay2] pass_enable_switch_instance {, pass_enable_switch_instance } ; | pass_switchtype pass_switch_instance { , pass_switch_instance } ; | pulldown [pulldown_strength] pull_gate_instance { , pull_gate_instance } ; | pullup [pullup_strength] pull_gate_instance { , pull_gate_instance } ; cmos_switch_instance ::= [ name_of_gate_instance ] ( output_terminal , input_terminal , ncontrol_terminal , pcontrol_terminal ) enable_gate_instance ::= [ name_of_gate_instance ] ( output_terminal , input_terminal , enable_terminal ) mos_switch_instance ::= [ name_of_gate_instance ] ( output_terminal , input_terminal , enable_terminal ) n_input_gate_instance ::= [ name_of_gate_instance ] ( output_terminal , input_terminal { , input_terminal } ) n_output_gate_instance ::= [ name_of_gate_instance ] ( output_terminal { , output_terminal } , input_terminal ) pass_switch_instance ::= [ name_of_gate_instance ] ( inout_terminal , inout_terminal ) pass_enable_switch_instance ::= [ name_of_gate_instance ] ( inout_terminal , inout_terminal , enable_terminal ) pull_gate_instance ::= [ name_of_gate_instance ] ( output_terminal ) name_of_gate_instance ::= gate_instance_identifier [ range ] pulldown_strength ::= (From Annex A - A.3.2) ( strength0 , strength1 ) | ( strength1 , strength0 ) | ( strength0 ) pullup_strength ::= ( strength0 , strength1 ) | ( strength1 , strength0 ) | ( strength1 ) enable_terminal ::= (From Annex A - A.3.3) expression inout_terminal ::= net_lvalue input_terminal ::= expression ncontrol_terminal ::= expression output_terminal ::= net_lvalue pcontrol_terminal ::= expression cmos_switchtype ::= (From Annex A - A.3.4) cmos | rcmos enable_gatetype ::= bufif0 | bufif1 | notif0 | notif1 mos_switchtype ::= nmos | pmos | rnmos | rpmos n_input_gatetype ::= and | nand | or | nor | xor | xnor n_output_gatetype ::= buf | not pass_en_switchtype ::= tranif0 | tranif1 | rtranif1 | rtranif0 pass_switchtype ::= tran | rtran Syntax 7-1—Syntax for gate instantiation 76 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 7.1.1 The gate type specification A gate or switch instance declaration shall begin with the keyword that specifies the gate or switch primitive being used by the instances that follow in the declaration. Table 31 lists the keywords that shall begin a gate or a switch instance declaration. Table 31—Built-in gates and switches n_input gates n_output gates three-state gates and buf bufif0 nand not bufif1 nor notif0 or notif1 xnor xor pull gates pulldown pullup MOS switches bidirectional switches cmos rtran nmos rtranif0 pmos rtranif1 rcmos rnmos rpmos tran tranif0 tranif1 Explanations of the built-in gates and switches shown in Table 31 begin in 7.2. 7.1.2 The drive strength specification An optional drive strength specification shall specify the strength of the logic values on the output terminals of the gate instance. Only the instances of the gate primitives shown in Table 32 can have the drive strength specification. Table 32—Valid gate types for strength specifications and nand buf not pulldown or nor bufif0 notif0 pullup xor xnor bufif1 notif1 The drive strength specification for a gate instance, with the exception of pullup and pulldown, shall have a strength1 specification and a strength0 specification. The strength1 specification shall specify the strength of signals with a logic value 1, and the strength0 specification shall specify the strength of signals with a logic value 0. The strength specification shall follow the gate type keyword and precede any delay specification. The strength0 specification can precede or follow the strength1 specification. The strength1 and strength0 specifications shall be separated by a comma and enclosed within a pair of parentheses. The pullup gate can have only strength1 specification; strength0 specification shall be optional. The pulldown gate can have only strength0 specification; strength1 specification shall be optional. The strength1 specification shall be one of the following keywords: supply1 strong1 pull1 weak1 The strength0 specification shall be one of the following keywords: supply0 strong0 pull0 weak0 Specifying highz1 as strength1 shall cause the gate or switch to output a logic value z in place of a 1. Specifying highz0 shall cause the gate to output a logic value z in place of a 0. The strength specifications Copyright © 2001 IEEE. All rights reserved. 77 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® (highz0, highz1) and (highz1, highz0) shall be considered invalid. In the absence of a strength specification, the instances shall have the default strengths strong1 and strong0. Example: The following example shows a drive strength specification in a declaration of an open collector nor gate: nor (highz1,strong0) n1(out1,in1,in2); In this example, the nor gate outputs a z in place of a 1. Logic strength modeling is discussed in more detail in 7.9 through 7.13. 7.1.3 The delay specification An optional delay specification shall specify the propagation delay through the gates and switches in a declaration. Gates and switches in declarations with no delay specification shall have no propagation delay. A delay specification can contain up to three delay values, depending on the gate type. The pullup and pulldown instance declarations shall not include delay specifications. Delays are discussed in more detail in 7.14. 7.1.4 The primitive instance identifier An optional name can be given to a gate or switch instance. If multiple instances are declared as an array of instances, an identifier shall be used to name the instances. 7.1.5 The range specification There are many situations when repetitive instances are required. These instances shall differ from each other only by the index of the vector to which they are connected. In order to specify an array of instances, the instance name shall be followed by the range specification. The range shall be specified by two constant expressions, left-hand index (lhi) and right-hand index (rhi), separated by a colon and enclosed within a pair of square brackets. A [lhi:rhi] range specification shall represent an array of abs(lhi-rhi)+1 instances. Neither of the two constant expressions are required to be zero, and lhi is not required to be larger than rhi. If both constant expressions are equal, only one instance shall be generated. An array of instances shall have a continuous range. One instance identifier shall be associated with only one range to declare an array of instances. The range specification shall be optional. If no range specification is given, a single instance shall be created. Example: A declaration shown below is illegal: nand #2 t_nand[0:3] ( ... ), t_nand[4:7] ( ... ); It could be declared correctly as one array of eight instances, or two arrays with unique names of four elements each: nand #2 t_nand[0:7]( ... ); nand #2 x_nand[0:3] ( ... ), y_nand[4:7] ( ... ); 78 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 7.1.6 Primitive instance connection list The terminal list describes how the gate or switch connects to the rest of the model. The gate or switch type can limit these expressions. The connection list shall be enclosed in a pair of parentheses, and the terminals shall be separated by commas. The output or bidirectional terminals shall always come first in the terminal list, followed by the input terminals. The terminal connections for an array of instances shall follow these rules: — The bit length of each port expression in the declared instance-array shall be compared with the bit length of each single-instance port or terminal in the instantiated module or primitive. — For each port or terminal where the bit length of the instance-array port expression is the same as the bit length of the single-instance port, the instance-array port expression shall be connected to each single-instance port. — If bit lengths are different, each instance shall get a part-select of the port expression as specified in the range, starting with the right-hand index. — Too many or too few bits to connect to all the instances shall be considered an error. An individual instance from an array of instances shall be referenced in the same manner as referencing an element of an array of regs. Examples: Example 1—The following declaration of nand_array declares four instances that can be referenced by nand_array[1], nand_array[2], nand_array[3], and nand_array[4] respectively. nand #2 nand_array[1:4]( ... ) ; Example 2—The two module descriptions that follow are equivalent except for indexed instance names, and they demonstrate the range specification and connection rules for declaring an array of instances: module driver (in, out, en); input [3:0] in; output [3:0] out; input en; bufif0 ar[3:0] (out, in, en); // array of three-state buffers endmodule module driver_equiv (in, out, en); input [3:0] in; output [3:0] out; input en; bufif0 ar3 (out[3], in[3], en); // each buffer declared separately bufif0 ar2 (out[2], in[2], en); bufif0 ar1 (out[1], in[1], en); bufif0 ar0 (out[0], in[0], en); endmodule Copyright © 2001 IEEE. All rights reserved. 79 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Example 3—The two module descriptions that follow are equivalent except for indexed instance names, and they demonstrate how different instances within an array of instances are connected when the port sizes do not match. module busdriver (busin, bushigh, buslow, enh, enl); input [15:0] busin; output [7:0] bushigh, buslow; input enh, enl; driver busar3 (busin[15:12], bushigh[7:4], enh); driver busar2 (busin[11:8], bushigh[3:0], enh); driver busar1 (busin[7:4], buslow[7:4], enl); driver busar0 (busin[3:0], buslow[3:0], enl); endmodule module busdriver_equiv (busin, bushigh, buslow, enh, enl); input [15:0] busin; output [7:0] bushigh, buslow; input enh, enl; driver busar[3:0] (.out({bushigh, buslow}), .in(busin), .en({enh, enh, enl, enl})); endmodule Example 4—This example demonstrates how a series of modules can be chained together. Figure 4 shows an equivalent schematic interconnection of DFF instances. module dffn (q, d, clk); parameter bits = 1; input [bits-1:0] d; output [bits-1:0] q; input clk ; DFF dff[bits-1:0] (q, d, clk); // create a row of D flip-flops endmodule module MxN_pipeline (in, out, clk); parameter M = 3, N = 4; // M=width,N=depth input [M-1:0] in; output [M-1:0] out; input clk; wire [M*(N-1):1] t; // #(M) redefines the bits parameter for dffn // create p[1:N] columns of dffn rows (pipeline) dffn #(M) p[1:N] ({out, t}, {t, in}, clk); endmodule 80 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C p[4] p[3] p[2] p[1] t[3] t[6] t[9] out[2] in[2] dff[2] dff[2] dff[2] dff[2] in[2:0] in[1] clk t[2] t[5] dff[1] dff[1] t[8] dff[1] dff[1] out[2:0] out[1] t[1] t[4] in[0] dff[0] dff[0] t[7] dff[0] out[0] dff[0] Figure 4—Schematic diagram of interconnections in array of instances 7.2 and, nand, nor, or, xor, and xnor gates The instance declaration of a multiple input logic gate shall begin with one of the following keywords: and nand nor or xor xnor The delay specification shall be zero, one, or two delays. If the specification contains two delays, the first delay shall determine the output rise delay, the second delay shall determine the output fall delay, and the smaller of the two delays shall apply to output transitions to x. If only one delay is specified, it shall specify both the rise delay and the fall delay. If there is no delay specification, there shall be no propagation delay through the gate. These six logic gates shall have one output and one or more inputs. The first terminal in the terminal list shall connect to the output of the gate and all other terminals connect to its inputs. The truth tables for these gates, showing the result of two input values, appear in Table 33. Copyright © 2001 IEEE. All rights reserved. 81 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Table 33—Truth tables for multiple input logic gates and 0 1 x z 0 0000 1 01xx x 0xxx z 0xxx or 0 1 x z 0 01xx 1 1111 x x1xx z x1xx xor 0 1 x z 0 01xx 1 10xx x xxxx z xxxx nand 0 1 x z 0 1111 1 10xx x 1xxx z 1xxx nor 0 1 x z 0 10xx 1 0000 x x0xx z x0xx xnor 0 1 x z 0 10xx 1 01xx x xxxx z xxxx Versions of these six logic gates having more than two inputs shall have a natural extension, but the number of inputs shall not alter propagation delays. Example: The following example declares a two input and gate: and a1 (out, in1, in2); The inputs are in1 and in2. The output is out. The instance name is a1. 7.3 buf and not gates The instance declaration of a multiple output logic gate shall begin with one of the following keywords: buf not The delay specification shall be zero, one, or two delays. If the specification contains two delays, the first delay shall determine the output rise delay, the second delay shall determine the output fall delay, and the smaller of the two delays shall apply to output transitions to x. If only one delay is specified, it shall specify both the rise delay and the fall delay. If there is no delay specification, there shall be no propagation delay through the gate. 82 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C These two logic gates shall have one input and one or more outputs. The last terminal in the terminal list shall connect to the input of the logic gate, and the other terminals shall connect to the outputs of the logic gate. Truth tables for these logic gates with one input and one output are shown in Table 34. Table 34—Truth tables for multiple output logic gates buf input output 0 0 1 1 x x z x not input output 0 1 1 0 x x z x Example: The following example declares a two output buf: buf b1 (out1, out2, in); The input is in. The outputs are out1 and out2. The instance name is b1. 7.4 bufif1, bufif0, notif1, and notif0 gates The instance declaration of these three-state logic gates shall begin with one of the following keywords: bufif0 bufif1 notif1 notif0 These four logic gates model three-state drivers. In addition to logic values 1 and 0, these gates can output z. The delay specification shall be zero, one, two, or three delays. If the delay specification contains three delays, the first delay shall determine the rise delay, the second delay shall determine the fall delay, the third delay shall determine the delay of transitions to z, and the smallest of the three delays shall determine the delay of transitions to x. If the specification contains two delays, the first delay shall determine the output rise delay, the second delay shall determine the output fall delay, and the smaller of the two delays shall apply to output transitions to x and z. If only one delay is specified, it shall specify the delay for all output transitions. If there is no delay specification, there shall be no propagation delay through the gate. Some combinations of data input values and control input values can cause these gates to output either of two values, without a preference for either value (see 7.10.2). These logic tables for these gates include two symbols representing such unknown results. The symbol L shall represent a result that has a value 0 or z. The symbol H shall represent a result that has a value 1 or z. Delays on transitions to H or L shall be treated the same as delays on transitions to x. Copyright © 2001 IEEE. All rights reserved. 83 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® These four logic gates shall have one output, one data input, and one control input. The first terminal in the terminal list shall connect to the output, the second terminal shall connect to the data input, and the third terminal shall connect to the control input. Table 35 presents the logic tables for these gates. Table 35—Truth tables for three-state logic gates bufif0 CONTROL 01xz D 0 0 z LL A 1 1 z HH T xxzxx A zxzxx bufif1 CONTROL 01xz D 0 z 0 LL A 1 z 1 HH T xzxxx A zzxxx notif0 CONTROL 01xz D 0 1 z HH A 1 0 z LL T xxzxx A zxzxx notif1 CONTROL 01xz D 0 z 1 HH A 1 z 0 LL T xzxxx A zzxxx Example: The following example declares an instance of bufif1: bufif1 bf1 (outw, inw, controlw); The output is outw, the input is inw, and the control is controlw. The instance name is bf1. 7.5 MOS switches The instance declaration of a MOS switch shall begin with one of the following keywords: cmos nmos pmos rcmos rnmos rpmos The cmos and rcmos switches are described in 7.7. The pmos keyword stands for the P-type metal-oxide semiconductor (PMOS) transistor and the nmos keyword stands for the N-type metal-oxide semiconductor (NMOS) transistor. PMOS and NMOS transistors have relatively low impedance between their sources and drains when they conduct. The rpmos keyword stands for resistive PMOS transistor and the rnmos keyword stands for resistive NMOS transistor. Resistive 84 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C PMOS and resistive NMOS transistors have significantly higher impedance between their sources and drains when they conduct than PMOS and NMOS transistors have. The load devices in static MOS networks are examples of rpmos and rnmos transistors. These four switches are unidirectional channels for data similar to the bufif gates. The delay specification shall be zero, one, two, or three delays. If the delay specification contains three delays, the first delay shall determine the rise delay, the second delay shall determine the fall delay, the third delay shall determine the delay of transitions to z, and the smallest of the three delays shall determine the delay of transitions to x. If the specification contains two delays, the first delay shall determine the output rise delay, the second delay shall determine the output fall delay, and the smaller of the two delays shall apply to output transitions to x and z. If only one delay is specified, it shall specify the delay for all output transitions. If there is no delay specification, there shall be no propagation delay through the switch. Some combinations of data input values and control input values can cause these switches to output either of two values, without a preference for either value. The logic tables for these switches include two symbols representing such unknown results. The symbol L represents a result that has a value 0 or z. The symbol H represents a result that has a value 1 or z. Delays on transitions to H and L shall be the same as delays on transitions to x. These four switches shall have one output, one data input, and one control input. The first terminal in the terminal list shall connect to the output, the second terminal shall connect to the data input, and the third terminal shall connect to the control input. The nmos and pmos switches shall pass signals from their inputs and through their outputs with a change in the strength of the signal in only one case, as discussed in 7.11. The rnmos and rpmos switches shall reduce the strength of signals that propagate through them, as discussed in 7.12. Table 36 presents the logic tables for these switches. Table 36—Truth tables for MOS switches pmos rpmos CONTROL 01x z D 0 0 z LL A 1 1 z HH T xxzxx A zzzzz nmos rnmos CONTROL 01xz D 0 z 0 LL A 1 z 1 HH T xzxxx A zzzzz Example: The following example declares a pmos switch: pmos p1 (out, data, control); The output is out, the data input is data, and the control input is control. The instance name is p1. Copyright © 2001 IEEE. All rights reserved. 85 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 7.6 Bidirectional pass switches The instance declaration of a bidirectional pass switch shall begin with one of the following keywords: tran rtran tranif1 rtranif1 tranif0 rtranif0 The bidirectional pass switches shall not delay signals propagating through them. When tranif0, tranif1, rtranif0, or rtranif1 devices are turned off they shall block signals, and when they are turned on they shall pass signals. The tran and rtran devices cannot be turned off, and they shall always pass signals. The delay specifications for tranif1, tranif0, rtranif1, and rtranif0 devices shall be zero, one, or two delays. If the specification contains two delays, the first delay shall determine the turn-on delay, and the second delay shall determine the turn-off delay, and the smaller of the two delays shall apply to output transitions to x and z. If only one delay is specified, it shall specify both the turn-on and the turn-off delays. If there is no delay specification, there shall be no turn-on or turn-off delay for the bidirectional pass switch. The bidirectional pass switches tran and rtran shall not accept delay specification. The tranif1, tranif0, rtranif1, and rtranif0 devices shall have three items in their terminal lists. The first two shall be bidirectional terminals that conduct signals to and from the devices, and the third terminal shall connect to a control input. The tran and rtran devices shall have terminal lists containing two bidirectional terminals. Both bidirectional terminals shall unconditionally conduct signals to and from the devices, allowing signals to pass in either direction through the devices. The bidirectional terminals of all six devices shall be connected only to scalar nets or bit-selects of vector nets. The tran, tranif0, and tranif1 devices shall pass signals with an alteration in their strength in only one case, as discussed in 7.11. The rtran, rtranif0, and rtranif1 devices shall reduce the strength of the signals passing through them according to rules discussed in 7.12. Example: The following example declares an instance of tranif1: tranif1 t1 (inout1,inout2,control); The bidirectional terminals are inout1 and inout2. The control input is control. The instance name is t1. 7.7 CMOS switches The instance declaration of a CMOS switch shall begin with one of the following keywords: cmos rcmos The delay specification shall be zero, one, two, or three delays. If the delay specification contains three delays, the first delay shall determine the rise delay, the second delay shall determine the fall delay, the third delay shall determine the delay of transitions to z, and the smallest of the three delays shall determine the delay of transitions to x. Delays in transitions to H or L are the same as delays in transitions to x. If the specification contains two delays, the first delay shall determine the output rise delay, the second delay shall determine the output fall delay, and the smaller of the two delays shall apply to output transitions to x and z. If only one delay is specified, it shall specify the delay for all output transitions. If there is no delay specification, there shall be no propagation delay through the switch. The cmos and rcmos switches shall have a data input, a data output, and two control inputs. In the terminal list, the first terminal shall connect to the data output, the second terminal shall connect to the data input, the 86 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C third terminal shall connect to the n-channel control input, and the last terminal shall connect to the p-channel control input. The cmos gate shall pass signals with an alteration in their strength in only one case, as discussed in 7.11. The rcmos gate shall reduce the strength of signals passing through it according to rules described in 7.12. The cmos switch shall be treated as the combination of a pmos switch and an nmos switch. The rcmos switch shall be treated as the combination of an rpmos switch and an rnmos switch. The combined switches in these configurations shall share data input and data output terminals, but they shall have separate control inputs. Example: The equivalence of the cmos gate to the pairing of an nmos gate and a pmos gate is shown in the following example: cmos (w, datain, ncontrol, pcontrol); is equivalent to: w nmos (w, datain, ncontrol); pmos (w, datain, pcontrol); ncontrol nmos pmos pcontrol datain 7.8 pullup and pulldown sources The instance declaration of a pullup or a pulldown source shall begin with one of the following keywords: pullup pulldown A pullup source shall place a logic value 1 on the nets connected in its terminal list. A pulldown source shall place a logic value 0 on the nets connected in its terminal list. The signals that these sources place on nets shall have pull strength in the absence of a strength specification. If conflicting strength specification is declared, it shall be ignored. There shall be no delay specifications for these sources. Example: The following example declares two pullup instances: pullup (strong1) p1 (neta), p2 (netb); In this example, the p1 instance drives neta and the p2 instance drives netb. Copyright © 2001 IEEE. All rights reserved. 87 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 7.9 Logic strength modeling The Verilog HDL provides for accurate modeling of signal contention, bidirectional pass gates, resistive MOS devices, dynamic MOS, charge sharing, and other technology-dependent network configurations by allowing scalar net signal values to have a full range of unknown values and different levels of strength or combinations of levels of strength. This multiple-level logic strength modeling resolves combinations of signals into known or unknown values to represent the behavior of hardware with improved accuracy. A strength specification shall have two components a) The strength of the 0 portion of the net value, called strength0, designated as one of the following: supply0 strong0 pull0 weak0 highz0 b) The strength of the 1 portion of the net value, called strength1, designated as one of the following: supply1 strong1 pull1 weak1 highz1 The combinations (highz0, highz1) and (highz1, highz0) shall be considered illegal. Despite this division of the strength specification, it is helpful to consider strength as a property occupying regions of a continuum in order to predict the results of combinations of signals. Table 37 demonstrates the continuum of strengths. The left column lists the keywords used in specifying strengths. The right column gives correlated strength levels. Table 37—Strength levels for scalar net signal values Strength name Strength level supply0 7 strong0 6 pull0 5 large0 4 weak0 3 medium0 2 small0 1 highz0 0 highz1 0 small1 1 medium1 2 weak1 3 large1 4 pull1 5 strong1 6 supply1 7 88 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C In Table 37, there are four driving strengths: supply strong pull weak Signals with driving strengths shall propagate from gate outputs and continuous assignment outputs. In Table 37, there are three charge storage strengths: large medium small Signals with the charge storage strengths shall originate in the trireg net type. It is possible to think of the strengths of signals in the preceding table as locations on the scale in Figure 5. strength0 strength1 76543210 01234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 Figure 5—Scale of strengths Discussions of signal combinations later in this section employs graphics similar to those used in Figure 5. If the signal value of a net is known, all of its strength levels shall be in either the strength0 part of the scale represented by Figure 5, or all strength levels shall be in its strength1 part. If the signal value of a net is unknown, it shall have strength levels in both the strength0 and the strength1 parts. A net with a signal value z shall have a strength level only in one of the 0 subdivisions of the parts of the scale. 7.10 Strengths and values of combined signals In addition to a signal value, a net shall have either a single unambiguous strength level or an ambiguous strength consisting of more than one level. When signals combine, their strengths and values shall determine the strength and value of the resulting signal in accordance with the principles in 7.10.1 through 7.10.4. 7.10.1 Combined signals of unambiguous strength This subclause deals with combinations of signals in which each signal has a known value and a single strength level. If two or more signals of unequal strength combine in a wired net configuration, the stronger signal shall dominate all the weaker drivers and determine the result. The combination of two or more signals of like value shall result in the same value with the greater of all the strengths. The combination of signals identical in strength and value shall result in the same signal. The combination of signals with unlike values and the same strength can have three possible results. Two of the results occur in the presence of wired logic and the third occurs in its absence. Wired logic is discussed in 7.10.4. The result in the absence of wired logic is the subject of Figure 7. Copyright © 2001 IEEE. All rights reserved. 89 IEEE Std 1364-2001 Version C Example: Pu1(5) St0(6) Su1(7) La1(4) IEEE STANDARD VERILOG® St0(6) Su1(7) Figure 6—Combining unequal strengths In Figure 6, the numbers in parentheses indicate the relative strengths of the signals. The combination of a pull1 and a strong0 results in a strong0, which is the stronger of the two signals. 7.10.2 Ambiguous strengths: sources and combinations There are several classifications of signals possessing ambiguous strengths — Signals with known values and multiple strength levels — Signals with a value x, which have strength levels consisting of subdivisions of both the strength1 and the strength0 parts of the scale of strengths in Figure 5 — Signals with a value L, which have strength levels that consist of high impedance joined with strength levels in the strength0 part of the scale of strengths in Figure 5 — Signals with a value H, which have strength levels that consist of high impedance joined with strength levels in the strength1 part of the scale of strengths in Figure 5 Many configurations can produce signals of ambiguous strength. When two signals of equal strength and opposite value combine, the result shall be a value x, along with the strength levels of both signals and all the smaller strength levels. Examples: Figure 7 shows the combination of a weak signal with a value 1 and a weak signal with a value 0 yielding a signal with weak strength and a value x. We1 WeX We0 Figure 7—Combination of signals of equal strength and opposite values This output signal is described in Figure 8. 90 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C strength0 strength1 76543210 01234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 Figure 8—Weak x signal strength An ambiguous signal strength can be a range of possible values. An example is the strength of the output from the three-state drivers with unknown control inputs as shown in Figure 9. St1 We0 X bufif1 X bufif0 StH StL Figure 9—Bufifs with control inputs of x The output of the bufif1 in Figure 9 is a strong H, composed of the range of values described in Figure 10. strength0 strength1 76543210 01234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 Figure 10—Strong H range of values The output of the bufif0 in Figure 9 is a strong L, composed of the range of values described in Figure 11. Copyright © 2001 IEEE. All rights reserved. 91 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® strength0 strength1 76543210 01234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 Figure 11—Strong L range of values The combination of two signals of ambiguous strength shall result in a signal of ambiguous strength. The resulting signal shall have a range of strength levels that includes the strength levels in its component signals. The combination of outputs from two three-state drivers with unknown control inputs, shown in Figure 12, is an example. X PuH Pu1 X 35X We0 WeL Figure 12—Combined signals of ambiguous strength In Figure 12, the combination of signals of ambiguous strengths produces a range that includes the extremes of the signals and all the strengths between them, as described in Figure 13. strength0 strength1 76543210 01234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 Figure 13—Range of strengths for an unknown signal The result is a value x because its range includes the values 1 and 0. The number 35, which precedes the x, is a concatenation of two digits. The first is the digit 3, which corresponds to the highest strength0 level for the result. The second digit, 5, corresponds to the highest strength1 level for the result. Switch networks can produce a ranges of strengths of the same value, such as the signals from the upper and lower configurations in Figure 14. 92 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C reg b =x Vcc pullup reg a =1 (6) Pu1 (5) 651 reg g =x 56X Pu0 (5) 530 reg d =0 reg e =0 and We0 (3) pulldown ground Figure 14—Ambiguous strengths from switch networks In Figure 14, the upper combination of a reg, a gate controlled by a reg of unspecified value, and a pullup produces a signal with a value of 1 and a range of strengths (651) described in Figure 15. strength0 strength1 76543210 01234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 Figure 15—Range of two strengths of a defined value In Figure 14, the lower combination of a pulldown, a gate controlled by a reg of unspecified value, and an and gate produces a signal with a value 0 and a range of strengths (530) described in Figure 16. strength0 strength1 76543210 01234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 Figure 16—Range of three strengths of a defined value Copyright © 2001 IEEE. All rights reserved. 93 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® When the signals from the upper and lower configurations in Figure 14 combine, the result is an unknown with a range (56x) determined by the extremes of the two signals shown in Figure 17. strength0 strength1 76543210 01234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 Figure 17—Unknown value with a range of strengths In Figure 14, replacing the pulldown in the lower configuration with a supply0 would change the range of the result to the range (StX) described in Figure 18. The range in Figure 18 is strong x, because it is unknown and the extremes of both its components are strong. The extreme of the output of the lower configuration is strong because the lower pmos reduces the strength of the supply0 signal. This modeling feature is discussed in 7.11. strength0 strength1 76543210 01234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 Figure 18—Strong X range Logic gates produce results with ambiguous strengths as well as three-state drivers. Such a case appears in Figure 19. The and gate N1 is declared with highz0 strength, and N2 is declared with weak0 strength. a=1 b=X c=0 d=0 StH N1 N2 We0 and (strong1,highz0) N1(a,b); and (strong1, weak0) N2(c,d); 36X Figure 19—Ambiguous strength from gates In Figure 19, reg b has an unspecified value, so input to the upper and gate is strong x. The upper and gate has a strength specification including highz0. The signal from the upper and gate is a strong H composed of the values as described in Figure 20. 94 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C strength0 strength1 76543210 01234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 Figure 20—Ambiguous strength signal from a gate HiZ0 is part of the result, because the strength specification for the gate in question specified that strength for an output with a value 0. A strength specification other than high impedance for the 0 value output results in a gate output value x. The output of the lower and gate is a weak 0 as described in Figure 21. strength0 strength1 76543210 01234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 Figure 21—Weak 0 When the signals combine, the result is the range (36x) as described in Figure 22. strength0 strength1 76543210 01234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 Figure 22—Ambiguous strength in combined gate signals Figure 22 presents the combination of an ambiguous signal and an unambiguous signal. Such combinations are the topic of 7.10.3. 7.10.3 Ambiguous strength signals and unambiguous signals The combination of a signal with unambiguous strength and known value with another signal of ambiguous strength presents several possible cases. To understand a set of rules governing this type of combination, it is necessary to consider the strength levels of the ambiguous strength signal separately from each other and relative to the unambiguous strength signal. When a signal of known value and unambiguous strength combines with a component of a signal of ambiguous strength, these shall be the effects: Copyright © 2001 IEEE. All rights reserved. 95 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® a) The strength levels of the ambiguous strength signal that are greater than the strength level of the unambiguous signal shall remain in the result. b) The strength levels of the ambiguous strength signal that are smaller than or equal to the strength level of the unambiguous signal shall disappear from the result, subject to rule c. c) If the operation of rule a and rule b results in a gap in strength levels because the signals are of opposite value, the signals in the gap shall be part of the result. The following figures show some applications of the rules. strength0 strength1 76543210 01234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 strength0 strength1 76543210 01234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 Combining the two signals above results in the following signal: strength0 strength1 76543210 01234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 Figure 23—Elimination of strength levels In Figure 23, the strength levels in the ambiguous strength signal that are smaller than or equal to the strength level of the unambiguous strength signal disappear from the result, demonstrating rule b. 96 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C strength0 strength1 7654321001234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 strength0 strength1 7654321001234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 Combining the two signals above results in the following signal: strength0 strength1 76543210 01234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 Figure 24—Result demonstrating a range and the elimination of strength levels of two values In Figure 24, rules a, b, and c apply. The strength levels of the ambiguous strength signal that are of opposite value and lesser strength than the unambiguous strength signal disappear from the result. The strength levels in the ambiguous strength signal that are less than the strength level of the unambiguous strength signal, and of the same value, disappear from the result. The strength level of the unambiguous strength signal and the greater extreme of the ambiguous strength signal define a range in the result. Copyright © 2001 IEEE. All rights reserved. 97 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® strength0 strength1 7654321001234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 strength0 strength1 7654321001234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 Combining the two signals above results in the following signal: strength0 strength1 7654321001234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 Figure 25—Result demonstrating a range and the elimination of strength levels of one value In Figure 25, rules a and b apply. The strength levels in the ambiguous strength signal that are less than the strength level of the unambiguous strength signal disappear from the result. The strength level of the unambiguous strength signal and the strength level at the greater extreme of the ambiguous strength signal define a range in the result. 98 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C strength0 strength1 76543210 01234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 strength0 strength1 76543210 01234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 Combining the two signals above results in the following signal: strength0 strength1 76543210 01234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 Figure 26—A range of both values In Figure 26, rules a, b, and c apply. The greater extreme of the range of strengths for the ambiguous strength signal is larger than the strength level of the unambiguous strength signal. The result is a range defined by the greatest strength in the range of the ambiguous strength signal and by the strength level of the unambiguous strength signal. 7.10.4 Wired logic net types The net types triand, wand, trior, and wor shall resolve conflicts when multiple drivers have the same strength. These net types shall resolve signal values by treating signals as inputs of logic functions. Examples: Consider the combination of two signals of unambiguous strength in Figure 27. Copyright © 2001 IEEE. All rights reserved. 99 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® strength0 strength1 7654321001234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 strength0 strength1 7654321001234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 wired AND logic value result: 0 wired OR logic value result: 1 Figure 27—Wired logic with unambiguous strength signals The combination of the signals in Figure 27, using wired and logic, produces a result with the same value as the result produced by an and gate with the value of the two signals as its inputs. The combination of signals using wired or logic produces a result with the same value as the result produced by an or gate with the values of the two signals as its inputs. The strength of the result is the same as the strength of the combined signals in both cases. If the value of the upper signal changes so that both signals in Figure 27 possess a value 1, then the results of both types of logic have a value 1. When ambiguous strength signals combine in wired logic, it is necessary to consider the results of all combinations of each of the strength levels in the first signal with each of the strength levels in the second signal, as shown in Figure 28. 100 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C strength0 strength1 7654321001234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 Signal 1 strength0 strength1 7654321001234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 Signal 2 The combinations of strength levels for and logic appear in the following chart: signal1 signal2 result strength value strength value strength value 5 0 5 1 5 0 6 0 5 1 6 0 The result is the following signal: strength0 strength1 7654321001234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 The combinations of strength levels for or logic appear in the following chart: signal1 signal2 result strength value strength value strength value 5 0 5 1 5 1 6 0 5 1 6 0 The result is the following signal: strength0 strength1 7654321001234567 Su0 St0 Pu0 La0 We0 Me0 Sm0 HiZ0 HiZ1 Sm1 Me1 We1 La1 Pu1 St1 Su1 Figure 28—Wired logic and ambiguous strengths Copyright © 2001 IEEE. All rights reserved. 101 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 7.11 Strength reduction by nonresistive devices The nmos, pmos, and cmos switches shall pass the strength from the data input to the output, except that a supply strength shall be reduced to a strong strength. The tran, tranif0, and tranif1 switches shall not affect signal strength across the bidirectional terminals, except that a supply strength shall be reduced to a strong strength. 7.12 Strength reduction by resistive devices The rnmos, rpmos, rcmos, rtran, rtranif1, and rtranif0 devices shall reduce the strength of signals that pass through them according to Table 38. Table 38—Strength reduction rules Input strength Reduced strength Supply drive Strong drive Pull drive Large capacitor Weak drive Medium capacitor Small capacitor High impedance Pull drive Pull drive Weak drive Medium capacitor Medium capacitor Small capacitor Small capacitor High impedance 7.13 Strengths of net types The tri0, tri1, supply0, and supply1 net types shall generate signals with specific strength levels. The trireg declaration can specify either of two signal strength levels other than a default strength level. 7.13.1 tri0 and tri1 net strengths The tri0 net type models a net connected to a resistive pulldown device. In the absence of an overriding source, such a signal shall have a value 0 and a pull strength. The tri1 net type models a net connected to a resistive pullup device. In the absence of an overriding source, such a signal shall have a value 1 and a pull strength. 7.13.2 trireg strength The trireg net type models charge storage nodes. The strength of the drive resulting from a trireg net that is in the charge storage state (that is, a driver charged the net and then went to high impedance) shall be one of these three strengths: large, medium, or small. The specific strength associated with a particular trireg net shall be specified by the user in the net declaration. The default shall be medium. The syntax of this specification is described in 3.4.1. 7.13.3 supply0 and supply1 net strengths The supply0 net type models ground connections. The supply1 net type models connections to power supplies. The supply0 and supply1 net types shall have supply driving strengths. 102 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 7.14 Gate and net delays Gate and net delays provide a means of more accurately describing delays through a circuit. The gate delays specify the signal propagation delay from any gate input to the gate output. Up to three values per output representing rise, fall, and turn-off delays can be specified (see 7.2 through 7.8). Net delays refer to the time it takes from any driver on the net changing value to the time when the net value is updated and propagated further. Up to three delay values per net can be specified. For both gates and nets, the default delay shall be zero when no delay specification is given. When one delay value is given, then this value shall be used for all propagation delays associated with the gate or the net. When two delays are given, the first delay shall specify the rise delay and the second delay shall specify the fall delay. The delay when the signal changes to high impedance or to unknown shall be the lesser of the two delay values. For a three-delay specification — The first delay refers to the transition to the 1 value (rise delay). — The second delay refers to the transition to the 0 value (fall delay). — The third delay refers to the transition to the high-impedance value. When a value changes to the unknown (x) value, the delay is the smallest of the three delays. The strength of the input signal shall not affect the propagation delay from an input to an output. Table 39 summarizes the from-to propagation delay choice for the two- and three-delay specifications. Table 39—Rules for propagation delays Delay used if there are From value: 0 0 0 1 1 1 x x x z z z To value: 1 x z 0 x z 0 1 z 0 1 x 2 delays d1 min(d1, d2) min(d1, d2) d2 min(d1, d2) min(d1, d2) d2 d1 min(d1, d2) d2 d1 min(d1, d2) 3 delays d1 min(d1, d2, d3) d3 d2 min(d1, d2, d3) d3 d2 d1 d3 d2 d1 min(d1, d2, d3) Copyright © 2001 IEEE. All rights reserved. 103 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Examples: Example 1—The following is an example of a delay specification with one, two, and three delays: and #(10) a1 (out, in1, in2); // only one delay and #(10,12) a2 (out, in1, in2); // rise and fall delays bufif0 #(10,12,11) b3 (out, in, ctrl);// rise, fall, and turn-off delays Example 2—The following example specifies a simple latch module with three-state outputs, where individual delays are given to the gates. The propagation delay from the primary inputs to the outputs of the module will be cumulative, and it depends on the signal path through the network. module tri_latch (qout, nqout, clock, data, enable); output qout, nqout; input clock, data, enable; tri qout, nqout; not #5 nand #(3,5) nand #(12,15) bufif1 #(3,7,13) n1 (ndata, data); n2 (wa, data, clock), n3 (wb, ndata, clock); n4 (q, nq, wa), n5 (nq, q, wb); q_drive (qout, q, enable), nq_drive (nqout, nq, enable); endmodule 7.14.1 min:typ:max delays The syntax for delays on gate primitives (including user-defined primitives; see Clause 8), nets, and continuous assignments shall allow three values each for the rising, falling, and turn-off delays. The minimum, typical, and maximum values for each delay shall be specified as constant expressions separated by colons. There shall be no required relation (e.g., min ≤ typ ≤ max) between the expressions for minimum, typical, and maximum delays. These can be any three constant expressions. Examples: The following example shows min:typ:max values for rising, falling, and turn-off delays: module iobuf (io1, io2, dir); ... bufif0 #(5:7:9, 8:10:12, 15:18:21) b1 (io1, io2, dir); bufif1 #(6:8:10, 5:7:9, 13:17:19) b2 (io2, io1, dir); ... endmodule The syntax for delay controls in procedural statements (see 9.7) also allows minimum, typical, and maximum values. These are specified by expressions separated by colons. The following example illustrates this concept. 104 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C parameter min_hi = 97, typ_hi = 100, max_hi = 107; reg clk; always begin #(95:100:105) clk = 1; #(min_hi:typ_hi:max_hi) clk = 0; end 7.14.2 trireg net charge decay Like all nets, the delay specification in a trireg net declaration can contain up to three delays. The first two delays shall specify the delay for transition to the 1 and 0 logic states when the trireg net is driven to these states by a driver. The third delay shall specify the charge decay time instead of the delay in a transition to the z logic state. The charge decay time specifies the delay between when the drivers of a trireg net turn off and when its stored charge can no longer be determined. A trireg net does not need a turn-off delay specification because a trireg net never makes a transition to the z logic state. When the drivers of a trireg net make transitions from the 1, 0, or x logic states to off, the trireg net shall retain the previous 1, 0, or x logic state that was on its drivers. The z value shall not propagate from the drivers of a trireg net to a trireg net. A trireg net can only hold a z logic state when z is the initial logic state of the trireg net or when the trireg net is forced to the z state with a force statement (see 9.3.2). A delay specification for charge decay models a charge storage node that is not ideal—a charge storage node whose charge leaks out through its surrounding devices and connections. The following subclauses describe the charge decay process and the delay specification for charge decay. 7.14.2.1 The charge decay process Charge decay is the cause of transition of a 1 or 0 that is stored in a trireg net to an unknown value (x) after a specified delay. The charge decay process shall begin when the drivers of the trireg net turn off and the trireg net starts to hold charge. The charge decay process shall end under the following two conditions: a) The delay specified by charge decay time elapses and the trireg net makes a transition from 1 or 0 to x. b) The drivers of trireg net turn on and propagate a 1, 0, or x into the trireg net. 7.14.2.2 The delay specification for charge decay time The third delay in a trireg net declaration shall specify the charge decay time. A three-valued delay specification in a trireg net declaration shall have the following form: #(d1, d2, d3) // (rise_delay, fall_delay, charge_decay_time) The charge decay time specification in a trireg net declaration shall be preceded by a rise and a fall delay specification. Examples: Example 1—The following example shows a specification of the charge decay time in a trireg net declaration: trireg (large) #(0,0,50) cap1; Copyright © 2001 IEEE. All rights reserved. 105 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® This example declares a trireg net named cap1. This trireg net stores a large charge. The delay specifications for the rise delay is 0, the fall delay is 0, and the charge decay time specification is 50 time units. Example 2—The next example presents a source description file that contains a trireg net declaration with a charge decay time specification. Figure 29 shows an equivalent schematic for the source description. gate data nmos1 trireg Figure 29—Trireg net with capacitance module capacitor; reg data, gate; // trireg declaration with a charge decay time of 50 time units trireg (large) #(0,0,50) cap1; nmos nmos1 (cap1, data, gate); // nmos that drives the trireg initial begin $monitor("%0d data=%v gate=%v cap1=%v", $time, data, gate, cap1); data = 1; // Toggle the driver of the control input to the nmos switch gate = 1; #10 gate = 0; #30 gate = 1; #10 gate = 0; #100 $finish; end endmodule 106 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 8. User-defined primitives (UDPs) This clause describes a modeling technique to augment the set of predefined gate primitives by designing and specifying new primitive elements called user-defined primitives (UDPs). Instances of these new UDPs can be used in exactly the same manner as the gate primitives to represent the circuit being modeled. The following two types of behavior can be represented in a user-defined primitive: a) Combinational—modeled by a combinational UDP b) Sequential—modeled by a sequential UDP A combinational UDP uses the value of its inputs to determine the next value of its output. A sequential UDP uses the value of its inputs and the current value of its output to determine the value of its output. Sequential UDPs provide a way to model sequential circuits such as flip-flops and latches. A sequential UDP can model both level-sensitive and edge-sensitive behavior. Each UDP has exactly one output, which can be in one of three states: 0, 1, or x. The three-state value z is not supported. In sequential UDPs, the output always has the same value as the internal state. The z values passed to UDP inputs shall be treated the same as x values. 8.1 UDP definition UDP definitions are independent of modules; they are at the same level as module definitions in the syntax hierarchy. They can appear anywhere in the source text, either before or after they are instantiated inside a module. They shall not appear between the keywords module and endmodule. NOTE—Implementations may limit the maximum number of UDP definitions in a model, but they shall allow at least 256. The formal syntax of the UDP definition is given in Syntax 8-1. Copyright © 2001 IEEE. All rights reserved. 107 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® udp_declaration ::= (From Annex A - A.5.1) { attribute_instance } primitive udp_identifier ( udp_port_list ) ; udp_port_declaration { udp_port_declaration } udp_body endprimitive | { attribute_instance } primitive udp_identifier ( udp_declaration_port_list ) ; udp_body endprimitive udp_port_list ::= (From Annex A - A.5.2) output_port_identifier , input_port_identifier { , input_port_identifier } udp_declaration_port_list ::= udp_output_declaration , udp_input_declaration { , udp_input_declaration } udp_port_declaration ::= udp_output_declaration ; | udp_input_declaration ; | udp_reg_declaration ; udp_output_declaration ::= { attribute_instance } output port_identifier | { attribute_instance } output reg port_identifier [ = constant_expression ] udp_input_declaration ::= { attribute_instance } input list_of_port_identifiers udp_reg_declaration ::= { attribute_instance } reg variable_identifier udp_body ::= (From Annex A - A.5.3) combinational_body | sequential_body combinational_body ::= table combinational_entry { combinational_entry } endtable combinational_entry ::= level_input_list : output_symbol ; sequential_body ::= [ udp_initial_statement ] table sequential_entry { sequential_entry } endtable udp_initial_statement ::= initial output_port_identifier = init_val ; init_val ::= 1'b0 | 1'b1 | 1'bx | 1'bX | 1'B0 | 1'B1 | 1'Bx | 1'BX | 1 | 0 sequential_entry ::= seq_input_list : current_state : next_state ; seq_input_list ::= level_input_list | edge_input_list level_input_list ::= level_symbol { level_symbol } edge_input_list ::= { level_symbol } edge_indicator { level_symbol } edge_indicator ::= ( level_symbol level_symbol ) | edge_symbol current_state ::= level_symbol next_state ::=output_symbol | output_symbol ::= 0 | 1 | x | X level_symbol ::= 0 | 1 | x | X | ? | b | B edge_symbol ::= r | R | f | F | p | P | n | N | * Syntax 8-1—Syntax for user-defined primitives 108 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 8.1.1 UDP header A UDP definition shall have one of two alternate forms. The first form shall begin with the keyword primitive, followed by an identifier, which shall be the name of the UDP. This in turn shall be followed by a comma-separated list of port names enclosed in parentheses, which shall be followed by a semicolon. The UDP definition header shall be followed by port declarations and a state table. The UDP definition shall be terminated by the keyword endprimitive. The second form shall begin with the keyword primitive, followed by an identifier, which shall be the name of the UDP. This in turn shall be followed by a comma separated list of port declarations enclosed in parenthesis, followed by a semicolon. The UDP definition header shall be followed by a state table. The UDP definition shall be terminated by the keyword endprimitive. UDPs have multiple input ports and exactly one output port; bidirectional inout ports are not permitted on UDPs. All ports of a UDP shall be scalar; vector ports are not permitted. The output port shall be the first port in the port list. 8.1.2 UDP port declarations UDPs shall contain input and output port declarations. The output port declaration begins with the keyword output, followed by one output port name. The input port declaration begins with the keyword input, followed by one or more input port names. Sequential UDPs shall contain a reg declaration for the output port, either in addition to the output declaration, when the UDP is declared using the first form of a UDP Header, or as part of the output_declaration, in either case. Combinational UDPs cannot contain a reg declaration. The initial value of the output port can be specified in an initial statement in a sequential UDP (see 8.1.3). NOTE—Implementations may limit the maximum number of inputs to a UDP, but they shall allow at least 9 inputs for sequential UDPs and 10 inputs for combinational UDPs. 8.1.3 Sequential UDP initial statement The sequential UDP initial statement specifies the value of the output port when simulation begins. This statement begins with the keyword initial. The statement that follows shall be an assignment statement that assigns a single-bit literal value to the output port. 8.1.4 UDP state table The state table defines the behavior of a UDP. It begins with the keyword table and is terminated with the keyword endtable. Each row of the table is terminated by a semicolon. Each row of the table is created using a variety of characters (see Table 40), which indicate input values and output state. Three states—0, 1, and x—are supported. The z state is explicitly excluded from consideration in user-defined primitives. A number of special characters are defined to represent certain combinations of state possibilities. These are described in Table 40. The order of the input state fields of each row of the state table is taken directly from the port list in the UDP definition header. It is not related to the order of the input port declarations. Combinational UDPs have one field per input and one field for the output. The input fields are separated from the output field by a colon (:). Each row defines the output for a particular combination of the input values (see 8.2). Copyright © 2001 IEEE. All rights reserved. 109 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Sequential UDPs have an additional field inserted between the input fields and the output field. This additional field represents the current state of the UDP and is considered equivalent to the current output value. It is delimited by colons. Each row defines the output based on the current state, particular combinations of input values, and at most one input transition (see 8.4). A row such as the one shown below is illegal: (01) (10) 0 : 0 : 1 ; If all input values are specified as x, then the output state shall be specified as x. It is not necessary to explicitly specify every possible input combination. All combinations of input values that are not explicitly specified result in a default output state of x. It is illegal to have the same combination of inputs, including edges, specified for different outputs. 8.1.5 Z values in UDP The z value in a table entry is not supported and it is considered illegal. The z values passed to UDP inputs shall be treated the same as x values. 8.1.6 Summary of symbols To improve the readability and to ease writing of the state table, several special symbols are provided. Table 40 summarizes the meaning of all the value symbols that are valid in the table part of a UDP definition. Symbol 0 1 x ? b (vw) * r f p n Table 40—UDP table symbols Interpretation Logic 0 Logic 1 Unknown Iteration of 0, 1, and x Iteration of 0 and 1 No change Value change from v to w Same as (??) Same as (01) Same as (10) Iteration of (01), (0 x) and (x1) Iteration of (10), (1x)and (x0 Comments Permitted in the input fields of all UDPs and in the current state field of sequential UDPs. Not permitted in output field. Permitted in the input fields of all UDPs and in the current state field of sequential UDPs. Not permitted in the output field. Permitted only in the output field of a sequential UDP. v and w can be any one of 0, 1, x, ?, or b, and are only permitted in the input field. Any value change on input. Rising edge on input. Falling edge on input. Potential positive edge on the input. Potential negative edge on the input. 110 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 8.2 Combinational UDPs In combinational UDPs, the output state is determined solely as a function of the current input states. Whenever an input state changes, the UDP is evaluated and the output state is set to the value indicated by the row in the state table that matches all the input states. All combinations of the inputs that are not explicitly specified will drive the output state to the unknown value x. Examples: The following example defines a multiplexer with two data inputs and a control input. primitive multiplexer (mux, control, dataA, dataB); output mux; input control, dataA, dataB; table // control dataA dataB mux 0 1 0 : 1; 0 1 1 : 1; 0 1 x : 1; 0 0 0 : 0; 0 0 1 : 0; 0 0 x : 0; 1 0 1 : 1; 1 1 1 : 1; 1 x 1 : 1; 1 0 0 : 0; 1 1 0 : 0; 1 x 0 : 0; x 0 0 : 0; x 1 1 : 1; endtable endprimitive The first entry in this example can be explained as follows: when control equals 0, and dataA equals 1, and dataB equals 0, then output mux equals 1. The input combination 0xx (control=0, dataA=x, dataB=x) is not specified. If this combination occurs during simulation, the value of output port mux will become x. Using ?, the description of a multiplexer can be abbreviated as primitive multiplexer (mux, control, dataA, dataB); output mux; input control, dataA, dataB; table // control dataA dataB mux 0 1 ? : 1 ; // ? = 0 1 x 0 0 ? : 0; 1 ? 1 : 1; 1 ? 0 : 0; x 0 0 : 0; x 1 1 : 1; endtable endprimitive Copyright © 2001 IEEE. All rights reserved. 111 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 8.3 Level-sensitive sequential UDPs Level-sensitive sequential behavior is represented the same way as combinational behavior, except that the output is declared to be of type reg, and there is an additional field in each table entry. This new field represents the current state of the UDP. The output field in a sequential UDP represents the next state. Example: Consider the example of a latch: primitive latch (q, clock, data); output q; reg q; input clock, data; table // clock data q q+ 0 1:?: 1; 0 0:?: 0; 1 ? : ? : - ; // - = no change endtable endprimitive This description differs from a combinational UDP model in two ways. First, the output identifier q has an additional reg declaration to indicate that there is an internal state q. The output value of the UDP is always the same as the internal state. Second, a field for the current state, which is separated by colons from the inputs and the output, has been added. 8.4 Edge-sensitive sequential UDPs In level-sensitive behavior, the values of the inputs and the current state are sufficient to determine the output value. Edge-sensitive behavior differs in that changes in the output are triggered by specific transitions of the inputs. This makes the state table a transition table. Each table entry can have a transition specification on at most one input. A transition is specified by a pair of values in parenthesis such as (01) or a transition symbol such as r. Entries such as the following are illegal: (01)(01)0 : 0 : 1 ; All transitions that do not affect the output shall be explicitly specified. Otherwise, such transitions cause the value of the output to change to x. All unspecified transitions default to the output value x. If the behavior of the UDP is sensitive to edges of any input, the desired output state shall be specified for all edges of all inputs. Example: The following example describes a rising edge D flip-flop: 112 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C primitive d_edge_ff (q, clock, data); output q; reg q; input clock, data; table // clock data q q+ // obtain output on rising edge of clock (01) 0 : ? : 0 ; (01) 1 : ? : 1 ; (0?) 1 : 1 : 1 ; (0?) 0 : 0 : 0 ; // ignore negative edge of clock (?0) ? : ? : - ; // ignore data changes on steady clock ? (??) : ? : - ; endtable endprimitive The terms such as (01) represent transitions of the input values. Specifically, (01) represents a transition from 0 to 1. The first line in the table of the preceding UDP definition is interpreted as follows: when clock changes value from 0 to 1, and data equals 0, the output goes to 0 no matter what the current state. The transition of clock from 0 to x with data equal to 0 and current state equal to 1 will result in the output q going to x. 8.5 Sequential UDP initialization The initial value on the output port of a sequential UDP can be specified with an initial statement that provides a procedural assignment. The initial statement is optional. Like initial statements in modules, the initial statement in UDPs begin with the keyword initial. The valid contents of initial statements in UDPs and the valid left-hand and right-hand sides of their procedural assignment statements differ from initial statements in modules. A partial list of differences between these two types of initial statements is described in Table 41. Table 41—Initial statements in UDPs and modules Initial statements in UDPs Initial statements in modules Contents limited to one procedural assignment statement Contents can be one procedural statement of any type or a block statement that contains more than one procedural statement The procedural assignment statement shall assign a value to a reg whose identifier matches the identifier of an output terminal Procedural assignment statements in initial statements can assign values to a reg whose identifier does not match the identifier of an output terminal The procedural assignment statement shall assign one of the following values: 1’b1, 1’b0, 1’bx, 1, 0 Procedural assignment statements can assign values of any size, radix, and value Copyright © 2001 IEEE. All rights reserved. 113 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Examples: Example 1—The following example shows a sequential UDP that contains an initial statement. primitive srff (q, s, r); output q; reg q; input s, r; initial q = 1’b1; table // s r q q+ 1 0:?:1; f 0:1:-; 0 r:?:0; 0 f:0:-; 1 1:?:0; endtable endprimitive The output q has an initial value of 1 at the start of the simulation; a delay specification on an instantiated UDP does not delay the simulation time of the assignment of this initial value to the output. When simulation starts, this value is the current state in the state table. Delays are not permitted in a UDP initial statement. Example 2—The following example and figure show how values are applied in a module that instantiates a sequential UDP with an initial statement. primitive dff1 (q, clk, d); input clk, d; output q; reg q; initial q = 1’b1; table // clk d q q+ r 0:?:0 ; r 1:?:1 ; f ?:?:- ; ? *:?:- ; endtable endprimitive module dff (q, qb, clk, d); input clk, d; output q, qb; dff1 g1 (qi, clk, d); buf #3 g2 (q, qi); not #5 g3 (qb, qi); endmodule The UDP dff1 contains an initial statement that sets the initial value of its output to 1. The module dff contains an instance of UDP dff1. Figure 30 shows the schematic of the preceding module and the simulation propagation times of the initial value of the UDP output. 114 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C module dff d UDP dff1 g1 clk buf g2 q #3 qi not g3 qb #5 qi 1 0 q 1 0 qb 1 0 0 3 5 simulation time Figure 30—Module schematic and simulation times of initial value propagation In Figure 30, the fanout from the UDP output qi includes nets q and qb. At simulation time 0, qi changes value to 1. That initial value of qi does not propagate to net q until simulation time 3, and it does not propagate to net qb until simulation time 5. 8.6 UDP instances The syntax for creating a UDP instance is shown in Syntax 8-2. udp_instantiation ::= (From Annex A- A.5.4) udp_identifier [ drive_strength ] [ delay2 ] [attribute_instance] udp_instance { , udp_instance } ; udp_instance ::= [ name_of_udp_instance ] ( output_terminal , input_terminal { , input_terminal } ) name_of_udp_instance ::= udp_instance_identifier [ range ] Syntax 8-2—Syntax for UDP instances Copyright © 2001 IEEE. All rights reserved. 115 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Instances of user-defined primitives are specified inside modules in the same manner as gates (see 7.1). The instance name is optional, just as for gates. The port connection order is as specified in the UDP definition. Only two delays can be specified because z is not supported for UDPs. An optional range may be specified for an array of UDP instances. The port connection rules remain the same as outlined in 7.1. Example: The following example creates an instance of the D-type flip-flop d_edge_ff (defined in 8.4). module flip; reg clock, data; parameter p1 = 10; parameter p2 = 33; parameter p3 = 12; d_edge_ff #p3 d_inst (q, clock, data); initial begin data = 1; clock = 1; #(20 * p1) $finish; end always #p1 clock = ~clock; always #p2 data = ~data; endmodule 8.7 Mixing level-sensitive and edge-sensitive descriptions UDP definitions allow a mixing of the level-sensitive and the edge-sensitive constructs in the same table. When the input changes, the edge-sensitive cases are processed first, followed by level-sensitive cases. Thus, when level-sensitive and edge-sensitive cases specify different output values, the result is specified by the level-sensitive case. Example: primitive jk_edge_ff (q, clock, j, k, preset, clear); output q; reg q; input clock, j, k, preset, clear; table // clock jk pc state output/next state ? ?? 01 : ? : 1 ; // preset logic ? ?? *1 : 1 : 1 ; ? ?? 10 : ? : 0 ; // clear logic ? ?? 1* : 0 : 0 ; r 00 00 : 0 : 1 ; // normal clocking cases r 00 11 : ? : - ; r 01 11 : ? : 0 ; r 10 11 : ? : 1 ; r 11 11 : 0 : 1 ; r 11 11 : 1 : 0 ; f ?? ?? : ? : - ; b *? ?? : ? : - ; // j and k transition cases b ?* ?? : ? : - ; endtable endprimitive 116 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C In this example, the preset and clear logic is level-sensitive. Whenever the preset and clear combination is 01, the output has value 1. Similarly, whenever the preset and clear combination has value 10, the output has value 0. The remaining logic is sensitive to edges of the clock. In the normal clocking cases, the flip-flop is sensitive to the rising clock edge, as indicated by an r in the clock field in those entries. The insensitivity to the falling edge of clock is indicated by a hyphen (-) in the output field (see Table 40) for the entry with an f as the value of clock. Remember that the desired output for this input transition shall be specified to avoid unwanted x values at the output. The last two entries show that the transitions in j and k inputs do not change the output on a steady low or high clock. 8.8 Level-sensitive dominance Table 42 shows level-sensitive and edge-sensitive entries in the example from 8.7, their level-sensitive or edge-sensitive behavior, and a case of input values that each includes. Table 42—Mixing of level-sensitive and edge-sensitive entries Entry ? ?? 01: ?: 1; f ?? ??: ?: -; Included case Behavior 0 00 01: 0: 1; f 00 01: 0: 0; Level-sensitive Edge-sensitive The included cases specify opposite next state values for the same input and current state combination. The level-sensitive included case specifies that when the inputs clock, jk, and pc values are 0, 00, and 01 and the current state is 0, the output changes to 1. The edge-sensitive included case specifies that when clock falls from 1 to 0, the other inputs jk and pc are 00 and 01, and the current state is 0, then the output changes to 0. When the edge-sensitive case is processed first, followed by the level-sensitive case, the output changes to 1. Copyright © 2001 IEEE. All rights reserved. 117 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 9. Behavioral modeling The language constructs introduced so far allow hardware to be described at a relatively detailed level. Modeling a circuit with logic gates and continuous assignments reflects quite closely the logic structure of the circuit being modeled; however, these constructs do not provide the power of abstraction necessary for describing complex high-level aspects of a system. The procedural constructs described in this section are well suited to tackling problems such as describing a microprocessor or implementing complex timing checks. This section starts with a brief overview of a behavioral model to provide a context for many types of behavioral statements in the Verilog HDL. 9.1 Behavioral model overview Verilog behavioral models contain procedural statements that control the simulation and manipulate variables of the data types previously described. These statements are contained within procedures. Each procedure has an activity flow associated with it. The activity starts at the control constructs initial and always. Each initial construct and each always construct starts a separate activity flow. All of the activity flows are concurrent to model the inherent concurrence of hardware. These constructs are formally described in 9.9. The following example shows a complete Verilog behavioral model. module behave; reg [1:0] a, b; initial begin a = ’b1; b = ’b0; end always begin #50 a = ~a; end always begin #100 b = ~b; end endmodule During simulation of this model, all of the flows defined by the initial and always constructs start together at simulation time zero. The initial constructs execute once, and the always constructs execute repetitively. In this model, the reg variables a and b initialize to 1 and 0 respectively at simulation time zero. The initial construct is then complete and does not execute again during this simulation run. This initial construct contains a begin-end block (also called a sequential block) of statements. In this begin-end block a is initialized first, followed by b. The always constructs also start at time zero, but the values of the variables do not change until the times specified by the delay controls (introduced by #) have elapsed. Thus, reg a inverts after 50 time units and reg b inverts after 100 time units. Since the always constructs repeat, this model will produce two square waves. The reg a toggles with a period of 100 time units, and reg b toggles with a period of 200 time units. The two always constructs proceed concurrently throughout the entire simulation run. 118 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 9.2 Procedural assignments As described in Clause 6, procedural assignments are used for updating reg, integer, time, real, realtime, and memory data types. There is a significant difference between procedural assignments and continuous assignments: — Continuous assignments drive nets and are evaluated and updated whenever an input operand changes value. — Procedural assignments update the value of variables under the control of the procedural flow constructs that surround them. The right-hand side of a procedural assignment can be any expression that evaluates to a value. The left-hand side shall be a variable that receives the assignment from the right-hand side. The left-hand side of a procedural assignment can take one of the following forms: — reg, integer, real, realtime, or time data type: an assignment to the name reference of one of these data types. — Bit-select of a reg, integer, or time data type: an assignment to a single bit that leaves the other bits untouched. — Part-select of a reg, integer, or time data type: a part-select of one or more contiguous bits that leaves the rest of the bits untouched. — Memory word: a single word of a memory. — Concatenation of any of the above: a concatenation of any of the previous four forms can be speci- fied, which effectively partitions the result of the right-hand side expression and assigns the partition parts, in order, to the various parts of the concatenation. NOTE—When the right-hand side evaluates to fewer bits than the left-hand side, then if the right-hand side is signed (see 4.5), it shall be sign-extended to the size of the left-hand side. The Verilog HDL contains two types of procedural assignment statements: — Blocking procedural assignment statements — Nonblocking procedural assignment statements Blocking and nonblocking procedural assignment statements specify different procedural flows in sequential blocks. 9.2.1 Blocking procedural assignments A blocking procedural assignment statement shall be executed before the execution of the statements that follow it in a sequential block (see 9.8.1). A blocking procedural assignment statement shall not prevent the execution of statements that follow it in a parallel block (see 9.8.2). The syntax for a blocking procedural assignment is given in Syntax 9-1. Copyright © 2001 IEEE. All rights reserved. 119 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® blocking_assignment ::= (From Annex A - A.6.2) variable_lvalue = [ delay_or_event_control ] expression delay_control ::= (From Annex A - A.6.5) # delay_value | # ( mintypmax_expression ) delay_or_event_control ::= delay_control | event_control | repeat ( expression ) event_control event_control ::= @ event_identifier | @ ( event_expression ) | @* | @ (*) event_expression ::= expression | hierarchical_identifier | posedge expression | negedge expression | event_expression or event_expression | event_expression , event_expression variable_lvalue ::= (From Annex A - A.8.5) hierarchical_variable_identifier | hierarchical_variable_identifier [ expression ] { [ expression ] } | hierarchical_variable_identifier [ expression ] { [ expression ] } [ range_expression ] | hierarchical_variable_identifier [ range_expression ] | variable_concatenation Syntax 9-1—Syntax for blocking assignments In this syntax, variable_lvalue is a data type that is valid for a procedural assignment statement, = is the assignment operator, and delay_or_event_control is the optional intra-assignment timing control. The control can be either a delay control (e.g., #6) or an event_control (e.g., @(posedge clk)). The expression is the right-hand side value that shall be assigned to the left-hand side. If variable_lvalue requires an evaluation, it shall be evaluated at the time specified by the intra-assignment timing control. The = assignment operator used by blocking procedural assignments is also used by procedural continuous assignments and continuous assignments. Example: The following examples show blocking procedural assignments. rega = 0; rega[3] = 1; // a bit-select rega[3:5] = 7; // a part-select mema[address] = 8’hff; // assignment to a mem element {carry, acc} = rega + regb; // a concatenation 120 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 9.2.2 The nonblocking procedural assignment The nonblocking procedural assignment allows assignment scheduling without blocking the procedural flow. The nonblocking procedural assignment statement can be used whenever several variable assignments within the same time step can be made without regard to order or dependence upon each other. The syntax for a nonblocking procedural assignment is given in Syntax 9-2. nonblocking_assignment ::= (From Annex A - A.6.2) variable_lvalue <= [ delay_or_event_control ] expression delay_control ::= (From Annex A - A.6.5) # delay_value | # ( mintypmax_expression ) delay_or_event_control ::= delay_control | event_control | repeat ( expression ) event_control event_control ::= @ event_identifier | @ ( event_expression ) | @* | @ (*) event_expression ::= expression | hierarchical_identifier | posedge expression | negedge expression | event_expression or event_expression | event_expression , event_expression variable_lvalue ::= (From Annex A - A.8.5) hierarchical_variable_identifier | hierarchical_variable_identifier [ expression ] { [ expression ] } | hierarchical_variable_identifier [ expression ] { [ expression ] } [ range_expression ] | hierarchical_variable_identifier [ range_expression ] | variable_concatenation Syntax 9-2—Syntax for nonblocking assignments In this syntax, variable_lvalue is a data type that is valid for a procedural assignment statement, <= is the nonblocking assignment operator, and delay_or_event_control is the optional intra-assignment timing control. If variable_lvalue requires an evaluation, it shall be evaluated at the same time as the expression on the right-hand side. The order of evaluation of the variable_lvalue and the expression on the right-hand side is undefined if timing control is not specified. The nonblocking assignment operator is the same operator as the less-than-or-equal-to relational operator. The interpretation shall be decided from the context in which <= appears. When <= is used in an expression, it shall be interpreted as a relational operator, and when it is used in a nonblocking procedural assignment, it shall be interpreted as an assignment operator. Copyright © 2001 IEEE. All rights reserved. 121 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® The nonblocking procedural assignments shall be evaluated in two steps as discussed in Clause 5. These two steps are shown in the following example. Example 1: module evaluates2 (out); At posedge c, the simulator output out; evaluates the right-hand sides reg a, b, c; Step 1: of the nonblocking assignments and schedules the initial begin a = 0; b = 1; assignments of the new values at the end of the nonblocking assign update events (see 5.4). c = 0; end When the simulator activates Step 2: the nonblocking assign update always c = #5 ~c; events, the simulator updates the left-hand side of each non- always @(posedge c) begin blocking assignment statement. a <= b; // evaluates, schedules, b <= a; // and executes in two steps end endmodule Nonblocking assignment schedules changes at time 5 a=0 b=1 Assignment values are: a=1 b=0 At the end of the time step means that the nonblocking assignments are the last assignments executed in a time step—with one exception. Nonblocking assignment events can create blocking assignment events. These blocking assignment events shall be processed after the scheduled nonblocking events. Unlike an event or delay control for blocking assignments, the nonblocking assignment does not block the procedural flow. The nonblocking assignment evaluates and schedules the assignment, but it does not block the execution of subsequent statements in a begin-end block. Example 2: //non_block1.v module non_block1; reg a, b, c, d, e, f; //blocking assignments initial begin a = #10 1; // a will be assigned 1 at time 10 b = #2 0; // b will be assigned 0 at time 12 c = #4 1; // c will be assigned 1 at time 16 end //non-blocking assignments initial begin d <= #10 1; // d will be assigned 1 at time 10 e <= #2 0; // e will be assigned 0 at time 2 f <= #4 1; // f will be assigned 1 at time 4 end endmodule scheduled changes at time 2 e=0 scheduled changes at time 4 f=1 scheduled changes at time 10 d=1 As shown in the previous example, the simulator evaluates and schedules assignments for the end of the current time step and can perform swapping operations with the nonblocking procedural assignments. 122 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Example 3: //non_block1.v module non_block1; Step 1: The simulator evaluates the righthand side of the nonblocking reg a, b; assignments and schedules the initial begin a = 0; assignments for the end of the current time step. b = 1; a <= b; // evaluates, schedules, and b <= a; // executes in two steps Step 2: end initial begin $monitor ($time, ,"a = %b b = %b", a, b); #100 $finish; At the end of the current time step, the simulator updates the left-hand side of each nonblocking assignment statement. end endmodule assignment values are: a=1 b=0 The order of the execution of distinct nonblocking assignments to a given variable shall be preserved. This means that if there is clear ordering of the execution of a set of nonblocking assigments, then the order of the resulting updates of the destination of the nonblocking assignments shall be the same as the ordering of the execution. Example 4: module multiple2; reg a; initial a = 1; // The assigned value of the reg is determinate initial begin a <= #4 0; // schedules a = 0 at time 4 a <= #4 1; // schedules a = 1 at time 4 end // At time 4, a = 1 endmodule If the simulator executes two procedural blocks concurrently, and if these procedural blocks contain nonblocking assignment operators to the same variable, the final value of that variable is indeterminate. For example, the value of reg a is indeterminate in the following example. Example 5: module multiple3 ; reg a; initial a = 1; initial a <= #4 0; initial a <= #4 1; // schedules 0 at time 4 // schedules 1 at time 4 // At time 4, a = ?? // The assigned value of the reg is indeterminate endmodule Copyright © 2001 IEEE. All rights reserved. 123 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® NOTE—The fact that two nonblocking assignments targeting the same variable are in different blocks is not by itself sufficient to make the order of assignments to a variable indeterminate. For example, the value of reg a at the end of time cycle 16 is determinate in the following example: module multiple2 ; reg a; initial #8 a <= #8 1;// executed at time 8; schedules // an update of 1 at time 16 initial #12 a <= #4 0;// executed at time 12; schedules // an update of 0 at time 16 // Because it is determinate that the update of a to // the value 1 is scheduled before the update of a to // the value 0, then it is determinate that a will have the // value 0 at the end of time slot 16. endmodule The following example shows how the value of i[0] is assigned to r1 and how the assignments are scheduled to occur after each time delay. Example 6: module multiple; reg r1; reg [2:0] i; initial begin // starts at time 0, does not hold the block r1 = 0; // makes assignments to r1 without cancelling previous assignments for (i = 0; i <= 5; i = i+1) r1 <= # (i*10) i[0]; end endmodule r1 0 10 20 30 40 50 9.3 Procedural continuous assignments The procedural continuous assignments (using keywords assign and force) are procedural statements that allow expressions to be driven continuously onto variables or nets. The syntax for these statements is given in Syntax 9-3. 124 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C net_assignment ::= (From Annex A - A.6.1) net_lvalue = expression procedural_continuous_assignments ::= (From Annex A - A.6.2) assign variable_assignment | deassign variable_lvalue | force variable_assignment | force net_assignment | release variable_lvalue | release net_lvalue variable_assignment ::= (From Annex A - A.6.3) variable_lvalue = expression net_lvalue ::= (From Annex A - A.8.5) hierarchical_net_identifier | hierarchical_net_identifier [ constant_expression ] { [ constant_expression ] } | hierarchical_net_identifier [ constant_expression ] { [ constant_expression ] } [ constant_range_expression ] | hierarchical_net_identifier [ constant_range_expression ] | net_concatenation variable_lvalue ::= hierarchical_variable_identifier | hierarchical_variable_identifier [ expression ] { [ expression ] } | hierarchical_variable_identifier [ expression ] { [ expression ] } [ range_expression ] | hierarchical_variable_identifier [ range_expression ] | variable_concatenation Syntax 9-3—Syntax for procedural continuous assignments The left-hand side of the assignment in the assign statement shall be a variable reference or a concatenation of variables. It shall not be a memory word (array reference) or a bit-select or a part-select of a variable. In contrast, the left-hand side of the assignment in the force statement can be a variable reference or a net reference. It can be a concatenation of any of the above. Bit-selects and part-selects of vector variables are not allowed. 9.3.1 The assign and deassign procedural statements The assign procedural continuous assignment statement shall override all procedural assignments to a variable. The deassign procedural statement shall end a procedural continuous assignment to a variable. The value of the variable shall remain the same until the reg is assigned a new value through a procedural assignment or a procedural continuous assignment. The assign and deassign procedural statements allow, for example, modeling of asynchronous clear/preset on a D-type edge-triggered flip-flop, where the clock is inhibited when the clear or preset is active. If the keyword assign is applied to a variable for which there is already a procedural continuous assignment, then this new procedural continuous assignment shall deassign the variable before making the new procedural continuous assignment. Copyright © 2001 IEEE. All rights reserved. 125 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Example: The following example shows a use of the assign and deassign procedural statements in a behavioral description of a D-type flip-flop with preset and clear inputs. module dff (q, d, clear, preset, clock); output q; input d, clear, preset, clock; reg q; always @(clear or preset) if (!clear) assign q = 0; else if (!preset) assign q = 1; else deassign q; always @(posedge clock) q = d; endmodule If either clear or preset is low, then the output q will be held continuously to the appropriate constant value and a positive edge on the clock will not affect q. When both the clear and preset are high, then q is deassigned. 9.3.2 The force and release procedural statements Another form of procedural continuous assignment is provided by the force and release procedural statements. These statements have a similar effect to the assign-deassign pair, but a force can be applied to nets as well as to variables. The left-hand side of the assignment can be a variable, a net, a constant bit-select of a vector net, a part-select of a vector net, or a concatenation. It cannot be a memory word (array reference) or a bit-select or a part-select of a vector variable. A force statement to a variable shall override a procedural assignment or procedural continuous assignment that takes place on the variable until a release procedural statement is executed on the variable. After the release procedural statement is executed, the variable shall not immediately change value (as would a net that is assigned with a procedural continuous assignment). The value specified in the force statement shall be maintained in the variable until the next procedural assignment takes place, except in the case where a procedural continuous assignment is active on the variable. A force procedural statement on a net overrides all drivers of the net—gate outputs, module outputs, and continuous assignments—until a release procedural statement is executed on the net. Releasing a variable that currently has an active procedural continuous assignment shall re-establish that assignment. 126 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE Example: IEEE Std 1364-2001 Version C module test; reg a, b, c, d; wire e; and and1 (e, a, b, c); initial begin $monitor("%d d=%b,e=%b", $stime, d, e); assign d = a & b & c; a = 1; b = 0; c = 1; #10; force d = (a | b | c); force e = (a | b | c); #10 $stop; release d; release e; #10 $finish; end endmodule Results: 0 d=0,e=0 10 d=1,e=1 20 d=0,e=0 In this example, an and gate instance and1 is “patched” as an or gate by a force procedural statement that forces its output to the value of its logical or inputs, and an assign procedural statement of logical and values is “patched” as an assign procedural statement of logical or values. The right-hand side of a procedural continuous assignment or a force statement can be an expression. This shall be treated just as a continuous assignment; that is, if any variable on the right-hand side of the assignment changes, the assignment shall be re-evaluated while the assign or force is in effect. For example: force a = b + f(c) ; Here, if b changes or c changes, a will be forced to the new value of the expression b+f(c). 9.4 Conditional statement The conditional statement (or if-else statement) is used to make a decision as to whether a statement is executed or not. Formally, the syntax is given in Syntax 9-4. Copyright © 2001 IEEE. All rights reserved. 127 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® conditional_statement ::= (From Annex A - A.6.6) if ( expression ) statement_or_null [ else statement_or_null ] | if_else_if_statement function_conditional_statement ::= (From Annex A - A.6.6) if ( expression ) function_statement_or_null [ else function_statement_or_null ] | function_if_else_if_statement Syntax 9-4—Syntax of if statement If the expression evaluates to true (that is, has a nonzero known value), the first statement shall be executed. If it evaluates to false (has a zero value or the value is x or z), the first statement shall not execute. If there is an else statement and expression is false, the else statement shall be executed. Since the numeric value of the if expression is tested for being zero, certain shortcuts are possible. For example, the following two statements express the same logic: if (expression) if (expression != 0) Because the else part of an if-else is optional, there can be confusion when an else is omitted from a nested if sequence. This is resolved by always associating the else with the closest previous if that lacks an else. In the example below, the else goes with the inner if, as shown by indentation. if (index > 0) if (rega > regb) result = rega; else // else applies to preceding if result = regb; If that association is not desired, a begin-end block statement shall be used to force the proper association, as shown below. if (index > 0) begin if (rega > regb) result = rega; end else result = regb; 9.4.1 If-else-if construct The following construction occurs so often that it is worth a brief separate discussion: 128 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C if_else_if_statement ::= (From Annex A - A.6.6) if ( expression ) statement_or_null { else if ( expression ) statement_or_null } [ else statement_or_null ] function_if_else_if_statement ::= (From Annex A - A.6.6) if ( expression ) function_statement_or_null { else if ( expression ) function_statement_or_null } [ else function_statement_or_null ] Syntax 9-5—Syntax of if-else-if construct This sequence of if statements (known as an if-else-if construct) is the most general way of writing a multiway decision. The expressions shall be evaluated in order; if any expression is true, the statement associated with it shall be executed, and this shall terminate the whole chain. Each statement is either a single statement or a block of statements. The last else part of the if-else-if construct handles the none-of-the-above or default case where none of the other conditions were satisfied. Sometimes there is no explicit action for the default; in that case, the trailing else statement can be omitted or it can be used for error checking to catch an impossible condition. Copyright © 2001 IEEE. All rights reserved. 129 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Example: The following module fragment uses the if-else statement to test the variable index to decide whether one of three modify_segn regs has to be added to the memory address, and which increment is to be added to the index reg. The first ten lines declare the regs and parameters. // declare regs and parameters reg [31:0] instruction, segment_area[255:0]; reg [7:0] index; reg [5:0] modify_seg1, modify_seg2, modify_seg3; parameter segment1 = 0, inc_seg1 = 1, segment2 = 20, inc_seg2 = 2, segment3 = 64, inc_seg3 = 4, data = 128; // test the index variable if (index < segment2) begin instruction = segment_area [index + modify_seg1]; index = index + inc_seg1; end else if (index < segment3) begin instruction = segment_area [index + modify_seg2]; index = index + inc_seg2; end else if (index < data) begin instruction = segment_area [index + modify_seg3]; index = index + inc_seg3; end else instruction = segment_area [index]; 9.5 Case statement The case statement is a multiway decision statement that tests whether an expression matches one of a number of other expressions and branches accordingly. The case statement has the syntax shown in Syntax 9-6. 130 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C case_statement ::= (From Annex A - A.6.7) case ( expression ) case_item { case_item } endcase | casez ( expression ) case_item { case_item } endcase | casex ( expression ) case_item { case_item } endcase case_item ::= expression { , expression } : statement_or_null | default [ : ] statement_or_null function_case_statement ::= case ( expression ) function_case_item { function_case_item } endcase | casez ( expression ) function_case_item { function_case_item } endcase | casex ( expression ) function_case_item { function_case_item } endcase function_case_item ::= expression { , expression } : function_statement_or_null | default [ : ] function_statement_or_null Syntax 9-6—Syntax for case statement The default statement shall be optional. Use of multiple default statements in one case statement shall be illegal. The case expression and the case item expression can be computed at runtime; neither expression is required to be a constant expression. Examples: A simple example of the use of the case statement is the decoding of reg rega to produce a value for result as follows: reg [15:0] rega; reg [9:0] result; case (rega) 16’d0: result = 10’b0111111111; 16’d1: result = 10’b1011111111; 16’d2: result = 10’b1101111111; 16’d3: result = 10’b1110111111; 16’d4: result = 10’b1111011111; 16’d5: result = 10’b1111101111; 16’d6: result = 10’b1111110111; 16’d7: result = 10’b1111111011; 16’d8: result = 10’b1111111101; 16’d9: result = 10’b1111111110; default result = ’bx; endcase Copyright © 2001 IEEE. All rights reserved. 131 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® The case item expressions shall be evaluated and compared in the exact order in which they are given. During the linear search, if one of the case item expressions matches the case expression given in parentheses, then the statement associated with that case item shall be executed. If all comparisons fail and the default item is given, then the default item statement shall be executed. If the default statement is not given and all of the comparisons fail, then none of the case item statements shall be executed. Apart from syntax, the case statement differs from the multiway if-else-if construct in two important ways: a) The conditional expressions in the if-else-if construct are more general than comparing one expression with several others, as in the case statement. b) The case statement provides a definitive result when there are x and z values in an expression. In a case expression comparison, the comparison only succeeds when each bit matches exactly with respect to the values 0, 1, x, and z. As a consequence, care is needed in specifying the expressions in the case statement. The bit length of all the expressions shall be equal so that exact bit-wise matching can be performed. The length of all the case item expressions, as well as the case expression in the parentheses, shall be made equal to the length of the longest case expression and case item expression. NOTE—The default length of x and z is the same as the default length of an integer. The reason for providing a case expression comparison that handles the x and z values is that it provides a mechanism for detecting such values and reducing the pessimism that can be generated by their presence. Examples: Example 1—The following example illustrates the use of a case statement to handle x and z values properly. case (select[1:2]) 2’b00: result = 0; 2’b01: result = flaga; 2’b0x, 2’b0z: result = flaga ? ’bx : 0; 2’b10: result = flagb; 2’bx0, 2’bz0: result = flagb ? ’bx : 0; default result = ’bx; endcase In this example, if select[1] is 0 and flaga is 0, then whether the value of select[2] is x or z, result should be 0—which is resolved by the third case. Example 2—The following example shows another way to use a case statement to detect x and z values. case (sig) 1’bz: $display("signal is floating"); 1’bx: $display("signal is unknown"); default: $display("signal is %b", sig); endcase 132 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 9.5.1 Case statement with don’t-cares Two other types of case statements are provided to allow handling of don’t-care conditions in the case comparisons. One of these treats high-impedance values (z) as don’t-cares, and the other treats both high-impedance and unknown (x) values as don’t-cares. These case statements can be used in the same way as the traditional case statement, but they begin with keywords casez and casex respectively. Don’t-care values (z values for casez, z and x values for casex) in any bit of either the case expression or the case items shall be treated as don’t-care conditions during the comparison, and that bit position shall not be considered. The don’t-care conditions in case expression can be used to control dynamically which bits should be compared at any time. The syntax of literal numbers allows the use of the question mark (?) in place of z in these case statements. This provides a convenient format for specification of don’t-care bits in case statements. Examples: Example 1—The following is an example of the casez statement. It demonstrates an instruction decode, where values of the most significant bits select which task should be called. If the most significant bit of ir is a 1, then the task instruction1 is called, regardless of the values of the other bits of ir. reg [7:0] ir; casez (ir) 8’b1???????: instruction1(ir); 8’b01??????: instruction2(ir); 8’b00010???: instruction3(ir); 8’b000001??: instruction4(ir); endcase Example 2—The following is an example of the casex statement. It demonstrates an extreme case of how don’t-care conditions can be dynamically controlled during simulation. In this case, if r = 8´b01100110, then the task stat2 is called. reg [7:0] r, mask; mask = 8’bx0x0x0x0; casex (r ^ mask) 8’b001100xx: stat1; 8’b1100xx00: stat2; 8’b00xx0011: stat3; 8’bxx010100: stat4; endcase 9.5.2 Constant expression in case statement A constant expression can be used for case expression. The value of the constant expression shall be compared against case item expressions. Copyright © 2001 IEEE. All rights reserved. 133 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Example: The following example demonstrates the usage by modeling a 3-bit priority encoder. reg [2:0] encode ; case (1) encode[2] : $display(“Select Line 2”) ; encode[1] : $display(“Select Line 1”) ; encode[0] : $display(“Select Line 0”) ; default $display(“Error: One of the bits expected ON”); endcase Note that the case expression is a constant expression (1). The case items are expressions (bit-selects) and are compared against the constant expression for a match. 9.6 Looping statements There are four types of looping statements. These statements provide a means of controlling the execution of a statement zero, one, or more times. forever repeat while for Continuously executes a statement. Executes a statement a fixed number of times. If the expression evaluates to unknown or high impedance, it shall be treated as zero, and no statement shall be executed. Executes a statement until an expression becomes false. If the expression starts out false, the statement shall not be executed at all. Controls execution of its associated statement(s) by a three-step process, as follows: a) Executes an assignment normally used to initialize a variable that controls the number of loops executed. b) Evaluates an expression—if the result is zero, the for-loop shall exit, and if it is not zero, the for-loop shall execute its associated statement(s) and then perform step c. If the expression evaluates to an unknown or high-impedance value, it shall be treated as zero. c) Executes an assignment normally used to modify the value of the loop-control variable, then repeats step b. 134 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE Syntax 9-7 shows the syntax for various looping statements. IEEE Std 1364-2001 Version C function_loop_statement ::= (From Annex A - A.6.8) forever function_statement | repeat ( expression ) function_statement | while ( expression ) function_statement | for ( variable_assignment ; expression ; variable_assignment ) function_statement loop_statement ::= forever statement | repeat ( expression ) statement | while ( expression ) statement | for ( variable_assignment ; expression ; variable_assignment ) statement Syntax 9-7—Syntax for the looping statements The rest of this clause presents examples for three of the looping statements. The forever loop should only be used in conjunction with the timing controls or the disable statement, therefore, this example is presented in 9.7.2. Examples: Example 1—Repeat statement: In the following example of a repeat loop, add and shift operators implement a multiplier. parameter size = 8, longsize = 16; reg [size:1] opa, opb; reg [longsize:1] result; begin : mult reg [longsize:1] shift_opa, shift_opb; shift_opa = opa; shift_opb = opb; result = 0; repeat (size) begin if (shift_opb[1]) result = result + shift_opa; shift_opa = shift_opa << 1; shift_opb = shift_opb >> 1; end end Copyright © 2001 IEEE. All rights reserved. 135 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Example 2—While statement: The following example counts the number of logic 1 values in rega. begin : count1s reg [7:0] tempreg; count = 0; tempreg = rega; while (tempreg) begin if (tempreg[0]) count = count + 1; tempreg = tempreg >> 1; end end Example 3—For statement: The for statement accomplishes the same results as the following pseudo-code that is based on the while loop: begin initial_assignment; while (condition) begin statement step_assignment; end end The for loop implements this logic while using only two lines, as shown in the pseudo-code below. for (initial_assignment; condition; step_assignment) statement 9.7 Procedural timing controls The Verilog HDL has two types of explicit timing control over when procedural statements can occur. The first type is a delay control, in which an expression specifies the time duration between initially encountering the statement and when the statement actually executes. The delay expression can be a dynamic function of the state of the circuit, but it can be a simple number that separates statement executions in time. The delay control is an important feature when specifying stimulus waveform descriptions. It is described in 9.7.1 and 9.7.7. The second type of timing control is the event expression, which allows statement execution to be delayed until the occurrence of some simulation event occurring in a procedure executing concurrently with this procedure. A simulation event can be a change of value on a net or variable (an implicit event) or the occurrence of an explicitly named event that is triggered from other procedures (an explicit event). Most often, an event control is a positive or negative edge on a clock signal. Event control is discussed in 9.7.2 through 9.7.7. The procedural statements encountered so far all execute without advancing simulation time. Simulation time can advance by one of the following three methods: — A delay control, which is introduced by the symbol # — An event control, which is introduced by the symbol @ — The wait statement, which operates like a combination of the event control and the while loop Syntax 9-8 describes timing control in procedural statements. 136 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C delay_control ::= (From Annex A - A.6.5) # delay_value | # ( mintypmax_expression ) delay_or_event_control ::= delay_control | event_control | repeat ( expression ) event_control event_control ::= @ event_identifier | @ ( event_expression ) | @* | @ (*) event_expression ::= expression | hierarchical_identifier | posedge expression | negedge expression | event_expression or event_expression | event_expression , event_expression Syntax 9-8—Syntax for procedural timing control The gate and net delays also advance simulation time, as discussed in Clause 6. The next subclauses discuss the three procedural timing control methods. 9.7.1 Delay control A procedural statement following the delay control shall be delayed in its execution with respect to the procedural statement preceding the delay control by the specified delay. If the delay expression evaluates to an unknown or high-impedance value, it shall be interpreted as zero delay. If the delay expression evaluates to a negative value, it shall be interpreted as a 2’s complement unsigned integer of the same size as a time variable. Specify parameters are permitted in the delay expression. They may be overridden by SDF annotation, in which case the expression is reevaluated. Examples: Example 1—The following example delays the execution of the assignment by 10 time units: #10 rega = regb; Example 2—The next three examples provide an expression following the number sign (#). Execution of the assignment is delayed by the amount of simulation time specified by the value of the expression. #d rega = regb; // d is defined as a parameter #((d+e)/2) rega = regb;// delay is average of d and e #regr regr = regr + 1; // delay is the value in regr Copyright © 2001 IEEE. All rights reserved. 137 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 9.7.2 Event control The execution of a procedural statement can be synchronized with a value change on a net or variable or the occurrence of a declared event. The value changes on nets and variable can be used as events to trigger the execution of a statement. This is known as detecting an implicit event. The event can also be based on the direction of the change—that is, towards the value 1 (posedge) or towards the value 0 (negedge). The behavior of posedge and negedge events is shown in Table 43 and can be described as follows: — A negedge shall be detected on the transition from 1 to x, z, or 0, and from x or z to 0 — A posedge shall be detected on the transition from 0 to x, z, or 1, and from x or z to 1 Table 43—Detecting posedge and negedge To 0 1 x z From 0 1 x z No edge posedge posedge posedge negedge No edge negedge negedge negedge posedge No edge No edge negedge posedge No edge No edge If the expression evaluates to more than a 1-bit result, the edge transition shall be detected on the least significant bit of the result. The change of value in any of the operands without a change in the value of the least significant bit of the expression result shall not be detected as an edge. Example: The following example shows illustrations of edge-controlled statements. @r rega = regb; // controlled by any value change in the reg r @(posedge clock) rega = regb; // controlled by posedge on clock forever @(negedge clock) rega = regb; // controlled by negative edge 9.7.3 Named events A new data type, in addition to nets and variables, called “event” can be declared. An identifier declared as an event data type is called a named event. A named event can be triggered explicitly. It can be used in an event expression to control the execution of procedural statements in the same manner as event controls described in 9.7.2. Named events can be made to occur from a procedure. This allows control over the enabling of multiple actions in other procedures. An event name shall be declared explicitly before it is used. Syntax 9-9 gives the syntax for declaring events. 138 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C event_declaration ::= (From Annex A - A.2.1.3) event list_of_event_identifiers ; list_of_event_identifiers ::= (From Annex A - A.2.3) event_identifier [ dimension { dimension }] { , event_identifier [ dimension { dimension }] } dimension ::= (From Annex A - A.2.5) [ dimension_constant_expression : dimension_constant_expression ] Syntax 9-9—Syntax for event declaration An event shall not hold any data. The following are the characteristics of a named event: — It can be made to occur at any particular time — It has no time duration — Its occurrence can be recognized by using the event control syntax described in 9.7.2. A declared event is made to occur by the activation of an event triggering statement with the syntax given in Syntax 9-10. An event is not made to occur by changing the index of an event array in an event control expression. event_trigger ::= (From Annex A - A.6.5) -> hierarchical_event_identifier ; Syntax 9-10—Syntax for event trigger An event-controlled statement (for example, @trig rega = regb;) shall cause simulation of its containing procedure to wait until some other procedure executes the appropriate event-triggering statement (for example, -> trig). Named events and event control give a powerful and efficient means of describing the communication between, and synchronization of, two or more concurrently active processes. A basic example of this is a small waveform clock generator that synchronizes control of a synchronous circuit by signalling the occurrence of an explicit event periodically while the circuit waits for the event to occur. 9.7.4 Event or operator The logical or of any number of events can be expressed such that the occurrence of any one of the events triggers the execution of the procedural statement that follows it. The keyword or or a comma character (,) is used as an event logical or operator. A combination of these can be used in the same event expression. Comma-separated sensitivity lists shall be synonymous to or-separated sensitivity lists. Copyright © 2001 IEEE. All rights reserved. 139 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Examples: The next two examples show the logical or of two and three events respectively. @(trig or enable) rega = regb; // controlled by trig or enable @(posedge clk_a or posedge clk_b or trig) rega = regb; The following examples show the use of the comma (,) as an event logical or operator. always @(a, b, c, d, e) always @(posedge clk, negedge rstn) always @(a or b, c, d or e) 9.7.5 Implicit event_expression list The event_expression list of an event control is a common source of bugs in RTL simulations. Users tend to forget to add some of the nets or variables read in the timing control statement. This is often found when comparing RTL and gate level versions of a design. The implicit event_expression, @*, is a convenient shorthand that eliminates these problems by adding all nets and variables which are read by the statement (which can be a statement group) of a procedural_timing_control_statement to the event_expression. All net and variable identifiers which appear in the statement will be automatically added to the event expression with these exceptions: — Identifiers which only appear in wait or event expressions. — Identifiers which only appear as a hierarchical_variable_identifier in the variable_lvalue of the left-hand side of assignments. Nets and variables which appear on the right-hand side of assignments, in function and task calls, in case and conditional expressions, as an index variable on the left-hand side of assignments, or as variables in case item expressions, shall all be included by these rules. Examples: Example 1 always @(*) // equivalent to @(a or b or c or d or f) y = (a & b) | (c & d) | myfunction(f); Example 2 always @* begin // equivalent to @(a or b or c or d or tmp1 or tmp2) tmp1 = a & b; tmp2 = c & d; y = tmp1 | tmp2; end Example 3 always @* begin // equivalent to @(b) @(i) kid = b; // i is not added to @* end 140 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Example 4 always @* begin // equivalent to @(a or b or c or d) x = a ^ b; @* // equivalent to @(c or d) x = c ^ d; end Example 5 always @* begin // same as @(a or en) y = 8’hff; y[a] = !en; end Example 6 always @* begin // same as @(state or go or ws) next = 4’b0; case (1’b1) state[IDLE]: if (go) next[READ] = 1’b1; else next[IDLE] = 1’b1; state[READ]: next[DLY ] = 1’b1; state[DLY ]: if (!ws) next[DONE] = 1’b1; else next[READ] = 1’b1; state[DONE]: next[IDLE] = 1’b1; endcase end 9.7.6 Level-sensitive event control The execution of a procedural statement can also be delayed until a condition becomes true. This is accomplished using the wait statement, which is a special form of event control. The nature of the wait statement is level-sensitive, as opposed to basic event control (specified by the @ character), which is edge-sensitive. The wait statement shall evaluate a condition, and, if it is false, the procedural statements following the wait statement shall remain blocked until that condition becomes true before continuing. The wait statement has the form given in Syntax 9-11. wait_statement ::= (From Annex A - A.6.5) wait ( expression ) statement_or_null Syntax 9-11—Syntax for wait statement Example: The following example shows the use of the wait statement to accomplish level-sensitive event control. begin wait (!enable) #10 a = b; #10 c = d; end If the value of enable is 1 when the block is entered, the wait statement will delay the evaluation of the Copyright © 2001 IEEE. All rights reserved. 141 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® next statement (#10 a = b;) until the value of enable changes to 0. If enable is already 0 when the begin-end block is entered, then the assignment “a = b;” is evaluated after a delay of 10 and no additional delay occurs. 9.7.7 Intra-assignment timing controls The delay and event control constructs previously described precede a statement and delay its execution. In contrast, the intra-assignment delay and event controls are contained within an assignment statement and modify the flow of activity in a different way. This subclause describes the purpose of intra-assignment timing controls and the repeat timing control that can be used in intra-assignment delays. An intra-assignment delay or event control shall delay the assignment of the new value to the left-hand side, but the right-hand side expression shall be evaluated before the delay, instead of after the delay. The syntax for intra-assignment delay and event control is given in Syntax 9-12. blocking_assignment ::= (From Annex A - A.6.2) variable_lvalue = [ delay_or_event_control ] expression nonblocking_assignment ::= variable_lvalue <= [ delay_or_event_control ] expression delay_control ::= (From Annex A - A.6.5) # delay_value | # ( mintypmax_expression ) delay_or_event_control ::= delay_control | event_control | repeat ( expression ) event_control event_control ::= @ event_identifier | @ ( event_expression ) | @* | @ (*) event_expression ::= expression | hierarchical_identifier | posedge expression | negedge expression | event_expression or event_expression | event_expression , event_expression Syntax 9-12—Syntax for intra-assignment delay and event control The intra-assignment delay and event control can be applied to both blocking assignments and nonblocking assignments. The event expression shall be resolved to a 1-bit value. The repeat event control shall specify an intra-assignment delay of a specified number of occurrences of an event. If the repeat count literal, or signed reg holding the repeat count, is less than or equal to 0 at the time of evaluation, the assignment occurs as if there is no repeat construct. Examples: repeat (-3) @ (event_expression) // will not execute event_expression. 142 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C repeat (a) @ (event_expression) // if a is assigned -3, it will execute the event_expression // if a is declared as an unsigned reg, // but not if it is signed. This construct is convenient when events have to be synchronized with counts of clock signals. Copyright © 2001 IEEE. All rights reserved. 143 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Examples: Table 44 illustrates the philosophy of intra-assignment timing controls by showing the code that could accomplish the same timing effect without using intra-assignment. Table 44—Intra-assignment timing control equivalence Intra-assignment timing control With intra-assignment construct a = #5 b; a = @(posedge clk) b; a = repeat(3) @(posedge clk) b; Without intra-assignment construct begin temp = b; #5 a = temp; end begin temp = b; @(posedge clk) a = temp; end begin temp = b; @(posedge clk); @(posedge clk); @(posedge clk) a = temp; end The next three examples use the fork-join behavioral construct. All statements between the keywords fork and join execute concurrently. This construct is described in more detail in 9.8.2. The following example shows a race condition that could be prevented by using intra-assignment timing control: fork #5 a = b; #5 b = a; join The code in this example samples and sets the values of both a and b at the same simulation time, thereby creating a race condition. The intra-assignment form of timing control used in the next example prevents this race condition. fork // data swap a = #5 b; b = #5 a; join Intra-assignment timing control works because the intra-assignment delay causes the values of a and b to be evaluated before the delay, and the assignments to be made after the delay. Some existing tools that implement intra-assignment timing control use temporary storage in evaluating each expression on the right-hand side. Intra-assignment waiting for events is also effective. In the following example, the right-hand side expressions are evaluated when the assignment statements are encountered, but the assignments are delayed until the rising edge of the clock signal. 144 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C fork // data shift a = @(posedge clk) b; b = @(posedge clk) c; join The following is an example of a repeat event control as the intra-assignment delay of a nonblocking assignment: a <= repeat(5) @(posedge clk) data; Figure 31 illustrates the activities that result from this repeat event control. data is evaluated clk data a Figure 31—Repeat event control utilizing a clock edge In this example, the value of data is evaluated when the assignment is encountered. After five occurrences of posedge clk, a is assigned the value of data. The following is an example of a repeat event control as the intra-assignment delay of a procedural assignment: a = repeat(num) @(clk) data; In this example, the value of data is evaluated when the assignment is encountered. After the number of transitions of clk equals the value of num, a is assigned the value of data. The following is an example of a repeat event control with expressions containing operations to specify both the number of event occurrences and the event that is counted: a <= repeat(a+b) @(posedge phi1 or negedge phi2) data; In this example, the value of data is evaluated when the assignment is encountered. After the sum of the positive edges of phi1 and the negative edges of phi2 equals the sum of a and b, a is assigned the value of data. Even if posedge phi1 and negedge phi2 occurred at the same simulation time, each will be detected separately. Copyright © 2001 IEEE. All rights reserved. 145 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 9.8 Block statements The block statements are a means of grouping two or more statements together so that they act syntactically like a single statement. There are two types of blocks in the Verilog HDL: — Sequential block, also called begin-end block — Parallel block, also called fork-join block The sequential block shall be delimited by the keywords begin and end. The procedural statements in sequential block shall be executed sequentially in the given order. The parallel block shall be delimited by the keywords fork and join. The procedural statements in parallel block shall be executed concurrently. 9.8.1 Sequential blocks A sequential block shall have the following characteristics: — Statements shall be executed in sequence, one after another — Delay values for each statement shall be treated relative to the simulation time of the execution of the previous statement — Control shall pass out of the block after the last statement executes Syntax 9-13 gives the formal syntax for a sequential block. function_seq_block ::= (From Annex A - A.6.3) begin [ : block_identifier { block_item_declaration } ] { function_statement } end seq_block ::= begin [ : block_identifier { block_item_declaration } ] { statement } end block_item_declaration ::= (From Annex A - A.2.8) { attribute_instance } block_reg_declaration | { attribute_instance } event_declaration | { attribute_instance } integer_declaration | { attribute_instance } local_parameter_declaration | { attribute_instance } parameter_declaration | { attribute_instance } real_declaration | { attribute_instance } realtime_declaration | { attribute_instance } time_declaration Syntax 9-13—Syntax for the sequential block Examples: Example 1—A sequential block enables the following two assignments to have a deterministic result: begin areg = breg; creg = areg; end // creg stores the value of breg 146 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The first assignment is performed and areg is updated before control passes to the second assignment. Example 2—Delay control can be used in a sequential block to separate the two assignments in time. begin areg = breg; @(posedge clock) creg = areg; // assignment delayed until end // posedge on clock Example 3—The following example shows how the combination of the sequential block and delay control can be used to specify a time-sequenced waveform. parameter d = 50; // d declared as a parameter and reg [7:0] r; // r declared as an 8-bit reg begin // a waveform controlled by sequential delay #d r = ’h35; #d r = ’hE2; #d r = ’h00; #d r = ’hF7; #d -> end_wave;//trigger an event called end_wave end 9.8.2 Parallel blocks A parallel block shall have the following characteristics: — Statements shall execute concurrently — Delay values for each statement shall be considered relative to the simulation time of entering the block — Delay control can be used to provide time-ordering for assignments — Control shall pass out of the block when the last time-ordered statement executes Syntax 9-14 gives the formal syntax for a parallel block. par_block ::= (From Annex A - A.6.3) fork [ : block_identifier { block_item_declaration } ] { statement } join block_item_declaration ::= (From Annex A - A.2.8) { attribute_instance } block_reg_declaration | { attribute_instance } event_declaration | { attribute_instance } integer_declaration | { attribute_instance } local_parameter_declaration | { attribute_instance } parameter_declaration | { attribute_instance } real_declaration | { attribute_instance } realtime_declaration | { attribute_instance } time_declaration Syntax 9-14—Syntax for the parallel block The timing controls in a fork-join block do not have to be ordered sequentially in time. Copyright © 2001 IEEE. All rights reserved. 147 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Example: The following example codes the waveform description shown in example 3 of 9.8.1 by using a parallel block instead of a sequential block. The waveform produced on the reg is exactly the same for both implementations. fork #50 r = ’h35; #100 r = ’hE2; #150 r = ’h00; #200 r = ’hF7; #250 -> end_wave; join 9.8.3 Block names Both sequential and parallel blocks can be named by adding : name_of_block after the keywords begin or fork. The naming of blocks serves several purposes: — It allows local variables, parameters, and named events to be declared for the block. — It allows the block to be referenced in statements such as the disable statement (see Clause 11). All variables shall be static—that is, a unique location exists for all variables and leaving or entering blocks shall not affect the values stored in them. The block names give a means of uniquely identifying all variables at any simulation time. 9.8.4 Start and finish times Both sequential and procedural blocks have the notion of a start and finish time. For sequential blocks, the start time is when the first statement is executed, and the finish time is when the last statement has been executed. For parallel blocks, the start time is the same for all the statements, and the finish time is when the last time-ordered statement has been executed. Sequential and parallel blocks can be embedded within each other, allowing complex control structures to be expressed easily and with a high degree of structure. When blocks are embedded within each other, the timing of when a block starts and finishes is important. Execution shall not continue to the statement following a block until the finish time for the block has been reached—that is, until the block has completely finished executing. Examples: Example 1—The following example shows the statements from the example in 9.8.2 written in the reverse order and still producing the same waveform. fork #250 -> end_wave; #200 r = ’hF7; #150 r = ’h00; #100 r = ’hE2; #50 r = ’h35; join 148 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Example 2—When an assignment is to be made after two separate events have occurred, known as the joining of events, a fork-join block can be useful. begin fork @Aevent; @Bevent; join areg = breg; end The two events can occur in any order (or even at the same simulation time) and the fork-join block will complete and the assignment will be made. In contrast to this, if the fork-join block was a begin-end block and the Bevent occurred before the Aevent, then the block would be waiting for the next Bevent. Example 3—This example shows two sequential blocks, each of which will execute when its controlling event occurs. Because the event controls are within a fork-join block, they execute in parallel and the sequential blocks can therefore also execute in parallel. fork @enable_a begin #ta wa = 0; #ta wa = 1; #ta wa = 0; end @enable_b begin #tb wb = 1; #tb wb = 0; #tb wb = 1; end join 9.9 Structured procedures All procedures in the Verilog HDL are specified within one of the following four statements: — initial construct — always construct — Task — Function The initial and always constructs are enabled at the beginning of a simulation. The initial construct shall execute only once and its activity shall cease when the statement has finished. In contrast, the always construct shall execute repeatedly. Its activity shall cease only when the simulation is terminated. There shall be no implied order of execution between initial and always constructs. The initial constructs need not be scheduled and executed before the always constructs. There shall be no limit to the number of initial and always constructs that can be defined in a module. Copyright © 2001 IEEE. All rights reserved. 149 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Tasks and functions are procedures that are enabled from one or more places in other procedures. Tasks and functions are described in Clause 10. 9.9.1 Initial construct The syntax for the initial construct is given in Syntax 9-15. initial_construct ::= (From Annex A - A.6.2) initial statement Syntax 9-15—Syntax for initial construct Examples: The following example illustrates use of the initial construct for initialization of variables at the start of simulation. initial begin areg = 0; // initialize a reg for (index = 0; index < size; index = index + 1) memory[index] = 0; //initialize memory word end Another typical usage of the initial construct is specification of waveform descriptions that execute once to provide stimulus to the main part of the circuit being simulated. initial begin inputs = ’b000000; //initialize at time zero #10 inputs = ’b011001; // first pattern #10 inputs = ’b011011; // second pattern #10 inputs = ’b011000; // third pattern #10 inputs = ’b001000; // last pattern end 9.9.2 Always construct The always construct repeats continuously throughout the duration of the simulation. Syntax 9-16 shows the syntax for the always construct. always_construct ::= (From Annex A - A.6.2) always statement Syntax 9-16—Syntax for always construct The always construct, because of its looping nature, is only useful when used in conjunction with some form of timing control. If an always construct has no control for simulation time to advance, it will create a simulation deadlock condition. 150 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The following code, for example, creates a zero-delay infinite loop. always areg = ~areg; Providing a timing control to the above code creates a potentially useful description as shown in the following: always #half_period areg = ~areg; Copyright © 2001 IEEE. All rights reserved. 151 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 10. Tasks and functions Tasks and functions provide the ability to execute common procedures from several different places in a description. They also provide a means of breaking up large procedures into smaller ones to make it easier to read and debug the source descriptions. This clause discusses the differences between tasks and functions, describes how to define and invoke tasks and functions, and presents examples of each. 10.1 Distinctions between tasks and functions The following rules distinguish tasks from functions: — A function shall execute in one simulation time unit; a task can contain time-controlling statements. — A function cannot enable a task; a task can enable other tasks and functions. — A function shall have at least one input type argument and shall not have an output or inout type argument; a task can have zero or more arguments of any type. — A function shall return a single value; a task shall not return a value. The purpose of a function is to respond to an input value by returning a single value. A task can support multiple goals and can calculate multiple result values. However, only the output or inout type arguments pass result values back from the invocation of a task. A function is used as an operand in an expression; the value of that operand is the value returned by the function. Example: Either a task or a function can be defined to switch bytes in a 16-bit word. The task would return the switched word in an output argument, so the source code to enable a task called switch_bytes could look like the following example: switch_bytes (old_word, new_word); The task switch_bytes would take the bytes in old_word, reverse their order, and place the reversed bytes in new_word. A word-switching function would return the switched word as the return value of the function. Thus, the function call for the function switch_bytes could look like the following example: new_word = switch_bytes (old_word); 10.2 Tasks and task enabling A task shall be enabled from a statement that defines the argument values to be passed to the task and the variables that receive the results. Control shall be passed back to the enabling process after the task has completed. Thus, if a task has timing controls inside it, then the time of enabling a task can be different from the time at which the control is returned. A task can enable other tasks, which in turn can enable still other tasks—with no limit on the number of tasks enabled. Regardless of how many tasks have been enabled, control shall not return until all enabled tasks have completed. 152 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 10.2.1 Task declarations The syntax for defining tasks is given in Syntax 10-1. task_declaration ::= (From Annex A - A.2.7) task [ automatic ] task_identifier ; { task_item_declaration } statement endtask | task [ automatic ] task_identifier ( task_port_list ) ; { block_item_declaration } statement endtask task_item_declaration ::= block_item_declaration | { attribute_instance } tf_ input_declaration ; | { attribute_instance } tf_output_declaration ; | { attribute_instance } tf_inout_declaration ; task_port_list ::= task_port_item { , task_port_item } task_port_item ::= { attribute_instance } tf_input_declaration | { attribute_instance } tf_output_declaration | { attribute_instance } tf_inout_declaration tf_input_declaration ::= input [ reg ] [ signed ] [ range ] list_of_port_identifiers | input [ task_port_type ] list_of_port_identifiers tf_output_declaration ::= output [ reg ] [ signed ] [ range ] list_of_port_identifiers | output [ task_port_type ] list_of_port_identifiers tf_inout_declaration ::= inout [ reg ] [ signed ] [ range ] list_of_port_identifiers | inout [ task_port_type ] list_of_port_identifiers task_port_type ::= time | real | realtime | integer block_item_declaration ::= (From Annex A - A.2.8) { attribute_instance } block_reg_declaration | { attribute_instance } event_declaration | { attribute_instance } integer_declaration | { attribute_instance } local_parameter_declaration | { attribute_instance } parameter_declaration | { attribute_instance } real_declaration | { attribute_instance } realtime_declaration | { attribute_instance } time_declaration block_reg_declaration ::= reg [ signed ] [ range ] list_of_block_variable_identifiers ; list_of_block_variable_identifiers ::= block_variable_type { , block_variable_type } block_variable_type ::= variable_identifier | variable_identifier dimension { dimension } Syntax 10-1—Syntax for task declaration Copyright © 2001 IEEE. All rights reserved. 153 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® There are two alternate task declaration syntaxes. The first syntax shall begin with the keyword task, followed by the optional keyword automatic, followed by a name for the task and a semicolon, and ending with the keyword endtask. The keyword automatic declares an automatic task that is reentrant with all the task declarations allocated dynamically for each concurrent task entry. Task item declarations can specify the following: — Input arguments — Output arguments — Inout arguments — All data types that can be declared in a procedural block The second syntax shall begin with the keyword task, followed by a name for the task and a parenthesis enclosed task_port_list. The task_port_list shall consist of zero or more comma separated task_port_items. There shall be a semicolon after the close parenthesis. The task body shall follow and then the keyword endtask. In both syntaxes, the port declarations shall have the same syntax as defined by the tf_input_declaration, tf_output_declaration and tf_inout_declaration, as detailed in Syntax 10-1 above. Tasks without the optional keyword automatic are static tasks, with all declared items being statically allocated. These items shall be shared across all uses of the task executing concurrently. Task with the optional keyword automatic are automatic tasks. All items declared inside automatic tasks are allocated dynamically for each invocation. Automatic task items can not be accessed by hierarchical references. Automatic tasks can be invoked through use of their hierarchical name. 10.2.2 Task enabling and argument passing The task enabling statement shall pass arguments as a comma-separated list of expressions enclosed in parentheses. The formal syntax of the task enabling statement is given in Syntax 10-2. task_enable ::= (From Annex A - A.6.9) hierarchical_task_identifier [ ( expression { , expression } ) ] ; Syntax 10-2—Syntax of the task enabling statement The list of arguments for a task enabling statement shall be optional. If the list of arguments is provided, the list shall be an ordered list of expressions that has to match the order of the list of arguments in the task definition. If an argument in the task is declared as an input, then the corresponding expression can be any expression. The order of evaluation of the expressions in the argument list is undefined. If the argument is declared as an output or an inout, then the expression shall be restricted to an expression that is valid on the left-hand side of a procedural assignment (see 9.2). The following items satisfy this requirement: — reg, integer, real, realtime, and time variables — Memory references — Concatenations of reg, integer, real, realtime and time variables — Concatenations of memory references — Bit-selects and part-selects of reg, integer, and time variables 154 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The execution of the task enabling statement shall pass input values from the expressions listed in the enabling statement to the arguments specified within the task. Execution of the return from the task shall pass values from the task output and inout type arguments to the corresponding variables in the task enabling statement. All arguments to the task shall be passed by value rather than by reference (that is, a pointer to the value). Examples: Example 1—The following example illustrates the basic structure of a task definition with five arguments. task my_task; input a, b; inout c; output d, e; begin . . . // statements that perform the work of the task ... c = foo1; // the assignments that initialize result regs d = foo2; e = foo3; end endtask Or using the second form of a task declaration, the task could be defined as: task my_task (input a, b, inout c, output d, e); begin . . . // statements that perform the work of the task ... c = foo1; // the assignments that initialize result regs d = foo2; e = foo3; end endtask The following statement enables the task: my_task (v, w, x, y, z); The task enabling arguments (v, w, x, y, and z) correspond to the arguments (a, b, c, d, and e) defined by the task. At task enabling time, the input and inout type arguments (a, b, and c) receive the values passed in v, w, and x. Thus, execution of the task enabling call effectively causes the following assignments: a = v; b = w; c = x; As part of the processing of the task, the task definition for my_task shall place the computed result values into c, d, and e. When the task completes, the following assignments to return the computed values to the calling process are performed: x = c; y = d; z = e; Copyright © 2001 IEEE. All rights reserved. 155 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Example 2—The following example illustrates the use of tasks by describing a traffic light sequencer: module traffic_lights; reg clock, red, amber, green; parameter on = 1, off = 0, red_tics = 350, amber_tics = 30, green_tics = 200; // initialize colors. initial red = off; initial amber = off; initial green = off; always begin red = on; light(red, red_tics); green = on; light(green, green_tics); amber = on; light(amber, amber_tics); end // sequence to control the lights. // turn red light on // and wait. // turn green light on // and wait. // turn amber light on // and wait. // task to wait for ’tics’ positive edge clocks // before turning ’color’ light off. task light; output color; input [31:0] tics; begin repeat (tics) @ (posedge clock); color = off; // turn light off. end endtask always begin #100 clock = 0; #100 clock = 1; end endmodule // traffic_lights. // waveform for the clock. 10.2.3 Task memory usage and concurrent activation A task may be enabled more than once concurrently. All variables of an automatic task shall be replicated on each concurrent task invocation to store state specific to that invocation. All variables of a static task shall be static in that there shall be a single variable corresponding to each declared local variable in a module instance, regardless of the number of concurrent activations of the task. However, static tasks in different instances of a module shall have separate storage from each other. Variables declared in static tasks shall retain their values between invocations. They shall be initialized to the default initialization value as described in 3.2.2. Variables declared in automatic tasks shall be initialized to the default initialization value whenever execution enters their scope. Because variables declared in automatic tasks are deallocated at the end of the task invocation, they shall not be used in certain constructs that might refer to them after that point. 156 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C — They shall not be assigned values using nonblocking assignments or procedural continuous assignments. — They shall not be referenced by procedural continuous assignments or procedural force statements. — They shall not be referenced in intra-assignment event controls of nonblocking assignments. — They shall not be traced with system tasks such as $monitor and $dumpvars. 10.3 Functions and function calling The purpose of a function is to return a value that is to be used in an expression. The rest of this clause explains how to define and use functions. Copyright © 2001 IEEE. All rights reserved. 157 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 10.3.1 Function declarations The syntax for defining a function is given in Syntax 10-3. function_declaration ::= (From Annex A - A.2.6) function [ automatic ] [ signed ] [ range_or_type ] function_identifier ; function_item_declaration { function_item_declaration } function_statement endfunction | function [ automatic ] [ signed ] [ range_or_type ] function_identifier ( function_port_list ) ; block_item_declaration { block_item_declaration } function_statement endfunction function_item_declaration ::= block_item_declaration | tf_input_declaration ; function_port_list ::= { attribute_instance } tf_input_declaration { , { attribute_instance }tf_input_declaration } tf_input_declaration ::= input [ reg ] [ signed ] [ range ] list_of_port_identifiers | input [ task_port_type ] list_of_port_identifiers range_or_type ::= range | integer | real | realtime | time block_item_declaration ::= (From Annex A - A.2.8) { attribute_instance } block_reg_declaration | { attribute_instance } event_declaration | { attribute_instance } integer_declaration | { attribute_instance } local_parameter_declaration | { attribute_instance } parameter_declaration | { attribute_instance } real_declaration | { attribute_instance } realtime_declaration | { attribute_instance } time_declaration block_reg_declaration ::= reg [ signed ] [ range ] list_of_block_variable_identifiers ; list_of_block_variable_identifiers ::= block_variable_type { , block_variable_type } block_variable_type ::= variable_identifier | variable_identifier dimension { dimension } Syntax 10-3—Syntax for function declaration A function definition shall begin with the keyword function, followed by the optional keyword automatic, followed by the optional signed designator, followed by the range or type of the return value from the func- tion, followed by the name of the function, and then either a semicolon, or a function port list enclosed in parenthesis, and then a semicolon, and then shall end with the keyword endfunction. The use of a range_or_type shall be optional. A function specified without a range or type defaults to a one bit reg for the return value. If used, range_or_type shall specify the return value of the function is a real, an integer, a time, a realtime or a value with a range of [n:m] bits. A function shall have at least one input declared. 158 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The keyword automatic declares a recursive function with all the function declarations allocated dynamically for each recursive call. Automatic function items can not be accessed by hierarchical references. Automatic functions can be invoked through the use of their hierarchical name. Function inputs shall be declared one of two ways. The first method shall have the name of the function followed by a semicolon. After the semicolon one or more input declarations optionally mixed with block item declarations shall follow. After the function item declarations there shall be a behavioral statement and then the endfunction keyword. The second method shall have the name of the function, followed by an open parenthesis, and one or more input declarations, separated by commas. After all the input declarations, there shall be a close parenthesis, and a semicolon. After the semicolon, there shall be zero or more block item declarations, followed by a behavioral statement, and then the endfunction keyword. Example: The following example defines a function called getbyte, using a range specification. function [7:0] getbyte; input [15:0] address; begin // code to extract low-order byte from addressed word ... getbyte = result_expression; end endfunction Or using the second form of a function declaration, the function could be defined as: function [7:0] getbyte (input [15:0] address); begin // code to extract low-order byte from addressed word ... getbyte = result_expression; end endfunction 10.3.2 Returning a value from a function The function definition shall implicitly declare a variable, internal to the function, with the same name as the function. This variable either defaults to a 1-bit reg or is the same type as the type specified in the function declaration. The function definition initializes the return value from the function by assigning the function result to the internal variable with the same name as the function. It is illegal to declare another object with the same name as the function in the scope where the function is declared. Inside a function, there is an implied variable with the name of the function, which may be used in expressions within the function. It is, therefore, also illegal to declare another object with the same name as the function inside the function scope. The following line from the example in 10.3.1 illustrates this concept: getbyte = result_expression; Copyright © 2001 IEEE. All rights reserved. 159 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 10.3.3 Calling a function A function call is an operand within an expression. The function call has the syntax given in Syntax 10-4. function_call ::= (From Annex A - A.8.2) hierarchical_function_identifier{ attribute_instance } ( expression { , expression } ) Syntax 10-4—Syntax for function call The order of evaluation of the arguments to a function call is undefined. Example: The following example creates a word by concatenating the results of two calls to the function getbyte (defined in 10.3.1): word = control ? {getbyte(msbyte), getbyte(lsbyte)}:0; 10.3.4 Function rules Functions are more limited than tasks. The following six rules govern their usage: a) A function definition shall not contain any time-controlled statements—that is, any statements introduced with #, @, or wait. b) Functions shall not enable tasks. c) A function definition shall contain at least one input argument. d) A function definition shall not have any argument declared as output or inout. e) A function definition shall include an assignment of the function result value to the internal variable that has the same name as the function name. f) A function shall not have any nonblocking assignments. Example: This example defines a function called factorial that returns an integer value. The factorial function is called iteratively and the results are printed. 160 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C module tryfact; // define the function function automatic integer factorial; input [31:0] operand; integer i; if (operand >= 2) factorial = factorial (operand - 1) * operand; else factorial = 1; endfunction // test the function integer result; integer n; initial begin for (n = 0; n <= 7; n = n+1) begin result = factorial(n); $display("%0d factorial=%0d", n, result); end end endmodule // tryfact The simulation results are as follows: 0 factorial=1 1 factorial=1 2 factorial=2 3 factorial=6 4 factorial=24 5 factorial=120 6 factorial=720 7 factorial=5040 10.3.5 Use of constant functions Constant function calls are used to support the building of complex calculations of values at elaboration time (see 12.1.3). A constant function call shall be a function invocation of a constant function local to the calling module where the arguments to the function are constant expressions. Constant functions are a subset of normal Verilog functions that shall meet the following constraints: — They shall contain no hierarchical references. — Any function invoked within a constant function shall be a constant function local to the current module. — All system tasks within a constant function shall be ignored. — All system functions within a constant function shall be illegal. — The only system task that may be invoked is $display, and it shall be ignored when invoked at elaboration time. Copyright © 2001 IEEE. All rights reserved. 161 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® — All parameter values used within the function shall be defined before the use of the invoking constant function call (i.e. any parameter use in the evaluation of a constant function call constitutes a use of that parameter at the site of the original constant function call). — All identifiers which are not parameters or functions shall be declared locally to the current function. — If they use any parameter value that is affected directly or indirectly by a defparam statement (see 12.2.1), the result is undefined. This can produce an error or the constant function can return an indeterminate value. — They shall not be declared inside a generate scope. — They shall not themselves use constant functions in any context requiring a constant expression. Constant function calls are evaluated at elaboration time. Their execution has no effect on the initial values of the variables used either at simulation time or among multiple invocations of a function at elaboration time. In each of these cases, the variables are initialized as they would be for normal simulation. Example: This example defines a function called clogb2 that returns an integer which has the value of the ceiling of the log base 2. module ram_model (address, write, chip_select, data); parameter data_width = 8; parameter ram_depth = 256; localparam addr_width = clogb2(ram_depth); input [addr_width - 1:0] address; input write, chip_select; inout [data_width - 1:0] data; //define the clogb2 function function integer clogb2; input [31:0] value; begin value = value - 1; for (clogb2 = 0; value > 0; clogb2 = clogb2 + 1) value = value >> 1; end endfunction reg [data_width - 1:0] data_store[0:ram_depth - 1]; //the rest of the ram model An instance of this ram_model with parameters assigned: ram_model #(32,421) ram_a0(a_addr,a_wr,a_cs,a_data); 162 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 11. Disabling of named blocks and tasks The disable statement provides the ability to terminate the activity associated with concurrently active procedures, while maintaining the structured nature of Verilog HDL procedural descriptions. The disable statement gives a mechanism for terminating a task before it executes all its statements, breaking from a looping statement, or skipping statements in order to continue with another iteration of a looping statement. It is useful for handling exception conditions such as hardware interrupts and global resets. The disable statement has the syntax form shown in Syntax 11-1. disable_statement ::= (From Annex A - A.6.5) disable hierarchical_task_identifier ; | disable hierarchical_block_identifier ; Syntax 11-1—Syntax of disable statement Either form of disable statement shall terminate the activity of a task or a named block. Execution shall resume at the statement following the block or following the task enabling statement. All activities enabled within the named block or task shall be terminated as well. If task enable statements are nested—that is, one task enables another, and that one enables yet another—then disabling a task within the chain shall disable all tasks downward on the chain. If a task is enabled more than once, then disabling such a task shall disable all activations of the task. The results of the following activities that may be initiated by a task are not specified if the task is disabled: — Results of output and inout arguments — Scheduled, but not executed, nonblocking assignments — Procedural continuous assignments (assign and force statements) The disable statement can be used within blocks and tasks to disable the particular block or task containing the disable statement. The disable statement can be used to disable named blocks within a function, but cannot be used to disable functions. In cases where a disable statement within a function disables a block or a task that called the function, the behavior is undefined. Disabling an automatic task or a block inside an automatic task proceeds as for regular tasks for all concurrent executions of the task. Examples: Example 1—This example illustrates how a block disables itself. begin : block_name rega = regb; disable block_name; regc = rega; // this assignment will never execute end Example 2—This example shows the disable statement being used within a named block in a manner similar to a forward goto. The next statement executed after the disable statement is the one following the named block. Copyright © 2001 IEEE. All rights reserved. 163 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® begin : block_name ... ... if (a == 0) disable block_name; ... end // end of named block // continue with code following named block ... Example 3—This example shows the disable statement being used as an early return from a task. However, a task disabling itself using a disable statement is not a short-hand for the return statement found in programming languages. task proc_a; begin ... ... if (a == 0) disable proc_a; // return if true ... ... end endtask Example 4—This example shows the disable statement being used in an equivalent way to the two statements continue and break in the C programming language. The example illustrates control code that would allow a named block to execute until a loop counter reaches n iterations or until the variable a is set to the value of b. The named block break contains the code that executes until a == b, at which point the disable break; statement terminates execution of that block. The named block continue contains the code that executes for each iteration of the for loop. Each time this code executes the disable continue; statement, the continue block terminates and execution passes to the next iteration of the for loop. For each iteration of the continue block, a set of statements executes if (a != 0). Another set of statements executes if(a!=b). begin : break for (i = 0; i < n; i = i+1) begin : continue @clk if (a == 0) // "continue" loop disable continue; statements statements @clk if (a == b) // "break" from loop disable break; statements statements end end 164 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Example 5—This example shows the disable statement being used to disable concurrently a sequence of timing controls and the task action, when the reset event occurs. The example shows a fork/join block within which is a named sequential block (event_expr) and a disable statement that waits for occurrence of the event reset. The sequential block and the wait for reset execute in parallel. The event_expr block waits for one occurrence of event ev1 and three occurrences of event trig. When these four events have happened, plus a delay of d time units, the task action executes. When the event reset occurs, regardless of events within the sequential block, the fork/join block terminates—including the task action. fork begin : event_expr @ev1; repeat (3) @trig; #d action (areg, breg); end @reset disable event_expr; join Example 6—The next example is a behavioral description of a retriggerable monostable. The named event retrig restarts the monostable time period. If retrig continues to occur within 250 time units, then q will remain at 1. always begin : monostable #250 q = 0; end always @retrig begin disable monostable; q = 1; end Copyright © 2001 IEEE. All rights reserved. 165 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 12. Hierarchical structures The Verilog HDL supports a hierarchical hardware description structure by allowing modules to be embedded within other modules. Higher-level modules create instances of lower-level modules and communicate with them through input, output, and bidirectional ports. These module input/output ports can be scalar or vector. As an example of a module hierarchy, consider a system consisting of printed circuit boards (PCBs). The system would be represented as the top-level module and would create instances of modules that represent the boards. The board modules would, in turn, create instances of modules that represent ICs, and the ICs could, in turn, create instances of modules such as flip-flops, mux’s, and alu’s. To describe a hierarchy of modules, the user provides textual definitions of the various modules. Each module definition stands alone; the definitions are not nested. Statements within the module definitions create instances of other modules, thus describing the hierarchy. 12.1 Modules This clause gives the formal syntax for a module definition and then gives the syntax for module instantiation, along with an example of a module definition and a module instantiation. A module definition shall be enclosed between the keywords module and endmodule. The identifier following the keyword module shall be the name of the module being defined. The optional list of parameter definitions shall specify an ordered list of the parameters for the module. The optional list of ports or port declarations shall specify an ordered list of the ports for the module. The order used in defining the list of parameters in the module_parameter_port_list and in the list of ports can be significant when instantiating the module (see 12.2.2.1 and 12.3.5). The identifiers in this list shall be declared in input, output, and inout statements within the module definition. Ports declared in the list of port declarations shall not be redeclared within the body of the module. The module items define what constitutes a module and they include many different types of declarations and definitions, many of which have already been introduced. The keyword macromodule can be used interchangeably with the keyword module to define a module. An implementation can choose to treat module definitions beginning with macromodule keyword differently. 166 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C module_declaration ::= (From Annex A - A.1.3) { attribute_instance } module_keyword module_identifier [ module_parameter_port_list ] [ list_of_ports ] ; { module_item } endmodule | { attribute_instance } module_keyword module_identifier [ module_parameter_port_list ] [ list_of_port_declarations ] ; { non_port_module_item } endmodule module_keyword ::= module | macromodule module_parameter_port_list ::= (From Annex A -A.1.4 # ( parameter_declaration { , parameter_declaration } ) list_of_ports ::= ( port { , port } ) list_of_port_declarations ::= ( port_declaration { , port_declaration } ) | ( ) port ::= [ port_expression ] | . port_identifier ( [ port_expression ] ) port_expression ::= port_reference | { port_reference { , port_reference } } port_reference ::= port_identifier | port_identifier [ constant_expression ] | port_identifier [ range_expression ] port_declaration ::= {attribute_instance} inout_declaration | {attribute_instance} input_declaration | {attribute_instance} output_declaration module_item ::= module_or_generate_item (From Annex A - A.1.5) | port_declaration ; | { attribute_instance } generated_instantiation | { attribute_instance } local_parameter_declaration | { attribute_instance } parameter_declaration | { attribute_instance } specify_block | { attribute_instance } specparam_declaration module_or_generate_item ::= { attribute_instance } module_or_generate_item_declaration | { attribute_instance } parameter_override | { attribute_instance } continuous_assign | { attribute_instance } gate_instantiation | { attribute_instance } udp_instantiation | { attribute_instance } module_instantiation | { attribute_instance } initial_construct | { attribute_instance } always_construct module_or_generate_item_declaration ::= net_declaration | reg_declaration | integer_declaration | real_declaration | time_declaration | realtime_declaration | event_declaration | genvar_declaration | task_declaration | function_declaration non_port_module_item ::= { attribute_instance } generated_instantiation | { attribute_instance } local_parameter_declaration | { attribute_instance } module_or_generate_item | { attribute_instance } parameter_declaration | { attribute_instance } specify_block | { attribute_instance } specparam_declaration parameter_override ::= defparam list_of_param_assignments ; Syntax 12-1—Syntax for module Copyright © 2001 IEEE. All rights reserved. 167 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® See 12.3 for the definitions of ports. 12.1.1 Top-level modules Top-level modules are modules that are included in the source text but are not instantiated, as described in 12.1.2. 12.1.2 Module instantiation Instantiation allows one module to incorporate a copy of another module into itself. Module definitions do not nest. That is, one module definition shall not contain the text of another module definition within its module-endmodule keyword pair. A module definition nests another module by instantiating it. The module instantiation statement creates one or more named instances of a defined module. For example, a counter module might instantiate a D flip-flop module to create multiple instances of the flipflop. Syntax 12-2 gives the syntax for specifying instantiations of modules. module_instantiation ::= (From Annex A - A.4.1) module_identifier [ parameter_value_assignment ] module_instance { , module_instance } ; parameter_value_assignment ::= # ( list_of_parameter_assignments ) list_of_parameter_assignments ::= ordered_parameter_assignment { , ordered_parameter_assignment } | named_parameter_assignment { , named_parameter_assignment } ordered_parameter_assignment ::= expression named_parameter_assignment ::= . parameter_identifier ( [ expression ] ) module_instance ::= name_of_instance ( [ list_of_port_connections ] ) name_of_instance ::= module_instance_identifier [ range ] list_of_port_connections ::= ordered_port_connection { , ordered_port_connection } | named_port_connection { , named_port_connection } ordered_port_connection ::= { attribute_instance } [ expression ] named_port_connection ::= { attribute_instance } .port_identifier ( [ expression ] ) Syntax 12-2—Syntax for module instantiation The instantiations of modules can contain a range specification. This allows an array of instances to be created. The array of instances are described in 7.1. The syntax and semantics of arrays of instances defined for gates and primitives apply for modules as well. One or more module instances (identical copies of a module) can be specified in a single module instantiation statement. 168 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The list of port connections shall be provided only for modules defined with ports. The parentheses, however, are always required. When a list of port connections is given using the ordered port connection method, the first element in the list shall connect to the first port declared in the module, the second to the second port, and so on. See 12.3 for a more detailed discussion of ports and port connection rules. A connection can be a simple reference to a variable or a net identifier, an expression, or a blank. An expression can be used for supplying a value to a module input port. A blank port connection shall represent the situation where the port is not to be connected. When connecting ports by name, an unconnected port can be indicated either by omitting it in the port list, or by providing no expression in the parentheses [i.e., .port_name ()]. Examples: Example 1—The following example illustrates a circuit (the lower-level module) being driven by a simple waveform description (the higher-level module) where the circuit module is instantiated inside the waveform module. // Lower level module: // module description of a nand flip-flop circuit module ffnand (q, qbar, preset, clear); output q, qbar; //declares 2 circuit output nets input preset, clear; //declares 2 circuit input nets // declaration of two nand gates and their interconnections nand g1 (q, qbar, preset), g2 (qbar, q, clear); endmodule // Higher-level module: // a waveform description for the nand flip-flop module ffnand_wave; wire out1, out2; //outputs from the circuit reg in1, in2; //variables to drive the circuit parameter d = 10; // instantiate the circuit ffnand, name it “ff”, // and specify the IO port interconnections ffnand ff(out1, out2, in1, in2); // define the waveform to stimulate the circuit initial begin #d in1 = 0; in2 = 1; #d in1 = 1; #d in2 = 0; #d in2 = 1; end endmodule Example 2—The following example creates two instances of the flip-flop module ffnand defined in example 1. It connects only to the q output in one instance and only to the qbar output in the other instance. Copyright © 2001 IEEE. All rights reserved. 169 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® // a waveform description for testing // the nand flip-flop, without the output ports module ffnand_wave; reg in1, in2; //variables to drive the circuit parameter d = 10; // make two copies of the circuit ffnand // ff1 has qbar unconnected, ff2 has q unconnected ffnand ff1(out1, , in1, in2), ff2(.qbar(out2), .clear(in2), .preset(in1), .q()); // ff3(.q(out3),.clear(in1),,,); is illegal // define the waveform to stimulate the circuit initial begin #d in1 = 0; in2 = 1; #d in1 = 1; #d in2 = 0; #d in2 = 1; end endmodule 12.1.3 Generated instantiation After a Verilog design has been parsed, but before simulation begins, the design must have the modules being instantiated linked to the modules being defined, the parameters propagated among the various modules, and hierarchical references resolved. This phase in understanding a Verilog description is termed elaboration. Generate instantiations are resolved during elaboration because that is when the parameters associated with a module become defined, hence, allowing the definition of the generated statements and declarations. Genvars are only defined during the evaluation of the generate instantiations and do not exist during simulation of a design. Generate statements facilitate the creation of parameterized models. When used with constant functions (see 10.3.5), parameters can be used to constrain other parameter(s) or localparam(s) in a generated design. All generate instantiations are coded within a module scope and require the keywords generate - endgenerate. Generate statements allow control over the declaration of variables, functions and tasks, as well as control over instantiations. Generated instantiations are one or more: modules, user defined primitives, Verilog gate primitives, continuous assignments, initial blocks and always blocks. Generated declarations and instantiations can be conditionally instantiated into a design. Generated variable declarations and instantiations can be multiply instantiated into a design. Generated instances have unique identifier names and can be referenced hierarchically as described in 12.4. To support the interconnection between structural elements and/or procedural blocks, generate statements permit the following Verilog data types to be declared within the generate scope: net, reg, integer, real, time, realtime, and event. Generated data types have unique identifier names and can be referenced hierarchically as described in 12.4 . Parameters may be redefined using defparam statements (see 12.2.1) or module instance parameter value assignments (see 12.2.2) within the generate scope. However, a defparam statement within the generate scope or within a hierarchy instantiated within the generate scope shall only modify the value of a parameter declared within the generate scope or within a hierarchy instantiated within the generate scope. 170 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Task and function declarations shall also be permitted within the generate scope, however not in a generate loop. Generated tasks and functions shall have unique identifier names and may be referenced hierarchically as described in 12.4. Module declarations and module items that shall not be permitted in a generate statement include: parameters, local parameters, input declarations, output declarations, inout declarations and specify blocks. Connections to generated module instances are handled the same way as they are handled with normal module instances as described in 12.1.2. Generated statements are created using one of the following three methods: generate-loop, generate-conditional, or generate-case. The syntax for generate instantiations is given in Syntax 12-3. Copyright © 2001 IEEE. All rights reserved. 171 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® module_item ::= (From Annex A - A.1.5) module_or_generate_item | port_declaration ; | { attribute_instance } generated_instantiation | { attribute_instance } local_parameter_declaration | { attribute_instance } parameter_declaration | { attribute_instance } specify_block | { attribute_instance } specparam_declaration module_or_generate_item ::= { attribute_instance } module_or_generate_item_declaration | { attribute_instance } parameter_override | { attribute_instance } continuous_assign | { attribute_instance } gate_instantiation | { attribute_instance } udp_instantiation | { attribute_instance } module_instantiation | { attribute_instance } initial_construct | { attribute_instance } always_construct module_or_generate_item_declaration ::= net_declaration | reg_declaration | integer_declaration | real_declaration | time_declaration | realtime_declaration | event_declaration | genvar_declaration | task_declaration | function_declaration generated_instantiation ::= (From Annex A -A.4.2) generate { generate_item } endgenerate generate_item_or_null ::= generate_item | ; generate_item ::= generate_conditional_statement | generate_case_statement | generate_loop_statement | generate_block | module_or_generate_item generate_conditional_statement ::= if ( constant_expression ) generate_item_or_null [ else generate_item_or_null ] generate_case_statement ::= case ( constant_expression ) genvar_case_item { genvar_case_item } endcase genvar_case_item ::= constant_expression { , constant_expression } : generate_item_or_null | default [ : ] generate_item_or_null generate_loop_statement ::= for ( genvar_assignment ; constant_expression ; genvar_assignment ) begin : generate_block_identifier { generate_item } end genvar_assignment ::= genvar_identifier = constant_expression generate_block ::= begin [ : generate_block_identifier ] { generate_item } end Syntax 12-3—Syntax for generate blocks 172 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 12.1.3.1 genvar - generate statement index variable An index variable that shall only be declared for use in generate statements shall be declared as a genvar and is referred to as a genvar in the rest of this section. The syntax for generate statement index variable declarations is given in Syntax 12-4. genvar_declaration ::= (From Annex A - A.2.1.3) genvar list_of_genvar_identifiers ; list_of_genvar_identifiers ::= (From Annex A - A.2.3) genvar_identifier { , genvar_identifier } Syntax 12-4—Syntax for generate statement index variable declaration A genvar shall be declared within the module where the genvar is used. A genvar can be declared either inside or outside of a generate scope. A genvar is an integer that is local to and shall only be used within a generate loop that uses it as an index variable. If any bit of the genvar ever is set to an X or Z or if the genvar is set to a negative value, this shall be an error. Genvars are only defined during the evaluation of the generate blocks (see 12.1.3), and do not exist during simulation of a Verilog design. The value of a genvar shall only be defined by a generate loop. Two generate loops using the same genvar as an index variable shall not be nested. The value of a genvar can be referenced in any context where the value of a parameter could be referenced. 12.1.3.2 generate-loop A generate-loop permits one or more variable declarations, modules, user defined primitives, gate primitives, continuous assignments, initial blocks and always blocks to be instantiated multiple times using a for-loop. The index loop variable used in a generate for-loop shall be declared as a genvar. Both genvar assignments in the for-loop shall assign to the same genvar, which is the loop index variable. The first genvar assignment in the for-loop shall not reference the loop index variable on the right-hand side. Examples: Example 1—A parameterized gray-code to binary-code converter module using a loop to generate continuous assignments module gray2bin1 (bin, gray); parameter SIZE = 8; // this module is parameterizable output [SIZE-1:0] bin; input [SIZE-1:0] gray; genvar i; generate for (i=0; i0) for (m=0; m ; Syntax 13-2—Syntax for declaring library in the library map file NOTES 1—The file_path uses file system-specific notation to specify an absolute or relative path to a particular file or set of files. The following shortcuts/wildcards can be used: ? * ... .. . single character wildcard (matches any single character) multiple character wildcard (matches any number of characters in a directory/file name) hierarchical wildcard (matches any number of hierarchical directories) specifies the parent directory specifies the directory containing the lib.map Paths which end in / shall include all files in the specified directory. Identical to /*. Paths which do not begin with / are relative to the directory in which the current lib.map file is located. 2—The paths ./*.v and *.v are identical and both specify all files with a .v suffix in the current directory. Any file encountered by the compiler which does not match any library’s file_path specification shall by default be compiled into a library named work. To perform the library mapping discussed in the example in 13.1, use the following library definitions in the lib.map file: library rtlLib *.v; // matches all files in the current directory with a .v suffix library gateLib ./*.vg; // matches all files in the current directory with a .vg suffix 13.2.1.1 File path resolution If a file name potentially matches multiple file path specifications, the path specifications shall be resolved in the following order: a) File path specifications which end with an explicit filename b) File path specifications which end with a wildcarded filename c) File path specifications which end with a directory If a file name matches path specifications in multiple library definitions (after the above resolution rules have been applied), it shall be an error. 202 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Using these rules with the library definitions in the lib.map file, all source files encountered by the parser/ compiler can be mapped to a unique library. Once the source descriptions have been mapped to libraries, the cells defined therein are available for binding. NOTE—Tool implementers may find it convenient to provide a command-line argument to explicitly specify the library into which the file being parsed is to be mapped, which shall override any library definitions in the lib.map file. If these libraries do not exist in the lib.map file, they can only be accessed via an explicit config. If multiple cells with the same name map to the same library, then the LAST cell encountered shall be written to the library. This is to support a “separate-compile” use-model (see 13.4.3), where it is assumed encountering a cell after it has previously been compiled is intended to be a recompiling of the cell. In the case where multiple modules with the same name are mapped to the same library in a single invocation of the compiler, then a warning message shall be issued. 13.2.2 Using multiple library mapping files In addition to specifying library mapping information, a lib.map file can also include references to other lib.map files. The include command is used to insert the entire contents of a library mapping file in another file during parsing. The result is as though the contents of the included mapping file appear in place of the include command. The syntax of a lib.map file is limited to library specifications, include statements, and standard Verilog comment syntax. Syntax 13-3 shows the syntax for the include command. include_statement ::= (From Annex A - A.1.1) include ; Syntax 13-3—Syntax for include command If the file path specification, whether in an include or library statement, describes a relative path, it shall be relative to the location of the file which contains the file path. Library providers shall include a local library mapping file in addition to the source contents of the library. Individual users can then simply include the provider’s library mapping file in their own map file to gain access to the contents of the provided library. 13.2.3 Mapping source files to libraries For each cell definition encountered during parsing/compiling, the name of the source file being parsed is compared to the file path specifications of the library declarations in all of the library map files being used. The cell is mapped into the library whose file path specification matches the source file name. 13.3 Configurations As mentioned in the introduction of this chapter, a configuration is simply a set of rules to apply when searching for library cells to which to bind instances. The syntax for configurations is shown in 13.3.1. 13.3.1 Basic configuration syntax The configuration syntax is shown in Syntax 13-4. Copyright © 2001 IEEE. All rights reserved. 203 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® config_declaration ::= (From Annex A -A.1.2) config config_identifier ; design_statement {config_rule_statement} endconfig design_statement ::= design { [library_identifier.]cell_identifier } ; config_rule_statement ::= default_clause liblist_clause | inst_clause liblist_clause | inst_clause use_clause | cell_clause liblist_clause | cell_clause use_clause Syntax 13-4—Syntax for configuration 13.3.1.1 Design statement The design statement names the library and cell of the top-level module or modules in the design hierarchy configured by the config. There shall be one and only one design statement, but multiple top-level modules can be listed in the design statement. The cell or cells identified can not be configurations themselves. It is possible the design identified can have the same name as configs, however. The design statement shall appear before any config rule statements in the config. If the library identifier is omitted, then the library which contains the config shall be used to search for the cell. 13.3.1.2 The default clause The syntax for the default clause is specified in Syntax 13-5. default_clause ::= (From Annex A - A.1.2) default Syntax 13-5—Syntax for default clause The default clause selects all instances which do not match a more specific selection clause. The use expansion clause (see 13.3.1.6) can not be used with a default selection clause. For other expansion clauses, there can not be more than one default clause which specifies the expansion clause. For simple design configurations, it might be sufficient to specify a default liblist (see 13.3.1.5). 13.3.1.3 The instance clause The instance clause is used to specify the specific instance to which the expansion clause shall apply.The syntax for the instance clause is specified in Syntax 13-6. 204 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C inst_clause ::= (From Annex A - A.1.2) instance inst_name inst_name ::= topmodule_identifier{.instance_identifier} Syntax 13-6—Syntax for instance clause The instance name associated with the instance clause is a Verilog hierarchical name, starting at the toplevel module of the config (i.e., the name of the cell in the design statement). 13.3.1.4 The cell clause The cell selection clause names the cell to which it applies. The syntax for the cell clause is specified in Syntax 13-7. cell_clause ::= (From Annex A - A.1.2) cell [ library_identifier.]cell_identifier Syntax 13-7—Syntax for cell clause If the optional library name is specified then the selection rule applies to any instance which is bound or is under consideration for being bound to the selected library and cell. It is an error if a library name is included in a cell selection clause and the corresponding expansion clause is a library list expansion clause. 13.3.1.5 The liblist clause The liblist clause defines an ordered set of libraries to be searched to find the current instance. The syntax for the liblist clause is specified in Syntax 13-8. liblist_clause ::= (From Annex A - A.1.2) liblist [{library_identifier}] Syntax 13-8—Syntax for liblist clause liblists are inherited hierarchically downward as instances are bound. When searching for a cell to bind to the current unbound instance, and in the absence of an applicable binding expansion clause, the specified library list is searched in the specified order. The current library list is selected by the selection clauses. If no library list clause is selected, or the selected library list is empty, then the library list contains the single name which is the library in which the cell containing the unbound instance is found (i.e., the parent cell’s library). Copyright © 2001 IEEE. All rights reserved. 205 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 13.3.1.6 The use clause The use clause specifies a specific binding for the selected cell. The syntax for the use clause is specified in Syntax 13-9. use_clause ::= (From Annex A - A.1.2) use [library_identifier.]cell_identifier[:config] Syntax 13-9—Syntax for use clause A use clause can only be used in conjunction with an instance or cell selection clause. It specifies the exact library and cell to which a selected cell or instance is bound. The use clause has no effect on the current value of the library list. It can be common in practice to specify multiple config rule statements, one of which specifies a binding and the other of which specifies a library list. If the lib.cell being referred to by the use clause is a config which has the same name as a module/macromodule/primitive in the same library, then the optional :config suffix can be added to the lib.cell to specify the config explicitly. If the library name is omitted, the library shall be inherited from the parent cell. NOTE—The binding statement can create situations where the unbound instance's module name and the cell name to which it is bound are different. 13.3.2 Hierarchical configurations For situations where it is desirable to specify a special set of configuration rules for a subsection of a design, it is possible to bind a particular instance directly to a configuration using the binding clause: instance top.a1.foo use lib1.foo:config; // bind to the config foo in library lib1 specifies the instance top.a1.foo is to be replaced with the design hierarchy specified by the configuration lib1.foo:config. The design statement in lib1.foo:config shall specify the actual binding for the instance top.a1.foo, and the rules specified in the config shall determine the configuration of all other subinstances under top.a1.foo. It shall be an error for an instance clause to specify a hierarchical path to an instance which occurs within a hierarchy specified by another config. config bot; design lib1.bot; default liblist lib1 lib2; instance bot.a1 liblist lib3; endconfig config top; design lib1.top; default liblist lib2 lib1; instance top.bot use lib1.bot:config; instance top.bot.a1 liblist lib4; // ERROR - can’t set liblist for top.bot.a1 from this config endconfig 206 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 13.4 Using libraries and configs The following subclause describes potential use-models for referencing configs on the command line. It is included for clarification purposes. The traditional Verilog simulation use-model takes a file-based approach, where the source descriptions for all cells in the design are specified on the command line for each invocation of the tool. With the advent of compiled-code simulators, the configuration mechanism shall also support a use-model which allows for the source files to be pre-compiled and then for the pre-compiled design objects to be referenced on the command line. This subclause shall explain how configurations can be used in both of these scenarios. 13.4.1 Precompiling in a single-pass use-model The single-pass use-model is the traditional use-model with which most users are familiar. In this use-model, all of the source description files shall be provided to the simulator via the command line, and only these source descriptions can be used to bind cell instances in the current design. A precompiling strategy in this scenario actually parses every cell description provided on the command line and maps it into the library without regard to whether the cell actually is used in the design. The tool can optionally check to see if the cell already exists in the library, and if it is up-to-date (i.e. the source description has not changed since the last time the cell was compiled) the tool can skip recompiling the cell. After all cells on the command line have been compiled, then the tool can locate the top-level cell (discussed in Clause 12), and proceed down the hierarchy, binding each instance as it is encountered in the hierarchy. NOTE—With this use-model it is not necessary for library objects to persist from one tool invocation to another (although for performance considerations it is recommended they do). 13.4.2 Elaboration-time compiling in a single-pass use-model An alternate strategy which can be used with a single-pass tool is to parse the source files only to find the top-level module(s), without actually compiling anything into the library during this scanning process. Once the top-level module(s) has been found, then it can be compiled into the library, and the tool can proceed down the hierarchy, only compiling the source descriptions necessary to bind the design successfully. Based on the binding rules in place, only the source files which match the current library specification need to be parsed to find the current cell’s source description to compile. As with the precompiled single-pass usemodel, it is not necessary for library cells to persist from one invocation to another using this strategy. 13.4.3 Precompiling using a separate compilation tool When using a separate compilation tool, it is essential library cells persist, and the compiled forms shall therefore exist somewhere in the file system. The exact format and location for holding these compiled forms shall be vendor/tool-specific. Using this separate compiler strategy, the source descriptions shall be parsed and compiled into the library using one or more invocations of the compiler tool. The only restriction is all cells in a design shall be precompiled prior to binding the design (typically via an invocation of a separate tool). Using this strategy, the tool which actually does the binding only needs to be told the top-level module(s) of the design to be bound, and then it shall use the precompiled form of the cell description(s) from the library to determine the subinstances and descend hierarchically down the design binding each cell as it is located. 13.4.4 Command line considerations In each of the three preceding strategies, the binding rules can either be specified via a config, or the default rules (from the library map file) can be used. In the single-pass use-models, the config can be specified by including its source description file on the command line. In the case where the config includes a design statement, then the specified cell shall be the top-level module, regardless of the presence of any uninstantiated cells in the rest of the source files. When using a separate compilation tool, the tool which actually does the binding only needs to be given the lib.cell specification for the top-level cell(s) and/or the config to be used. In this strategy, the config itself shall also be precompiled. Copyright © 2001 IEEE. All rights reserved. 207 IEEE Std 1364-2001 Version C 13.5 Configuration examples Consider the following set of source descriptions: Example: IEEE STANDARD VERILOG® file top.v module top(...); ... adder a1(...); adder a2(...); endmodule module foo(...); ... // rtl endmodule file adder.v module adder(...); ... // rtl foo f1(...); foo f2(...); endmodule module foo(...); ... // rtl endmodule file adder.vg module adder(...); ... // gate-level foo f1(...); foo f2(...); endmodule module foo(...); ... // gate-level endmodule file lib.map library rtlLib top.v; library aLib adder.*; library gateLib adder.vg; All of the examples in this section shall assume the top.v, adder.v and adder.vg files get compiled with the given lib.map file. This yields the following library structure: rtlLib.top // from top.v rtlLib.foo // from top.v aLib.adder // from adder.v aLib.foo // rtl from adder.v gateLib.adder // from adder.vg gateLib.foo // from adder.vg 13.5.1 Default configuration from library map file With no configuration, the libraries are searched according to the library declaration order in the library map file. This means all instances of module adder shall use aLib.adder (since aLib is the first library specified which contains a cell named adder), and all instances of module foo shall use rtlLib.foo (since rtlLib is the first library which contains foo). 13.5.2 Using the default clause To always use the foo definition from file adder.v, use the following simple configuration: config cfg1; design rtlLib.top default liblist aLib rtlLib; endconfig The default liblist statement overrides the library search order in the lib.map file, so aLib is always searched before rtlLib. Since the gateLib library is not included in the liblist, the gate-level descriptions of adder and foo shall not be used. To use the gate-level representations of adder and foo, add to the config as follows: config cfg2; design rtlLib.top default liblist gateLib aLib rtlLib; endconfig 208 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C This shall cause the gate representation always to be taken before the rtl representation, using the module definitions for adder and foo from adder.vg. The rtl view of top shall be taken since there is no gate representation available. 13.5.3 Using the cell clause To modify the config to use the rtl view of adder and the gate-level representation of foo from gateLib: config cfg3; design rtlLib.top default liblist aLib rtlLib; cell foo use gateLib.foo; endconfig The cell clause selects all cells named foo and explicitly binds them to the gate representation in gateLib. 13.5.4 Using the instance clause To modify the config so the top.a1 adder (and its descendants) use the gate representation and the top.a2 adder (and its descendants) use the rtl representation from aLib: config cfg4 design rtlLib.top default liblist gateLib rtlLib; instance top.a2 liblist aLib; endconfig Since the liblist is inherited, all of the descendants of top.a2 inherit its liblist from the instance selection clause. 13.5.5 Using a hierarchical config Now suppose all this work has only been on the adder module by itself and a config which uses the rtlLib.foo cell for f1, and the gateLib.foo cell for f2 has already been developed. Then use: config cfg5; design aLib.adder; default liblist gateLib aLib; instance adder.f1 liblist rtlLib; endconfig To use this configuration cfg5 for the top.a2 instance of adder and take the full default aLib adder for the top.a1 instance, use the following config: config cfg6; design rtlLib.top; default liblist aLib rtlLib; instance top.a2 use work.cfg5:config endconfig The binding clause specifies the work.cfg5:config configuration is to be used to resolve the bindings of instance top.a2 and its descendants. It is the design statement in config cfg5 which defines the exact binding for the top.a2 instance itself. The rest of cfg5 defines the rules to bind the descendants of top.a2. Notice the instance clause in cfg5 is relative to its own top-level module, adder. Copyright © 2001 IEEE. All rights reserved. 209 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 13.6 Displaying library binding information It shall be possible to display the actual library binding information for module instances during simulation. The format specifier %l or %L shall print out the library.cell binding information for the module instance containing the display (or other textual output) command. This is similar to the %m format specifier which prints out the hierarchical path name of the module containing it. It shall also be able to use the VPI interface to display the binding information. The following new vpiProperties shall exist for objects of type vpiModule: — vpiUseBinding - the library.cell binding information for a module instance — vpiLibrary - the library name into which the module was compiled — vpiCell - the name of the cell bound to the module instance — vpiConfig - the library.cell name of the config controlling the binding of the module instance These properties shall be of string type, similar to the vpiName and vpiFullName properties. 13.7 Library mapping examples In the absence of a configuration, it is possible to perform basic control of the library searching order when binding a design. When a config is used, the config overrides the rules specified here. 13.7.1 Using the command line to control library searching In the absence of a configuration, it shall be necessary for all compliant tools to provide a mechanism of specifying a library search order on the command line which overrides the default order from the library mapping file. This mechanism shall include specification of library names only, with the definitions of these libraries to be taken from the library mapping file. NOTE—It is recommended all compliant tools use "-L " to specify this search order. 13.7.2 File path specification examples Example: Given the following set of files: /proj/lib1/rtl/a.v /proj/lib2/gates/a.v /proj/lib1/rtl/b.v /proj/lib2/gates/b.v From the /proj library, the following absolute file_path_specs are resolved as shown: /proj/lib*/*/a.v =/proj/lib1/rtl/a.v, /proj/lib2/gates/a.v .../a.v =/proj/lib1/rtl/a.v, /proj/lib2/gates/a.v /proj/.../b.v =/proj/lib1/rtl/b.v, /proj/lib2/gates/b.v .../rtl/*.v =/proj/lib1/rtl/a.v, /proj/lib1/rtl/b.v 210 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C From the /proj/lib1 directory, the following relative file_path_specs are resolved as shown: ../lib2/gates/*.v = /proj/lib2/gates/a.v, /proj/lib2/gates/b.v ./rtl/?.v = /proj/lib1/rtl/a.v, /proj/lib1/rtl/b.v ./rtl/ = /proj/lib1/rtl/a.v, /proj/lib1/rtl/b.v 13.7.3 Resolving multiple path specifications Example: library lib1 “/proj/lib1/foo*.v”; library lib2 “/proj/lib1/foo.v”; library lib3 “../lib1/”; library lib4 “/proj/lib1/*ver.v”; When evaluated from the directory /proj/tb directory, the following source files shall map into the specified library: ../lib1/foobar.v - lib1 // potentially matches lib1 and lib3. Since lib1 includes a filename and lib3 only specifies a directory; lib1 takes precedence /proj/lib1/foo.v - lib2 // takes precedence over lib1 and lib3 path specifications /proj/lib1/bar.v - lib3 /proj/lib1/barver.v - lib4 // takes precedence over lib3 path specification /proj/lib1/foover.v - ERROR // matches lib1 and lib4 /test/tb/tb.v - work // does not match any library specifications. Copyright © 2001 IEEE. All rights reserved. 211 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 14. Specify blocks Two types of HDL constructs are often used to describe delays for structural models such as ASIC cells. They are as follows: — Distributed delays, which specify the time it takes events to propagate through gates and nets inside the module (see 7.14) — Module path delays, which describe the time it takes an event at a source (input port or inout port) to propagate to a destination (output port or inout port) This clause describes how paths are specified in a module and how delays are assigned to these paths. 14.1 Specify block declaration A block statement called the specify block is the vehicle for describing paths between a source and a destination and for assigning delays to these paths. The syntax for specify block is shown in Syntax 14-1. specify_item ::= (From Annex A - A.7.1) specparam_declaration | pulsestyle_declaration | showcancelled_declaration | path_declaration | system_timing_check Syntax 14-1—Syntax of specify block The specify block shall be bounded by the keywords specify and endspecify, and it shall appear inside a module declaration. The specify block can be used to perform the following tasks: — Describe various paths across the module. — Assign delays to those paths. — Perform timing checks to ensure that events occurring at the module inputs satisfy the timing con- straints of the device described by the module (see Clause15). The paths described in the specify block, called module paths, pair a signal source with a signal destination. The source may be unidirectional (an input port) or bidirectional (an inout port) and is referred to as the module path source. Similarly, the destination may be unidirectional (an output port) or bidirectional (an inout port) and is referred to as the module path destination. Example: specify specparam tRise_clk_q = 150, tFall_clk_q = 200; specparam tSetup = 70; (clk => q) = (tRise_clk_q, tFall_clk_q); $setup(d, posedge clk, tSetup); endspecify 212 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The first two lines following the keyword specify declare specify parameters, which are discussed in 3.11.3. The line following the declarations of specify parameters describes a module path and assigns delays to that module path. The specify parameters determine the delay assigned to the module path. Specifying module paths is presented in 14.2. Assigning delays to module paths is discussed in 14.3. The line preceding the keyword endspecify instantiates one of the system timing checks, which are discussed further in Clause 15. 14.2 Module path declarations There are two steps required to set up module path delays in a specify block: a) Describe the module paths b) Assign delays to those paths (see 14.3) The syntax of the module path declaration is described in Syntax 14-2. path_declaration ::= (From Annex A - A.7.2) simple_path_declaration ; | edge_sensitive_path_declaration ; | state_dependent_path_declaration ; Syntax 14-2—Syntax of the module path declaration A module path may be described as a simple path, an edge sensitive path, or a state dependent path. A module path shall be defined inside a specify block as a connection between a source signal and a destination signal. Module paths can connect any combination of vectors and scalars. Example: Figure 35 illustrates a circuit with module path delays. More than one source (A, B, C, and D) may have a module path to the same destination (Q), and different delays may be specified for each input to output path. 10 12 18 22 A B C D n = module path delay MODULE PATHS: from A to Q Q from B to Q from C to Q from D to Q Figure 35—Module path delays Copyright © 2001 IEEE. All rights reserved. 213 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 14.2.1 Module path restrictions Module paths have the following restrictions: — The module path source shall be a net that is connected to a module input port or inout port. — The module path destination shall be a net or variable that is connected to a module output port or inout port. — The module path destination shall have only one driver inside the module. 14.2.2 Simple module paths The syntax for specifying a simple module path is given in Syntax 14-3. simple_path_declaration ::= (From Annex A - A.7.2) parallel_path_description = path_delay_value | full_path_description = path_delay_value parallel_path_description ::= ( specify_input_terminal_descriptor [ polarity_operator ] => specify_output_terminal_descriptor ) full_path_description ::= ( list_of_path_inputs [ polarity_operator ] *> list_of_path_outputs ) list_of_path_inputs ::= specify_input_terminal_descriptor { , specify_input_terminal_descriptor } list_of_path_outputs ::= specify_output_terminal_descriptor { , specify_output_terminal_descriptor } specify_input_terminal_descriptor ::= (From Annex A - A.7.3) input_identifier | input_identifier [ constant_expression ] | input_identifier [ range_expression ] specify_output_terminal_descriptor ::= output_identifier | output_identifier [ constant_expression ] | output_identifier [ range_expression ] input_identifier ::= input_port_identifier | inout_port_identifier output_identifier ::= output_port_identifier | inout_port_identifier polarity_operator ::= (From Annex A - A.7.4) +|- Syntax 14-3—Syntax for simple module path Simple path can be declared in one of the two forms: — source *> destination — source => destination The symbols *> and => each represent a different kind of connection between the module path source and the module path destination. The operator *> establishes a full connection between source and destination. The operator => establishes a parallel connection between source and destination. Refer to 14.2.5 for a description of full connection and parallel connection paths. 214 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE Example: The following three examples illustrate valid simple module path declarations. IEEE Std 1364-2001 Version C (A => Q) = 10; (B => Q) = (12); (C, D *> Q) = 18; 14.2.3 Edge-sensitive paths When a module path is described using an edge transition at the source, it is called an edge-sensitive path. The edge-sensitive path construct is used to model the timing of input to output delays, which only occur when a specified edge occurs at the source signal. The syntax of the edge-sensitive path declaration is shown in Syntax 14-4. edge_sensitive_path_declaration ::= (From Annex A - A.7.4) parallel_edge_sensitive_path_description = path_delay_value | full_edge_sensitive_path_description = path_delay_value parallel_edge_sensitive_path_description ::= ( [ edge_identifier ] specify_input_terminal_descriptor => specify_output_terminal_descriptor [ polarity_operator ] : data_source_expression ) full_edge_sensitive_path_description ::= ( [ edge_identifier ] list_of_path_inputs *> list_of_path_outputs [ polarity_operator ] : data_source_expression ) data_source_expression ::= expression edge_identifier ::= posedge | negedge Syntax 14-4—Syntax of the edge-sensitive path declaration The edge identifier may be one of the keywords posedge or negedge, associated with an input terminal descriptor, which may be any input port or inout port. If a vector port is specified as the input terminal descriptor, the edge transition shall be detected on the least significant bit. If the edge transition is not specified, the path shall be considered active on any transition at the input terminal. An edge-sensitive path may be specified with full connections (*>) or parallel connections (=>). For parallel connections (=>), the destination shall be any scalar output or inout port or the bit-select of a vector output or inout port. For full connections (*>), the destination shall be a list of one or more of the vector or scalar output and inout ports, and bit-selects or part-selects of vector output and inout ports. Refer to 14.2.5 for a description of parallel paths and full connection paths. The data source expression is an arbitrary expression, which serves as a description of the flow of data to the path destination. This arbitrary data path description does not affect the actual propagation of data or events through the model; how an event at the data path source propagates to the destination depends on the internal logic of the module. The polarity operator describes whether the data path is inverting or noninverting. Copyright © 2001 IEEE. All rights reserved. 215 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Examples: Example 1—The following example demonstrates an edge-sensitive path declaration with a positive polarity operator: ( posedge clock => ( out +: in ) ) = (10, 8); In this example, at the positive edge of clock, a module path extends from clock to out using a rise delay of 10 and a fall delay of 8. The data path is from in to out, and in is not inverted as it propagates to out. Example 2—The following example demonstrates an edge-sensitive path declaration with a negative polarity operator: ( negedge clock[0] => ( out -: in ) ) = (10, 8); In this example, at the negative edge of clock[0], a module path extends from clock[0] to out using a rise delay of 10 and a fall delay of 8. The data path is from in to out, and in is inverted as it propagates to out. Example 3—The following example demonstrates an edge-sensitive path declaration with no edge identifier: ( clock => ( out : in ) ) = (10, 8); In this example, at any change in clock, a module path extends from clock to out. 14.2.4 State-dependent paths A state-dependent path makes it possible to assign a delay to a module path that affects signal propagation delay through the path only if specified conditions are true. A state-dependent path description includes the following items: — A conditional expression that, when evaluated true, enables the module path — A module path description — A delay expression that applies to the module path The syntax for the state-dependent path declaration is shown in Syntax 14-5. state_dependent_path_declaration ::= (From Annex A - A.7.4) if ( module_path_expression ) simple_path_declaration | if ( module_path_expression ) edge_sensitive_path_declaration | ifnone simple_path_declaration Syntax 14-5—Syntax of state-dependent paths 14.2.4.1 Conditional expression The operands in the conditional expression shall be constructed from the following: — Scalar or vector module input ports or inout ports or their bit-selects or part-selects — Locally defined variables or nets or their bit-selects or part-selects — Compile time constants (constant numbers and specify parameters) 216 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Table 46 contains a list of valid operators that may be used in conditional expressions: Table 46—List of valid operators in state dependent path delay expression Operator Description Operator Description ~ & | ^ ^~ ~^ == != && || ! bit-wise negation bit-wise and bit-wise or bit-wise xor bit-wise xnor logical equality logical inequality logical and logical or logical not & | ^ ~& ~| ^~ ~^ {} { {} } ?: reduction and reduction or reduction xor reduction nand reduction nor reduction xnor concatenation replication conditional A conditional expression shall evaluate to true (1) for the state-dependent path to be assigned a delay value. If the conditional expression evaluates to x or z, it shall be treated as true. If the conditional expression evaluates to multiple bits, the least significant bit shall represent the result. The conditional expression can have any number of operands and operators. 14.2.4.2 Simple state-dependent paths If the path description of a state-dependent path is a simple path, then it is called a simple state-dependent path. The simple path description is discussed in 14.2.2. Examples: Example 1—The following example uses state-dependent paths to describe the timing of an XOR gate. module XORgate (a, b, out); input a, b; output out; xor x1 (out, a, b); specify specparam noninvrise = 1, noninvfall = 2; specparam invertrise = 3, invertfall = 4; if (a) (b => out) = (invertrise, invertfall); if (b) (a => out) = (invertrise, invertfall); if (~a)(b => out) = (noninvrise, noninvfall); if (~b)(a => out) = (noninvrise, noninvfall); endspecify endmodule In this example, first two state-dependent paths describe a pair of output rise and fall delay times when the XOR gate (x1) inverts a changing input. The last two state-dependent paths describe another pair of output rise and fall delay times when the XOR gate buffers a changing input. Copyright © 2001 IEEE. All rights reserved. 217 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Example 2—The following example models a partial ALU. The state-dependent paths specify different delays for different ALU operations. module ALU (o1, i1, i2, opcode); input [7:0] i1, i2; input [2:1] opcode; output [7:0] o1; //functional description omitted specify // add operation if (opcode == 2’b00) (i1,i2 *> o1) = (25.0, 25.0); // pass-through i1 operation if (opcode == 2’b01) (i1 => o1) = (5.6, 8.0); // pass-through i2 operation if (opcode == 2’b10) (i2 => o1) = (5.6, 8.0); // delays on opcode changes (opcode => o1) = (6.1, 6.5); endspecify endmodule In the preceding example, the first three path declarations declare paths extending from operand inputs i1 and i2 to the o1 output. The delays on these paths are assigned to operations on the basis of the operation specified by the inputs on opcode. The last path declaration declares a path from the opcode input to the o1 output. 14.2.4.3 Edge-sensitive state-dependent paths If the path description of a state-dependent path describes an edge-dependent path, then the state-dependent path is called an edge-sensitive state-dependent path. The edge-sensitive paths are discussed in 14.2.3. Different delays can be assigned to the same edge-sensitive path as long as the following criteria are met: — The edge, condition, or both make each declaration unique. — The port is referenced in the same way in all path declarations (entire port, bit-select, or part-select). Examples: Example 1 if ( !reset && !clear ) ( posedge clock => ( out +: in ) ) = (10, 8) ; In this example, if the positive edge of clock occurs when reset and clear are low, and a module path extends from clock to out using a rise delay of 10 and a fall delay of 8. 218 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Example 2—The following example shows four edge-sensitive path declarations. Note that each path has a unique edge or condition. specify ( posedge clk => ( q[0] : data ) ) = (10, 5); ( negedge clk => ( q[0] : data ) ) = (20, 12); if (reset) ( posedge clk => ( q[0] : data ) ) = (15, 8); if (!reset && cntrl) ( posedge clk => ( q[0] : data ) ) = (6, 2); endspecify Example 3—The two state-dependent path declarations shown below are not legal because even though they have different conditions, the destinations are not specified in the same way: the first destination is a partselect, the second is a bit-select. specify if (reset) (posedge clk => (q[3:0]:data)) = (10,5); if (!reset) (posedge clk => (q[0]:data)) = (15,8); endspecify 14.2.4.4 The ifnone condition The ifnone keyword is used to specify a default state-dependent path delay when all other conditions for the path are false. The ifnone condition shall specify the same module path source and destination as the statedependent module paths. The following rules apply to module paths specified with the ifnone condition: — Only simple module paths may be described with an ifnone condition. — The state-dependent paths that correspond to the ifnone path may be either simple module paths or edge-sensitive paths. — If there are no corresponding state-dependent module paths to the ifnone module path, then the ifnone module path shall be treated the same as an unconditional simple module path. — It is illegal to specify both an ifnone condition for a module path and an unconditional simple mod- ule path for the same module path. Copyright © 2001 IEEE. All rights reserved. 219 IEEE Std 1364-2001 Version C Examples: Example 1—The following are valid state-dependent path combinations. IEEE STANDARD VERILOG® if (C1) (IN => OUT) = (1,1); ifnone (IN => OUT) = (2,2); // add operation if (opcode == 2’b00) (i1,i2 *> o1) = (25.0, 25.0); // pass-through i1 operation if (opcode == 2’b01) (i1 => o1) = (5.6, 8.0); // pass-through i2 operation if (opcode == 2’b10) (i2 => o1) = (5.6, 8.0); // all other operations ifnone (i2 => o1) = (15.0, 15.0); (posedge CLK => (Q +: D)) = (1,1); ifnone (CLK => Q) = (2,2); Example 2—The following module path description combination is illegal because it combines a statedependent path using an ifnone condition and an unconditional path for the same module path. if (a) (b => out) = (2,2); if (b) (a => out) = (2,2); ifnone (a => out) = (1,1); (a => out) = (1,1); 14.2.5 Full connection and parallel connection paths The operator *> shall be used to establish a full connection between source and destination. In a full connection, every bit in the source shall connect to every bit in the destination. The module path source need not have the same number of bits as the module path destination. The full connection can handle most types of module paths, since it does not restrict the size or number of source signals and destination signals. The following situations require the use of full connections: — To describe a module path between a vector and a scalar — To describe a module path between vectors of different sizes — To describe a module path with multiple sources or multiple destinations in a single statement (see 14.2.6) The operator => shall be used to establish a parallel connection between source and destination. In a parallel connection, each bit in the source shall connect to one corresponding bit in the destination. Parallel module paths can be created only between sources and destinations that contain the same number of bits. Parallel connections are more restrictive than full connections. They only connect one source to one destination, where each signal contains the same number of bits. Therefore, a parallel connection may only be used to describe a module path between two vectors of the same size. Since scalars are one bit wide, either *> or => may be used to set up bit-to-bit connections between two scalars. 220 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Examples: Example 1—Figure 36 illustrates how a parallel connection differs from a full connection between two 4-bit vectors. Parallel module path Input bits 0 1 2 3 Output bits 0 1 2 3 Full module path Input bits 0 1 2 3 Output bits 0 1 2 3 N = number of bits = 4 Number of paths = N = 4 Use => to define path bit-to-bit connections Number of paths = N * N = 16 Use *> to define path bit-to-vector connections Figure 36—The difference between parallel and full connection paths Example 2—The following example shows module paths for a 2:1 multiplexor with two 8-bit inputs and one 8-bit output. module mux8 (in1, in2, s, q) ; output [7:0] q; input [7:0] in1, in2; input s; // Functional description omitted ... specify (in1 => q) = (3, 4) ; (in2 => q) = (2, 3) ; (s *> q) = 1; endspecify endmodule The module path from s to q uses a full connection (*>) because it connects a scalar source—the 1-bit select line—to a vector destination—the 8-bit output bus. The module paths from both input lines in1 and in2 to q use a parallel connection (=>) because they set up parallel connections between two 8-bit buses. 14.2.6 Declaring multiple module paths in a single statement Multiple module paths may be described in a single statement by using the symbol *> to connect a commaseparated list of sources to a comma-separated list of destinations. When describing multiple module paths in one statement, the lists of sources and destinations may contain a mix of scalars and vectors of any size. Copyright © 2001 IEEE. All rights reserved. 221 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® The connection in a multiple module path declaration is always a full connection. Example: (a, b, c *> q1, q2) = 10; is equivalent to the following six individual module path assignments: (a *> q1) = 10 ; (b *> q1) = 10 ; (c *> q1) = 10 ; (a *> q2) = 10 ; (b *> q2) = 10 ; (c *> q2) = 10 ; 14.2.7 Module path polarity The polarity of a module path is an arbitrary specification indicating whether or not the direction of a signal transition is inverted as it propagates from the input to the output. This arbitrary polarity description does not affect the actual propagation of data or events through the model; how a rise or a fall at the source propagates to the destination depends on the internal logic of the module. Module paths may specify any of three polarities: — Unknown polarity — Positive polarity — Negative polarity 14.2.7.1 Unknown polarity By default, module paths shall have unknown polarity—that is, a transition at the path source may propagate to the destination in an unpredictable way, as follows: — A rise at the source may cause either a rise transition, a fall transition, or no transition at the destination. — A fall at the source may cause either a rise transition, a fall transition, or no transition at the destination. A module path specified either as a full connection or a parallel connection, but without a polarity operator + or -, shall be treated as a module path with unknown polarity. 14.2.7.2 Positive polarity For module paths with positive polarity, any transition at the source may cause the same transition at the destination, as follows: — A rise at the source may cause either a rise transition or no transition at the destination. — A fall at the source may cause either a fall transition or no transition at the destination. A module path with positive polarity shall be specified by prefixing the + polarity operator to => or *>. 222 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 14.2.7.3 Negative polarity For module paths with negative polarity, any transition at the source may cause the opposite transition at the destination, as follows: — A rise at the source may cause either a fall transition or no transition at the destination. — A fall at the source may cause either a rise transition or no transition at the destination. A module path with negative polarity shall be specified by prefixing the - polarity operator to => or *>. Examples: The following examples show each type of path polarity: // Positive polarity (In1 +=> q) = In_to_q ; (s +*> q) = s_to_q ; // Negative polarity (In1 -=> q) = In_to_q ; (s -*> q) = s_to_q ; // Unknown polarity (In1 => q) = In_to_q ; (s *> q) = s_to_q ; 14.3 Assigning delays to module paths The delays that occur at the module outputs where paths terminate shall be specified by assigning delay values to the module path descriptions. The syntax for specifying delay values is shown in Syntax 14-6. path_delay_value ::= (From Annex A - A.7.4) list_of_path_delay_expressions | ( list_of_path_delay_expressions ) list_of_path_delay_expressions ::= t_path_delay_expression | trise_path_delay_expression , tfall_path_delay_expression | trise_path_delay_expression , tfall_path_delay_expression , tz_path_delay_expression | t01_path_delay_expression , t10_path_delay_expression , t0z_path_delay_expression , tz1_path_delay_expression , t1z_path_delay_expression , tz0_path_delay_expression | t01_path_delay_expression , t10_path_delay_expression , t0z_path_delay_expression , tz1_path_delay_expression , t1z_path_delay_expression , tz0_path_delay_expression , t0x_path_delay_expression , tx1_path_delay_expression , t1x_path_delay_expression , tx0_path_delay_expression , txz_path_delay_expression , tzx_path_delay_expression t_path_delay_expression ::= path_delay_expression Syntax 14-6—Syntax for path delay value Copyright © 2001 IEEE. All rights reserved. 223 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® In module path delay assignments, a module path description (see 14.2) is specified on the left-hand side, and one or more delay values are specified on the right-hand side. The delay values may be optionally enclosed in a pair of parentheses. There may be one, two, three, six, or twelve delay values assigned to a module path, as described in 14.3.1. The delay values shall be constant expressions containing literals or specparams, and there may be a delay expression of the form min:typ:max. Example: specify // Specify Parameters specparam tRise_clk_q = 45:150:270, tFall_clk_q=60:200:350; specparam tRise_Control = 35:40:45, tFall_control=40:50:65; // Module Path Assignments (clk => q) = (tRise_clk_q, tFall_clk_q); (clr, pre *> q) = (tRise_control, tFall_control); endspecify In the example above, the specify parameters declared following the specparam keyword specify values for the module path delays. The module path assignments assign those module path delays to the module paths. 14.3.1 Specifying transition delays on module paths Each path delay expression may be a single value—representing the typical delay—or a colon-separated list of three values—representing a minimum, typical, and maximum delay, in that order. If the path delay expression results in a negative value, it shall be treated as zero. Table 47 describes how different path delay values shall be associated with various transitions. The path delay expression names refer to the names used in Syntax 14-6. Table 47—Associating path delay expressions with transitions Number of path delay expressions specified Transitions 1 0 -> 1 t 1 -> 0 t 0 -> z t z -> 1 t 1 -> z t z -> 0 t 0 -> x * x -> 1 * 1 -> x * x -> 0 * x -> z * z -> x * * See 14.3.2. 2 trise tfall trise trise tfall tfall * * * * * * 3 trise tfall tz trise tz tfall * * * * * * 6 t01 t10 t0z tz1 t1z tz0 * * * * * * 12 t01 t10 t0z tz1 t1z tz0 t0x tx1 t1x tx0 txz tzx 224 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE Example: IEEE Std 1364-2001 Version C // one expression specifies all transitions (C => Q) = 20; (C => Q) = 10:14:20; // two expressions specify rise and fall delays specparam tPLH1 = 12, tPHL1 = 25; specparam tPLH2 = 12:16:22, tPHL2 = 16:22:25; (C => Q) = ( tPLH1, tPHL1 ) ; (C => Q) = ( tPLH2, tPHL2 ) ; // three expressions specify rise, fall, and z transition delays specparam tPLH1 = 12, tPHL1 = 22, tPz1 = 34; specparam tPLH2 = 12:14:30, tPHL2 = 16:22:40, tPz2 = 22:30:34; (C => Q) = (tPLH1, tPHL1, tPz1); (C => Q) = (tPLH2, tPHL2, tPz2); // six expressions specify transitions to/from 0, 1, and z specparam t01 = 12, t10 = 16, t0z = 13, tz1 = 10, t1z = 14, tz0 = 34 ; (C => Q) = ( t01, t10, t0z, tz1, t1z, tz0) ; specparam T01 = 12:14:24, T10 = 16:18:20, T0z = 13:16:30 ; specparam Tz1 = 10:12:16, T1z = 14:23:36, Tz0 = 15:19:34 ; (C => Q) = ( T01, T10, T0z, Tz1, T1z, Tz0) ; // twelve expressions specify all transition delays explicitly specparam t01=10, t10=12, t0z=14, tz1=15, t1z=29, tz0=36, t0x=14, tx1=15, t1x=15, tx0=14, txz=20, tzx=30 ; (C => Q) = (t01, t10, t0z, tz1, t1z, tz0, t0x, tx1, t1x, tx0, txz, tzx) ; 14.3.2 Specifying x transition delays If the x transition delays are not explicitly specified, the calculation of delay values for x transitions is based on the following two pessimistic rules: — Transitions from a known state to x shall occur as quickly as possible—that is, the shortest possible delay shall be used for any transition to x. — Transitions from x to a known state shall take as long as possible—that is, the longest possible delay shall be used for any transition from x. Table 48 presents the general algorithm for calculating delay values for x transitions, along with specific examples. The following two groups of x transitions are represented in the table: a) Transition from a known state s to x: s -> x b) Transition from x to a known state s: x -> s Copyright © 2001 IEEE. All rights reserved. 225 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Table 48—Calculating delays for x transitions X transition Delay value General algorithm s -> x x -> s minimum (s -> other known signals) maximum (other known signals -> s) Specific transitions 0 -> x 1 -> x z -> x x -> 0 x -> 1 x -> z minimum (0 -> z delay, 0 -> 1 delay) minimum (1 -> z delay, 1 -> 0 delay) minimum (z -> 1 delay, z -> 0 delay) maximum (z -> 0 delay, 1 -> 0 delay) maximum (z -> 1 delay, 0 -> 1 delay) maximum (1 -> z delay, 0 -> z delay) Usage: (C => Q) = (5, 12, 17, 10, 6, 22) ; 0 -> x 1 -> x z -> x x -> 0 x -> 1 x -> z minimum (17, 5) = 5 minimum (6, 12) = 6 minimum (10, 22) = 10 maximum (22, 12) = 22 maximum (10, 5) = 10 maximum (6, 17) = 17 14.3.3 Delay selection The simulator shall determine the proper delay to use when a specify path output must be scheduled to transition. There may be specify paths to the output from more than one input, and the simulator must decide which specify path to use. The simulator shall do this by first determining which specify paths to the output are active. Active specify paths are those whose input has transitioned most recently in time, and which have either no condition or whose conditions are true. In the presence of simultaneous input transitions, it is possible for many specify paths to an output to be simultaneously active. Once the active specify paths are identified, a delay must be selected from among them. This is done by comparing the correct delay for the specific transition being scheduled from each specify path, and choosing the smallest. Examples: Example 1: (A => Y) = (6, 9); (B => Y) = (5, 11); 226 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C For a Y transition from 0 to 1, if A transitioned more recently than B a delay of 6 will be chosen. But if B transitioned more recently than A, a delay of 5 will be chosen. And if the last time they transitioned A and B did so simultaneously, then the smallest of the two rise delays would be chosen, which is the rise delay from B of 5. The fall delay from A of 9 would be chosen if Y was instead to transition from 1 to 0. Example 2: if (MODE < 5) (A => Y) = (5, 9); if (MODE < 4) (A => Y) = (4, 8); if (MODE < 3) (A => Y) = (6, 5); if (MODE < 2) (A => Y) = (3, 2); if (MODE < 1) (A => Y) = (7, 7); Anywhere from zero to five of these specify paths might be active depending upon the value of MODE. For instance, when MODE is 2 the first three specify paths are active. A rise transition would select a delay of 4, because that is the smallest rise delay among the first three. A fall transition would select a delay of 5, because that is the smallest fall delay among the first three. 14.4 Mixing module path delays and distributed delays If a module contains module path delays and distributed delays (delays on primitive instances within the module), the larger of the two delays for each path shall be used. Examples: Example 1—Figure 37 illustrates a simple circuit modeled with a combination of distributed delays and path delays (only the D input to Q output path is illustrated). Here, the delay on the module path from input D to output Q = 22, while the sum of the distributed delays = 0 + 1 = 1. Therefore, a transition on Q caused by a transition on D will occur 22 time units after the transition on D. 22 A B 0 n = module path delay 1Q n = distributed delay C D 0 Figure 37—Module path delays longer than distributed delays Example 2—In Figure 38, the delay on the module path from D to Q = 22, but the distributed delays along that module path now add up to 10 + 20 = 30. Therefore, an event on Q caused by an event on D will occur 30 time units after the event on D. Copyright © 2001 IEEE. All rights reserved. 227 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 22 A B 10 n = module path delay 20 Q n = distributed delay C D 10 Figure 38—Module path delays shorter than distributed delays 14.5 Driving wired logic Module path output nets shall not have more than one driver within the module. Therefore, wired logic is not allowed at module path outputs. Figure 39 illustrates a violation of this wired-output rule and a method of avoiding the rule violation. E F S G H E F S G H (a) (b) Figure 39—Legal and illegal module paths In Figure 39 (a), any module path to S is illegal because the path destination has two drivers. Assuming signal S in Figure 39 (a) is a wired-and, this limitation can be circumvented by replacing wired logic with gated logic to create a single driver to the output. Figure 39 (b) shows how adding a third and gate—the shaded gate—solves the problem for the module in Figure 39 (a). The example in Figure 40 is also illegal. In this example, when the outputs Q and R are wired together, it creates a condition where both paths have multiple drivers from within the same module. 228 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C A B Q C D R Figure 40—Illegal module paths Although multiple output drivers to a path destination are prohibited inside the same module, they are allowed outside the module. The example in Figure 41 is legal since Q and R each have only one driver within the module in which the module paths are specified. A B Q C D E F R G H Figure 41—Legal module paths 14.6 Detailed control of pulse filtering behavior Two consecutive scheduled transitions closer together in time than the module path delay is deemed a pulse. By default, pulses on a module path output are rejected. Consecutive transitions cannot be closer together than the module path delay, and this is known as the inertial delay model of pulse propagation. Pulse width ranges control how to handle a pulse presented at a module path output. They are: — A pulse width range for which a pulse shall be rejected — A pulse width range for which a pulse shall be allowed to propagate to the path destination — A pulse width range for which a pulse shall generate a logic x on the path destination Copyright © 2001 IEEE. All rights reserved. 229 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Two pulse limit values define the pulse width ranges associated with each module path transition delay. The pulse limit values are called the error limit (e-limit) and the rejection limit (r-limit). The e-limit must always be at least as large as the r-limit. Pulses greater than or equal to the e-limit pass unfiltered. Pulses less than the e-limit but greater than or equal to the r-limit are filtered to X. Pulses less than the r-limit are rejected and no pulse emerges. By default, both the e-limit and the r-limit are set equal to the delay. These default values yield full inertial pulse behavior, rejecting all pulses smaller than the delay. Example: (A =>Y ) = 7, 9; A pulse width = 4 // Module path // delay for a buffer // Pulse considered Y’ // at module path output pulse width = 4 // Pulse is filtered Y The rise delay from input A to output Y is 7, and the fall delay is 9. By default, the e-limit and the r-limit for the rise delay are both 7. The e-limit and the r-limit for the fall delay are both 9. The pulse limits associated with the delay forming the trailing edge of the pulse determine if and how the pulse should be filtered. Waveform Y' shows the waveform resulting from no pulse filtering. The width of the pulse is 2, which is less than the reject limit for the rise delay of 7, and so the pulse is filtered as shown in waveform Y. There are three ways to modify the pulse limits from their default values. First, the Verilog language provides the PATHPULSE$ specparam to modify the pulse limits from their default values. Second, invocation options can specify percentages applying to all module path delays to form the corresponding e-limits and r-limits. Third, SDF annotation can individually annotate the e-limit and r-limit of each module path transition delay. 14.6.1 Specify block control of pulse limit values Pulse limit values may be set from within the specify block with the PATHPULSE$ specparam. The syntax for using PATHPULSE$ to specify the reject limit and error limit values is given in Syntax 14-7. pulse_control_specparam ::= (From Annex A - A.2.4) PATHPULSE$ = ( reject_limit_value [ , error_limit_value ] ) ; | PATHPULSE$specify_input_terminal_descriptor$specify_output_terminal_descriptor = ( reject_limit_value [ , error_limit_value ] ) ; error_limit_value ::= limit_value reject_limit_value ::= limit_value limit_value ::= constant_mintypmax_expression Syntax 14-7—Syntax for PATHPULSE$ pulse control 230 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C If only the reject limit value is specified, it shall apply to both the reject limit and the error limit. The reject limit and error limit may be specified for a specific module path. When no module path is specified, the reject limit and error limit shall apply to all module paths defined in a module. If both path-specific PATHPULSE$ specparams and a non-path-specific PATHPULSE$ specparam appear in the same module, then the path-specific specparams shall take precedence for the specified paths. The module path input terminals and output terminals shall conform to the rules for module path inputs and outputs, with the following restriction: the terminals may not be a bit-select or part-select of a vector. When a module path declaration declares multiple paths, the PATHPULSE$ specparam shall only be specified for the first path input terminal and the first path output terminal. The reject limit and error limit specified shall apply to all other paths in the multiple path declaration. A PATHPULSE$ specparam which specifies anything other than the first path input and path output terminals shall be ignored. Example: In the following example, the path (clk=>q) acquires a reject limit of 2 and an error limit of 9, as defined by the first PATHPULSE$ declaration. The paths (clr*>q) and (pre*>q) receive a reject limit of 0 and an error limit of 4, as specified by the second PATHPULSE$ declaration. The path (data=>q) is not explicitly defined in any of the PATHPULSE$ declarations, and so it acquires reject and error limit of 3, as defined by the last PATHPULSE$ declaration. specify (clk => q) = 12; (data => q) = 10; (clr, pre *> q) = 4; specparam PATHPULSE$clk$q = (2,9), PATHPULSE$clr$q = (0,4), PATHPULSE$ = 3; endspecify 14.6.2 Global control of pulse limit values Two invocation options can specify percentages applying globally to all module path transition delays. The error limit invocation option specifies the percentage of each module path transition delay used for its error limit value. The reject limit invocation option specifies the percentage of each module path transition delay used for its reject limit value. The percentage values shall be an integer between 0 and 100. The default values for both the reject and error limit invocation options are 100%. When neither option is present then 100% of each module transition delay is used as the reject and error limits. It is an error if the error limit percentage is smaller than the reject limit percentage. In such cases the error limit percentage is set equal to the reject limit percentage. When both PATHPULSE$ and global pulse limit invocation options are present, the PATHPULSE$ values shall take precedence. 14.6.3 SDF annotation of pulse limit values SDF annotation can be used to specify the pulse limit values of module path transition delays. Clause 16 describes this in greater detail. Copyright © 2001 IEEE. All rights reserved. 231 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® When both PATHPULSE$, global pulse limit invocation options, and SDF annotation of pulse limit values are present, SDF annotation values shall take precedence. 14.6.4 Detailed pulse control capabilities The default style of pulse filtering behavior has two drawbacks. First, pulse filtering to the X state may be insufficiently pessimistic with an X state duration too short to be useful. Second, unequal delays can result in pulse rejection whenever the trailing edge precedes the leading edge, leaving no indication that a pulse was rejected. This section introduces more detailed pulse control capabilities. 14.6.4.1 On-event versus on-detect pulse filtering When an output pulse must be filtered to X, greater pessimism can be expressed if the module path output transitions immediately to X (on-detect) instead of at the already scheduled transition time of the leading edge of the pulse (on-event). The on-event method of pulse filtering to X is the default. When an output pulse must be filtered to X, the leading edge of the pulse becomes a transition to X and the trailing edge a transition from X. The times of transition of the edges do not change. Just like on-event, the on-detect method of pulse filtering changes the leading edge of the pulse into a transition to X and the trailing edge to a transition from X, but the time of the leading edge is changed to occur immediately upon detection of the pulse. Figure 42 illustrates this behavior using a simple buffer with asymmetric rise/fall times and both the r-limits and e-limits equal to 0. An output waveform is shown for both on-detect and on-event approaches. rise/fall 4/6 in out 10 12 14 18 in out (on-event) (default) out (on-detect) Figure 42—On-detect -vs.- on-event On-detect versus on-event behavior can be selected in two different ways. First, one may be selected globally for all module path outputs through use of the on-detect or on-event invocation option. Second, one may be selected locally through use of specify block pulse style declarations. 232 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE The syntax for pulse style declarations is shown in Syntax 14-8. IEEE Std 1364-2001 Version C pulsestyle_declaration ::= (From Annex A- A.7.1) pulsestyle_onevent list_of_path_outputs ; | pulsestyle_ondetect list_of_path_outputs ; Syntax 14-8—Syntax for pulse style declarations It is an error if a module path output appears in a pulse style declaration after it has already appeared in a module path declaration. The pulse style invocation options take precedence over pulse style specify block declarations. 14.6.4.2 Negative pulse detection When the delays to a module path output are unequal, it is possible for the trailing edge of a pulse to be scheduled for a time earlier than the schedule time of the leading edge, yielding a pulse with a negative width. Under normal operation, if the schedule for a trailing pulse edge is earlier than the schedule for a leading pulse edge, then the leading edge is cancelled. No transition takes place when the initial and final states of the pulse are the same, leaving no indication a schedule was ever present. Negative pulses can be indicated with the X state by use of the showcancelled style of behavior. When the trailing edge of a pulse would be scheduled before the leading edge, this style causes the leading edge to be scheduled to X, and the trailing edge to be scheduled from X. With on-event pulse style, the schedule to X replaces the leading edge schedule. With on-detect pulse style, the schedule to X is made immediately upon detection of the negative pulse. showcancelled behavior can be enabled in two different ways. First, it may be enabled globally for all module path outputs through use of the showcancelled and noshowcancelled invocation options. Second, it may be enabled locally through use of specify block showcancelled declarations. The syntax for showcancelled declarations is shown in Syntax 14-9. showcancelled_declaration ::= (From Annex A- A.7.1) showcancelled list_of_path_outputs ; | noshowcancelled list_of_path_outputs ; Syntax 14-9—Syntax for showcancelled declarations It is an error if a module path output appears in a showcancelled declaration after it has already appeared in a module path declaration. The showcancelled invocation options take precedence over the showcancelled specify block declarations. The showcancelled behavior is illustrated in Figure 43, which shows a narrow pulse presented at the input to a buffer with unequal rise/fall delays. This causes the trailing edge of the pulse to be scheduled earlier than leading edge. The leading edge of the input pulse schedules an output event 6 units later at the point marked by A. The pulse trailing edge occurs one time unit later, which schedules an output event 4 units later marked by point B. This second schedule on the output is for a time prior to the already existing schedule for the leading output pulse edge. Copyright © 2001 IEEE. All rights reserved. 233 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® The output waveform is shown for three different operating modes. The first waveform shows the default behavior with showcancelled behavior not enabled and with the default on-event style. The second waveform shows showcancelled behavior in conjunction with on-event. The last waveform shows showcancelled behavior in conjunction with on-detect. (in=>out)=(4,6); in out in 10 11 15 16 out (default) BA out (showcancelled with on-event) out (showcancelled with on-detect) Figure 43—Current event cancellation problem and correction This same situation can also arise with nearly simultaneous input transitions, which is defined as two inputs transitioning closer together in time than the difference in their respective delays to the output. Figure 44 shows waveforms for a 2-input NAND gate where initially A is high and B is low. B transitions 0->1 at time 10, causing a 1->0 output schedule at time 24. A transitions 1->0 at time 12, causing a 0->1 schedule at time 22. Arrows mark the output transitions caused by the transitions on inputs A and B. The output waveform is shown for three different operating modes. The first waveform shows the default behavior with showcancelled behavior not enabled and with the default on-event style. The second shows showcancelled behavior in conjunction with on-event. The third shows showcancelled behavior in conjunction with on-detect. 234 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C (A=>Q) = 10; (B=>Q) = 14; 10 12 A 22 24 B . out (default) out (showcancelled with on-event) out (showcancelled with on-detect) Figure 44—NAND gate with nearly simultaneous input switching where one event is scheduled prior to another that has not matured One drawback of the on-event style with showcancelled behavior is that as the output pulse edges draw closer together, the duration of the resulting X state becomes smaller. Figure 45 illustrates how the on-detect style solves this problem. Copyright © 2001 IEEE. All rights reserved. 235 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® (A=>Q) = 10 (B=>Q) = 14 10 14 24 A B out (default) out (showcancelled with on-event) out (showcancelled with on-detect) Figure 45—Input NAND gate with nearly simultaneous input switching with output event scheduled at same time. Examples: Example 1: specify (a=>out)=(2,3); (b =>out)=(3,4); endspecify Since no pulse style or showcancelled declarations appear within the specify block, the compiler applies the default modes of on-event and noshowcancelled. Example 2: specify (a=>out)=(2,3); showcancelled out; (b =>out)=(3,4); endspecify 236 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C This showcancelled declaration is in error because it follows use of out in a module path declaration. It would be contradictory for out to have noshowcancelled behavior from input a, but showcancelled behavior from input b. Example 2—Both these specify blocks produce the same result. Outputs out and out_b are both declared showcancelled and on_detect. specify showcancelled out; pulsestyle_ondetect out; (a =>out)=(2,3); (a=>out)=(4,5); showcancelled out_b; pulsestyle_ondetect out_b; (b=>out_b)=(5,6); (b=>out_b)=(3,4); endspecify specify showcancelled out,out_b; pulsestyle_ondetect out,out_b; (a =>out)=(2,3); (b=>out)=(3,4); (a =>out_b)=(3,4); (b=>out_b)=(5,6); endspecify Copyright © 2001 IEEE. All rights reserved. 237 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 15. Timing checks This section describes how timing checks are used in specify blocks to determine if signals obey the timing constraints. 15.1 Overview Timing checks can be placed in specify blocks to verify the timing performance of a design by making sure critical events occur within given time limits. The syntax for system timing checks is given in Syntax 15-1. 238 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C system_timing_check ::= (From Annex A - A.7.5.1) $setup_timing_check | $hold_timing_check | $setuphold_timing_check | $recovery_timing_check | $removal_timing_check | $recrem_timing_check | $skew_timing_check | $timeskew_timing_check | $fullskew_timing_check | $period_timing_check | $width_timing_check | $nochange_timing_check $setup_timing_check ::= $setup ( data_event , reference_event , timing_check_limit [ , [ notify_reg ] ] ) ; $hold_timing_check ::= $hold ( reference_event , data_event , timing_check_limit [ , [ notify_reg ] ] ) ; $setuphold_timing_check ::= $setuphold ( reference_event , data_event , timing_check_limit , timing_check_limit [ , [ notify_reg ] [ , [ stamptime_condition ] [ , [ checktime_condition ] [ , [ delayed_reference ] [ , [ delayed_data ] ] ] ] ] ] ) ; $recovery_timing_check ::= $recovery ( reference_event , data_event , timing_check_limit [ , [ notify_reg ] ] ) ; $removal_timing_check ::= $removal ( reference_event , data_event , timing_check_limit [ , [ notify_reg ] ] ) ; $recrem_timing_check ::= $recrem ( reference_event , data_event , timing_check_limit , timing_check_limit [ , [ notify_reg ] [ , [ stamptime_condition ] [ , [ checktime_condition ] [ , [ delayed_reference ] [ , [ delayed_data ] ] ] ] ] ] ) ; $skew_timing_check ::= $skew ( reference_event , data_event , timing_check_limit [ , [ notify_reg ] ] ) ; $timeskew_timing_check ::= $timeskew ( reference_event , data_event , timing_check_limit [ , [ notify_reg ] [ , [ event_based_flag ] [ , [ remain_active_flag ] ] ] ] ) ; $fullskew_timing_check ::= $fullskew ( reference_event , data_event , timing_check_limit , timing_check_limit [ , [ notify_reg ] [ , [ event_based_flag ] [ , [ remain_active_flag ] ] ] ] ) ; $period_timing_check ::= $period ( controlled_reference_event , timing_check_limit [ , [ notify_reg ] ] ) ; $width_timing_check ::= $width ( controlled_reference_event , timing_check_limit , threshold [ , [ notify_reg ] ] ) ; $nochange_timing_check ::= $nochange ( reference_event , data_event , start_edge_offset , end_edge_offset [ , [ notify_reg ] ] ) ; Syntax 15-1—Syntax for system timing checks Copyright © 2001 IEEE. All rights reserved. 239 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® The syntax for check time conditions and timing check events is given in Syntax 15-2. checktime_condition ::= (From Annex A - A.7.5.2) mintypmax_expression controlled_reference_event ::= controlled_timing_check_event data_event ::= timing_check_event delayed_data ::= terminal_identifier | terminal_identifier [ constant_mintypmax_expression ] delayed_reference ::= terminal_identifier | terminal_identifier [ constant_mintypmax_expression ] end_edge_offset ::= mintypmax_expression event_based_flag ::= constant_expression notify_reg ::= variable_identifier reference_event ::= timing_check_event remain_active_flag ::= constant_mintypmax_expression stamptime_condition ::= mintypmax_expression start_edge_offset ::= mintypmax_expression threshold ::=constant_expression timing_check_limit ::= expression timing_check_event ::= (From Annex A - A.7.5.3) [timing_check_event_control] specify_terminal_descriptor [ &&& timing_check_condition ] controlled_timing_check_event ::= timing_check_event_control specify_terminal_descriptor [ &&& timing_check_condition ] timing_check_event_control ::= posedge | negedge | edge_control_specifier specify_terminal_descriptor ::= specify_input_terminal_descriptor | specify_output_terminal_descriptor edge_control_specifier ::= edge [ edge_descriptor [ , edge_descriptor ] ] edge_descriptora ::= 01 | 10 | z_or_x zero_or_one | zero_or_one z_or_x zero_or_one ::= 0 | 1 z_or_x ::= x | X | z | Z timing_check_condition ::= scalar_timing_check_condition | ( scalar_timing_check_condition ) scalar_timing_check_condition ::= expression | ~ expression | expression == scalar_constant | expression === scalar_constant | expression != scalar_constant | expression !== scalar_constant scalar_constant ::= 1'b0 | 1'b1 | 1'B0 | 1'B1 | 'b0 | 'b1 | 'B0 | 'B1 | 1 | 0 aEmbedded spaces are illegal. Syntax 15-2—Syntax for check time conditions and timing check events 240 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C For ease of presentation, the timing checks are divided into two groups. The first group of timing checks are described in terms of stability time windows: $setup $recovery $hold $removal $setuphold $recrem The timing checks in the second group check clock and control signals, and are described in terms of the difference in time between two events (the $nochange check involves three events): $skew $width $timeskew $period $fullskew $nochange Although they begin with a $, timing checks are not system tasks. The leading $ is present because of historical reasons, and timing checks shall not be confused with system tasks. In particular, no system task can appear in a specify block, and no timing check can appear in procedural code. Some timing checks can accept negative limit values. This topic is discussed in detail in 15.8. All timing checks have both a reference event and a data event, and boolean conditions can be associated with each. Some of the checks have two signal arguments, one of which is the reference event and the other the data event. Other checks have only one signal argument, and the reference and data events are derived from it. Reference events and data events shall only be detected by timing checks when their associated conditions are true. See 15.6 for more information about conditions in timing checks. Timing check evaluation is based upon the times of two events, called the timestamp event and the timecheck event. A transition on the timestamp event signal causes the simulator to record (stamp) the time of transition for future use in evaluating the timing check. A transition on the timecheck event signal causes the simulator to actually evaluate the timing check to determine whether a violation has taken place. For some checks the reference event is always the timestamp event, while the data event is always the timecheck event, while for other checks the reverse is true. And for yet other checks the decision as to which is the timestamp and which the timecheck event is based upon factors to be discussed later in greater detail. Every timing check can include an optional notifier which toggles whenever the timing check detects a violation. The model can use the notifier to make behavior a function of timing check violations. Notifiers are discussed in greater detail in 15.5. Like expressions for module path delays, timing check limit values are constant expressions which can include specparams. 15.2 Timing checks using a stability window These timing checks are discussed in this section: $setup $recovery $hold $removal $setuphold $recrem These checks accept two signals, the reference event and the data event, and define a time window with respect to one signal while checking the time of transition of the other signal with respect to the window. In general they all perform the following steps: a) Define a time window with respect to the reference signal using the specified limit or limits; b) Check the time of transition of the data signal with respect to the time window; c) Report a timing violation if the data signal transitions within the time window. Copyright © 2001 IEEE. All rights reserved. 241 IEEE Std 1364-2001 Version C 15.2.1 $setup The $setup timing check syntax is shown in Syntax 15-3. IEEE STANDARD VERILOG® $setup_timing_check ::= (From Annex A - A.7.5.1) $setup ( data_event , reference_event , timing_check_limit [ , [ notify_reg ] ] ) ; data_event ::= (From Annex A - A.7.5.2) timing_check_event notify_reg ::= variable_identifier reference_event ::= timing_check_event timing_check_limit ::= expression Syntax 15-3—Syntax for $setup Table 49 defines the $setup timing check. Table 49—$setup arguments Argument Description data_event reference_event limit notifier (optional) Timestamp event Timecheck event Non-negative constant expression Reg The data event is usually a data signal, while the reference event is usually a clock signal. The endpoints of the time window are determined as follows: (beginning of time window) = (timecheck time) - limit (end of time window) = (timecheck time) The $setup timing check reports a timing violation in the following case: (beginning of time window) < (timestamp time) < (end of time window) The endpoints of the time window are not part of the violation region. When the limit is zero, the $setup check shall never issue a violation. 15.2.2 $hold The $hold timing check syntax is shown in Syntax 15-4. 242 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C $hold_timing_check ::= (From Annex A - A.7.5.1) $hold ( reference_event , data_event , timing_check_limit [ , [ notify_reg ] ] ) ; data_event ::= (From Annex A - A.7.5.2) timing_check_event notify_reg ::= variable_identifier reference_event ::= timing_check_event timing_check_limit ::= expression Syntax 15-4—Syntax for $hold Table 50 defines the $hold timing check. Table 50—$hold arguments Argument Description reference_event data_event limit notifier (optional) Timestamp event Timecheck event Non-negative constant expression Reg The data event is usually a data signal, while the reference event is usually a clock signal. The endpoints of the time window are determined as follows: (beginning of time window) = (timestamp time) (end of time window) = (timestamp time) + limit The $hold timing check reports a timing violation in the following case: (beginning of time window) <= (timecheck time) < (end of time window) Only the end of the time window is not part of the violation region. When the limit is zero, the $hold check shall never issue a violation. 15.2.3 $setuphold The $setuphold timing check syntax is shown in Syntax 15-5. Copyright © 2001 IEEE. All rights reserved. 243 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® $setuphold_timing_check ::= (From Annex A - A.7.5.1) $setuphold ( reference_event , data_event , timing_check_limit , timing_check_limit [ , [ notify_reg ] [ , [ stamptime_condition ] [ , [ checktime_condition ] [ , [ delayed_reference ] [ , [ delayed_data ] ] ] ] ] ] ) ; checktime_condition ::= (From Annex A - A.7.5.2) mintypmax_expression data_event ::= timing_check_event delayed_data ::= terminal_identifier | terminal_identifier [ constant_mintypmax_expression ] delayed_reference ::= terminal_identifier | terminal_identifier [ constant_mintypmax_expression ] notify_reg ::= variable_identifier reference_event ::= timing_check_event stamptime_condition ::= mintypmax_expression timing_check_limit ::= expression Syntax 15-5—Syntax for $setuphold Table 51 defines the $setuphold timing check. Table 51—$setuphold arguments Argument reference_event data_event setup_limit hold_limit notifier (optional) timestamp_cond (optional) timecheck_cond (optional) delayed_reference (optional) delayed_data (optional) Description Timecheck or timestamp event when setup limit is positive Timestamp event when setup limit is negative Timecheck or timestamp event when hold limit is positive Timestamp event when hold limit is negative Constant expression Constant expression Reg Timestamp condition for negative timing checks Timecheck condition for negative timing checks Delayed reference signal for negative timing checks Delayed data signal for negative timing checks The $setuphold timing check can accept negative limit values. This is discussed in greater detail in 15.8. The data event is usually a data signal, while the reference event is usually a clock signal. 244 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C When both the setup limit and the hold limit are positive, either the reference event or the data event can be the timecheck event. It shall depend upon which occurs first in the simulation. When either the setup limit or the hold limit is negative the restriction becomes: setup_limit + hold_limit > (simulation unit of precision) The $setuphold timing check combines the functionality of the $setup and $hold timing checks into a single timing check. Therefore, the following invocation: $setuphold( posedge clk, data, tSU, tHLD ); is equivalent in functionality to the following, if tSU and tHLD are not negative: $setup( data, posedge clk, tSU ); $hold( posedge clk, data, tHLD ); When both setup and hold limits are positive and the data event occurs first, the endpoints of the time window are determined as follows: (beginning of time window) = (timecheck time) - limit (end of time window) = (timecheck time) And the $setuphold timing check reports a timing violation in the following case: (beginning of time window) < (timecheck time) <= (end of time window) Only the beginning of the time window is not part of the violation region. The $setuphold check shall report a timing violation when the reference and data events occur simultaneously. When both setup and hold limits are positive and the data event occurs second, the endpoints of the time window are determined as follows: beginning of time window) = (timestamp time) (end of time window) = (timestamp time) + limit And the $setuphold timing check reports a timing violation in the following case: (beginning of time window) <= (timecheck time) < (end of time window) Only the end of the time window is not part of the violation region. The $setuphold check shall report a timing violation when the reference and data events occur simultaneously. When both limits are zero, the $setuphold check shall never issue a violation. 15.2.4 $removal The $removal timing check syntax is shown in Syntax 15-6. Copyright © 2001 IEEE. All rights reserved. 245 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® $removal_timing_check ::= (From Annex A - A.7.5.1) $removal ( reference_event , data_event , timing_check_limit [ , [ notify_reg ] ] ) ; data_event ::= (From Annex A - A.7.5.2) timing_check_event notify_reg ::= variable_identifier reference_event ::= timing_check_event timing_check_limit ::= expression Syntax 15-6—Syntax for $removal Table 52 defines the $removal timing check. Table 52—$removal arguments Argument Description reference_event data_event limit notifier (optional) Timestamp event Timecheck event Non-negative constant expression Reg The reference event is usually a control signal like clear, reset or set, while the data event is usually a clock signal. The endpoints of the time window are determined as follows: (beginning of time window) = (timecheck time) - limit (end of time window) = (timecheck time) The $removal timing check reports a timing violation in the following case: (beginning of time window) < (timestamp time) < (end of time window) The endpoints of the time window are not part of the violation region. When the limit is zero, the $removal check shall never issue a violation. 15.2.5 $recovery The $recovery timing check syntax is shown in Syntax 15-7. 246 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C $recovery_timing_check ::= (From Annex A - A.7.5.1) $recovery ( reference_event , data_event , timing_check_limit [ , [ notify_reg ] ] ) ; data_event ::= (From Annex A - A.7.5.2) timing_check_event notify_reg ::= variable_identifier reference_event ::= timing_check_event timing_check_limit ::= expression Syntax 15-7—Syntax for $recovery Table 53 defines the $recovery timing check. Table 53—$recovery arguments Argument Description reference_event Timestamp event data_event Timecheck event limit Non-negative constant expression notifier (optional) Reg The reference event is usually a control signal like clear, reset or set, while the data event is usually a clock signal. The endpoints of the time window are determined as follows: (beginning of time window) = (timestamp time) (end of time window) = (timestamp time) + limit The $recovery timing check reports a timing violation in the following case: (beginning of time window) <= (timecheck time) < (end of time window) Only the end of the time window is not part of the violation region. When the limit is zero, the $recovery check shall never issue a violation. 15.2.6 $recrem The $recrem timing check syntax is shown in Syntax 15-8. Copyright © 2001 IEEE. All rights reserved. 247 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® $recrem_timing_check ::= (From Annex A - A.7.5.1) $recrem ( reference_event , data_event , timing_check_limit , timing_check_limit [ , [ notify_reg ] [ , [ stamptime_condition ] [ , [ checktime_condition ] [ , [ delayed_reference ] [ , [ delayed_data ] ] ] ] ] ] ) ; checktime_condition ::= (From Annex A - A.7.5.2) mintypmax_expression data_event ::= timing_check_event delayed_data ::= terminal_identifier | terminal_identifier [ constant_mintypmax_expression ] delayed_reference ::= terminal_identifier | terminal_identifier [ constant_mintypmax_expression ] notify_reg ::= variable_identifier reference_event ::= timing_check_event stamptime_condition ::= mintypmax_expression timing_check_limit ::= expression Syntax 15-8—Syntax for $recrem Table 54 defines the $recrem timing check. Table 54—$recrem arguments Argument reference_event data_event recovery_limit removal_limit notifier (optional) timestamp_cond (optional) timecheck_cond (optional) delayed_reference (optional) delayed_data (optional) Description Timecheck or timestamp event when removal limit is positive Timestamp event when removal limit is negative Timecheck or timestamp event when recovery limit is positive Timestamp event when recovery limit is negative Constant expression Constant expression Reg Timestamp condition for negative timing checks Timecheck condition for negative timing checks Delayed reference signal for negative timing checks Delayed data signal for negative timing checks The $recrem timing check can accept negative limit values. This is discussed in greater detail in 15.8. When both the removal limit and the recovery limit are positive, either the reference event or the data event can be the timecheck event. It shall depend upon which occurs first in the simulation. 248 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C When either the removal limit or the recovery limit is negative the restriction becomes: removal_limit + recovery_limit > (simulation unit of precision) The $recrem timing check combines the functionality of the $removal and $recovery timing checks into a single timing check. Therefore, the following invocation: $recrem( posedge clear, posedge clk, tREC, tREM ); is equivalent in functionality to the following, if tREC and tREM are not negative: $removal( posedge clear, posedge clk, tREM ); $recovery( posedge clear, posedge clk, tREC ); When both removal and recovery limits are positive and the data event occurs first, the endpoints of the time window are determined as follows: (beginning of time window) = (timecheck time) - limit (end of time window) = (timecheck time) And the $recrem timing check reports a timing violation in the following case: (beginning of time window) < (timecheck time) <= (end of time window) Only the beginning of the time window is not part of the violation region. The $recrem check shall report a timing violation when the reference and data events occur simultaneously. When both removal and recovery limits are positive and the data event occurs second, the endpoints of the time window are determined as follows: (beginning of time window) = (timestamp time) (end of time window) = (timestamp time) + limit And the $recrem timing check reports a timing violation in the following case: (beginning of time window) <= (timecheck time) < (end of time window) Only the end of the time window is not part of the violation region. The $recrem check shall report a timing violation when the reference and data events occur simultaneously. When both limits are zero, the $recrem check shall never issue a violation. 15.3 Timing checks for clock and control signals The following timing checks are discussed in this section: $skew $timeskew $fullskew $period $width $nochange These checks accept one or two signals and verify transitions on them are never separated by more than the limit. For those checks specifying only one signal, the reference event and data event are derived from that one signal. In general these checks all perform the following steps: Copyright © 2001 IEEE. All rights reserved. 249 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® a) Determine the elapsed time between two events; b) Compare the elapsed time to the specified limit; c) Report a timing violation if the elapsed time violates the limit. The skew checks have two different violation detection mechanisms, event-based and timer-based. Eventbased skew checking is performed only when a signal transitions, while timer-based skew checking takes place as soon as the simulation time equal to the skew limit has elapsed. The $nochange check involves three events rather than two. 15.3.1 $skew The $skew timing check syntax is shown in Syntax 15-9. $skew_timing_check ::= (From Annex A - A.7.5.1) $skew ( reference_event , data_event , timing_check_limit [ , [ notify_reg ] ] ) ; data_event ::= (From Annex A - A.7.5.2) timing_check_event notify_reg ::= variable_identifier reference_event ::= timing_check_event timing_check_limit ::= expression Syntax 15-9—Syntax for $skew Table 55 defines the $skew timing check. Table 55—$skew arguments Argument reference_event data_event limit notifier (optional) Description Timestamp event Timecheck event Non-negative constant expression Reg The $skew timing check reports a violation in the following case: (timecheck time) - (timestamp time) > limit Simultaneous transitions on the reference and data signals can never cause $skew to report a timing violation, even when the skew limit value is zero. 250 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The $skew timing check is event-based; it is evaluated only after a data event. If there is never a data event (i.e., the data event is infinitely late), the $skew timing check shall never be evaluated, and no timing violation shall ever be reported. In contrast, the $timeskew and $fullskew checks are timer-based by default, and they shall be used if violation reports are absolutely required and the data event can be very late or even absent altogether. These checks are discussed in 15.3.2 and 15.3.3. $skew shall wait indefinitely for the data event once it has detected a reference event and it shall not report a timing violation until the data event takes place. A second consecutive reference event shall cancel the old wait for the data event and begin a new one. After a reference event, the $skew timing check shall never stop checking data events for a timing violation. $skew shall report timing violations for all data events occurring beyond the limit after a reference event. 15.3.2 $timeskew The syntax for $timeskew is shown in Syntax 15-10. $timeskew_timing_check ::= (From Annex A - A.7.5.1) $timeskew ( reference_event , data_event , timing_check_limit [ , [ notify_reg ] [ , [ event_based_flag ] [ , [ remain_active_flag ] ] ] ] ) ; data_event ::= (From Annex A - A.7.5.2) timing_check_event event_based_flag ::= constant_expression notify_reg ::= variable_identifier reference_event ::= timing_check_event remain_active_flag ::= constant_mintypmax_expression timing_check_limit ::= expression Syntax 15-10—Syntax for $timeskew Table 56 defines the $timeskew timing check arguments. Table 56—$timeskew arguments Argument reference_event data_event limit notifier (optional) event_based_flag (optional) remain_active_flag (optional) Description Timestamp event Timecheck event Non-negative constant expression Reg Constant expression Constant expression Copyright © 2001 IEEE. All rights reserved. 251 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® The $timeskew timing check reports a violation only in the following cases: (timecheck time) - (timestamp time) > limit Simultaneous transitions on the reference and data signals can never cause $timeskew to report a timing violation, even when the skew limit value is zero. The default behavior for $timeskew is timer-based. Violations are reported immediately upon an elapse of time after the reference event equal to the limit, and the check shall become dormant and report no more violations (even in response to data events) until after the next reference event. This check shall also become dormant if it detects a reference event when its condition is false. The $timeskew check's default timer-based behavior can be altered to event-based using the event based flag. It behaves like the $skew check when both the event based flag and the remain active flag are set. The $timeskew check behaves like the $skew check when only the event based flag is set, except it becomes dormant after reporting the first violation. Example: $timeskew (posedge CP &&& MODE, negedge CPN, 50); MODE CP 50 F AB CPN C DE Figure 46—Sample $timeskew Case 1: Event based flag and remain active flag not set. After the first reference event on CP at A, a violation is reported at B as soon as 50 time units have passed. No further violations are reported. Case 2: Event based flag set, remain active flag not set. The negative transition on CPN at point C shall cause a timing violation. Subsequent negative transitions at points D and E do not cause violations. The second reference event at F occurs while MODE is false, turning the $timeskew check dormant, and no further violations are reported. Case 3: Event based flag set, remain active flag set. The first three negative transitions on CPN at points C, D and E shall cause timing violations. The second reference event at F occurs while MODE is false, turning the $timeskew check dormant, and no further violations are reported. Case 4: Event based flag and remain active flag both set. Every negative edge on CPN is reported as a violation, which is identical to $skew behavior. 252 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE 15.3.3 $fullskew The syntax for $fullskew is shown in Syntax 15-11. IEEE Std 1364-2001 Version C $fullskew_timing_check ::= (From Annex A - A.7.5.1) $fullskew ( reference_event , data_event , timing_check_limit , timing_check_limit [ , [ notify_reg ] [ , [ event_based_flag ] [ , [ remain_active_flag ] ] ] ] ) ; data_event ::= (From Annex A - A.7.5.2) timing_check_event event_based_flag ::= constant_expression notify_reg ::= variable_identifier reference_event ::= timing_check_event remain_active_flag ::= constant_mintypmax_expression timing_check_limit ::= expression Syntax 15-11—Syntax for $fullskew Table 57 defines the $fullskew timing check arguments. Table 57—$fullskew arguments Argument Description reference_event data_event limit 1 limit 2 notifier (optional) event_based_flag (optional) remain_active_flag (optional) Timestamp or timecheck event Timestamp or timecheck event Non-negative constant expression Non-negative constant expression Reg Constant expression Constant expression $fullskew is identical to $timeskew except the reference and data events can transition in either order. The first limit is the maximum time by which the data event can follow the reference event. The second limit is the maximum time by which the reference event can follow the data event. The reference event is the timestamp event and the data event is the timecheck event when the reference event precedes the data event. The data event is the timestamp event and the reference event is the timecheck event when the data event precedes the reference event. The $fullskew timing check reports a violation only in the following case, where limit is set to limit1 when the reference event transitions first, and to limit2 when the data event transitions first: (timecheck time) - (timestamp time) > limit Copyright © 2001 IEEE. All rights reserved. 253 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Simultaneous transitions on the reference and data signals shall never cause $fullskew to report a timing violation, even when the skew limit value is zero. The default behavior for $fullskew is timer-based. Violations shall be reported immediately upon an elapse of time after the timestamp event equal to the limit. It then becomes dormant and reports no more violations, even in response to timecheck events, until after the next timestamp event. This check shall also become dormant if it detects a timestamp event when the associated condition is false. The $fullskew check's default timer-based behavior can be altered to event-based using the event based flag. It behaves like the $skew check when both the event based flag and the remain active flag are set. The $timeskew check behaves like the $skew check when only the event based flag is set, except it becomes dormant after it reports the first violation. Example: $fullskew (posedge CP &&& MODE, negedge CPN, 50, 70); MODE CP J 50 A BB 70 70 CPN D C E F GHI Figure 47—Sample $fullskew Case 1: Event based flag and remain active flag not set. The transition at A of CP while MODE is true begins a wait for a negative transition on CPN, and a violation is reported at B as soon as a period of time equal to 50 time units has passed. This resets the check and readies it for the next active transition. A negative transition on CPN occurs next at C, beginning a wait for a positive transition on CP while MODE is true. At D a time equal to 70 time units has passed without a positive edge on CP while MODE is true, so a violation is reported and the check is again reset to await the next active transition. A transition on CPN at E also results in a timing violation, as does the transition at F, because even though CP transitions, MODE is no longer true. Transitions at G and H also result in timing violations, but not the transition at I, because it is followed by a positive transition on CP while MODE is true. Case 2: Event based flag set, remain active flag not set. The transition at A of CP while MODE is true begins a wait for a negative transition on CPN, and a violation is reported at C on CPN because it occurs beyond the 50 time unit limit. This transition at C also begins a wait of 70 time units for a positive transition on CP while MODE is true. But for transitions on CPN at B 254 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C through H there is no positive transition on CP while MODE is true, and so no timing violations are reported. The transition at I on CPN begins a wait of 70 time units, and this is satisfied by the positive transition on CP at J while MODE is true. Case 3: Event based flag and remain active flag both set. The transition at A of CP while MODE is true begins a wait for a negative transition on CPN, and a violation is reported at C on CPN, and it shall also begin a wait for a positive transition on CP while MODE is true. No such transition on CP ever takes place after CPN transitions C through H, but no violations are reported because CP never experiences a positive transition while MODE is true. Transition I also reports no violation because a positive transition at I on CP while MODE is true occurs within the 70 time unit skew limit. 15.3.4 $width The $width timing check syntax is shown in Syntax 15-12. $width_timing_check ::= (From Annex A - A.7.5.1) $width ( controlled_reference_event , timing_check_limit , threshold [ , [ notify_reg ] ] ) ; controlled_reference_event ::= (From Annex A - A.7.5.2) controlled_timing_check_event notify_reg ::= variable_identifier threshold ::= constant_expression timing_check_limit ::= expression Syntax 15-12—Syntax for $width If the comma before the threshold is present, the comma before the notifier shall also be present, even though both arguments are optional. Table 58 defines the $width timing check. Table 58—$width arguments Argument reference_event (data_event - implicit) limit threshold (optional) notifier (optional) Description Timestamp edge triggered event Timecheck edge triggered event Non-negative constant expression Non-negative constant expression Reg The $width timing check monitors the width of signal pulses by measuring the time from the timestamp event to the timecheck event. Since a data event is not passed to $width, it is derived from the reference event, as follows: data event = reference event signal with opposite edge Copyright © 2001 IEEE. All rights reserved. 255 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Because of the way the data event is derived for $width, an edge triggered event has to be passed as the reference event. A compilation error shall occur if the reference event is not an edge specification. While the $width timing check can be defined in terms of a time window, it is simpler to express it as the difference between the timecheck and timestamp times. The $width timing check reports a violation in the following case: threshold < (timecheck time) - (timestamp time) < limit The pulse width has to be greater than or equal to limit in order to avoid a timing violation, but no violation is reported for glitches smaller than the threshold. The threshold argument shall be included if the notifier argument is required. It is permissible to not specify both the threshold and notifier arguments, making the default value for the threshold zero (0). If the notifier is present, a non-null value for the threshold shall also be present. Here is a legal $width check when the notifier is required and the threshold is not: $width (posedge clk, 6, 0, ntfr_reg); The data event and the reference event shall never occur at the same simulation time because these events are triggered by opposite transitions. Example: The following example demonstrates some examples of legal and illegal calls: // Legal Calls $width ( negedge clr, lim ); $width ( negedge clr, lim, thresh, notif ); $width ( negedge clr, lim, 0, notif ); // Illegal Calls $width ( negedge clr, lim, , notif ); $width ( negedge clr, lim, notif ); 15.3.5 $period The $period timing check syntax is shown in Syntax 15-13. $period_timing_check ::= (From Annex A - A.7.5.1) $period ( controlled_reference_event , timing_check_limit [ , [ notify_reg ] ] ) ; controlled_reference_event ::= (From Annex A - A.7.5.2) controlled_timing_check_event notify_reg ::= variable_identifier timing_check_limit ::= expression Syntax 15-13—Syntax for $period 256 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Table 59 defines the $period timing check. Table 59—$period arguments Argument Description reference_event (data_event - implicit) limit notifier (optional) Timestamp edge triggered event Timestamp edge triggered event Non-negative constant expression Reg Since the data event is not passed as an argument to $period, it is derived from the reference event, as follows: data event = reference event signal with the same edge Because of the way the data event is derived for $period, an edge triggered event shall be passed as the reference event. A compilation error shall occur if the reference event is not an edge specification. While the $period timing check can be defined in terms of a time window, it is simpler to express it as the difference between the timecheck and timestamp times.The $period timing check reports a violation in the following case: (timecheck time) - (timestamp time) < limit 15.3.6 $nochange The $nochange syntax is shown in Syntax 15-14. $nochange_timing_check ::= (From Annex A - A.7.5.1) $nochange ( reference_event , data_event , start_edge_offset , end_edge_offset [ , [ notify_reg ] ] ) ; data_event ::= (From Annex A - A.7.5.2) timing_check_event end_edge_offset ::= mintypmax_expression notify_reg ::= variable_identifier reference_event ::= timing_check_event start_edge_offset ::= mintypmax_expression Syntax 15-14—Syntax for $nochange Copyright © 2001 IEEE. All rights reserved. 257 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Table 60 defines the $nochange timing check arguments. Table 60—$nochange arguments Argument Description reference_event data_event start_edge_offset end_edge_offset notifier (optional) Edge triggered timestamp and/or timecheck event Timestamp or timecheck event Constant expression Constant expression Reg The $nochange timing check reports a timing violation if the data event occurs during the specified level of the control signal (the reference event). The reference event can be specified with the posedge or the negedge keyword, but the edge control specifiers (see 15.4) can not be used. The start edge and end edge offsets can expand or shrink the timing violation region, which is defined by the duration of the reference event signal after the edge. For example, if the reference event is a posedge, then the duration is the period during which the reference signal is high. A positive offset for start edge extends the region by starting the timing violation region earlier; a negative offset for start edge shrinks the region by starting the region later. Similarly, a positive offset for the end edge extends the timing violation region by ending it later, while a negative offset for the end edge shrinks the region by ending it earlier. If both the offsets are zero, the size of the region shall not change. Unlike other timing checks, $nochange involves three, rather than two, transitions. The leading edge of the reference event defines the beginning of the time window, while the trailing edge of the reference event defines the end of the time window. A violation results if the data event occurs anytime within the time window. The endpoints of the time window are determined as follows: (beginning of time window) = (leading reference edge time) start_edge_offset (end of time window) = (trailing reference edge time) + end_edge_offset The $nochange timing check reports a timing violation in the following case: (beginning of time window) < (data event time) < (end of time window) The endpoints of the time window are not included. The values of start_edge_offset and end_edge_offset play a significant role in determining which signal, the reference event or the data event, is the timestamp or timecheck event. Example: $nochange( posedge clk, data, 0, 0) ; In this example, $nochange system task shall report a violation if the data signal changes while clk is high. It shall not be a violation if posedge clk and a transition on data occur simultaneously. 15.4 Edge-control specifiers The edge-control specifiers can be used to control events in timing checks based on specific edge transitions 258 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C between 0, 1, and x. Syntax 15-15 shows the syntax for edge-control specifiers. edge_control_specifier ::= (From Annex A - A.7.5.3) edge [ edge_descriptor [ , edge_descriptor ] ] edge_descriptora ::= 01 | 10 | z_or_x zero_or_one | zero_or_one z_or_x zero_or_one ::= 0 | 1 z_or_x ::= x | X | z | Z aEmbedded spaces are illegal. Syntax 15-15—Syntax for edge control specifier Edge-control specifiers contain the keyword edge followed by a square bracketed list of from one to six pairs of edge transitions between 0, 1 and x, as follows: 01 Transition from 0 to 1 0x Transition from 0 to x 10 Transition from 1 to 0 1x Transition from 1 to x x0 Transition from x to 0 x1 Transition from x to 1 Edge transitions involving z are treated the same way as edge transitions involving x. The posedge and negedge keywords can be used as a shorthand for certain edge-control specifiers. For example, the construct: posedge clr is equivalent to the following: edge[01, 0x, x1] clr Similarly, the construct negedge clr is the same as the following: edge[10, x0, 1x] clr However, edge-control specifiers offer the flexibility to declare edge transitions other than posedge and negedge. Copyright © 2001 IEEE. All rights reserved. 259 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 15.5 Notifiers: user-defined responses to timing violations Timing check notifiers detect timing check violations behaviorally, and, therefore, take an action as soon as a violation occurs. Such notifiers can be used to print an informative error message describing the violation or to propagate an x value at the output of the device which reported the violation. The notifier is a reg—declared in the module where timing check tasks are invoked—which is passed as the last argument to a system timing check. Whenever a timing violation occurs, the system task updates the value of the notifier. The notifier is an optional argument to all system timing checks and can be omitted from the system task call without adversely affecting its operation. Table 61 shows how the notifier values are toggled when timing violations occur. Table 61—User-defined responses to timing violations BEFORE violation AFTER violation x 0 0 1 1 0 z z Examples: Example 1 $setup( data, posedge clk, 10, notify_reg ) ; $width( posedge clk, 16, notify_reg ) ; Example 2—Consider a more complex example of how to use notifiers in a behavioral model. The following example uses a notifier to set the D flip-flop output to x when a timing violation occurs in an edge-sensitive UDP. 260 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C primitive posdff_udp(q, clock, data, preset, clear, notifier); output q; reg q; input clock, data, preset, clear, notifier; table //clock data p c notifier state q //------------------------------------- r 0 11 ? : ? :0; r 1 11 ? : ? :1; p 1 ?1 ? : 1 :1; p 0 1? ? : 0 :0; n ? ?? ? : ? :-; ? * ?? ? : ? :-; ? ? 01 ? : ? :1; ? ? *1 ? : 1 :1; ? ? 10 ? : ? :0; ? ? 1* ? : 0 :0; ? ? ? ? * : ? : x ; // At any notifier event // output x endtable endprimitive Copyright © 2001 IEEE. All rights reserved. 261 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® module dff(q, qbar, clock, data, preset, clear); output q, qbar; input clock, data, preset, clear; reg notifier; and (enable, preset, clear); not (qbar, ffout); buf (q, ffout); posdff_udp (ffout, clock, data, preset, clear, notifier); specify // Define timing check specparam values specparam tSU = 10, tHD = 1, tPW = 25, tWPC = 10, tREC = 5; // Define module path delay rise and fall min:typ:max values specparam tPLHc = 4:6:9 , tPHLc = 5:8:11; specparam tPLHpc = 3:5:6 , tPHLpc = 4:7:9; // Specify module path delays (clock *> q,qbar) = (tPLHc, tPHLc); (preset,clear *> q,qbar) = (tPLHpc, tPHLpc); // Setup time : data to clock, only when preset and clear are 1 $setup(data, posedge clock &&& enable, tSU, notifier); // Hold time: clock to data, only when preset and clear are 1 $hold(posedge clock, data &&& enable, tHD, notifier); // Clock period check $period(posedge clock, tPW, notifier); // Pulse width : preset, clear $width(negedge preset, tWPC, 0, notifier); $width(negedge clear, tWPC, 0, notifier); // Recovery time: clear or preset to clock $recovery(posedge preset, posedge clock, tREC, notifier); $recovery(posedge clear, posedge clock, tREC, notifier); endspecify endmodule NOTE—This model applies to edge-sensitive UDPs only; for level-sensitive models, an additional UDP for x propagation has to be generated. 15.5.1 Requirements for accurate simulation In order to accurately model negative value timing checks: a) A timing violation shall be triggered if the signal changes in the violation window, exclusive of the endpoints. Violation windows smaller than two units of simulation precision can not yield timing violations. b) The value of the latched data shall be the one which is stable during the violation window, again, exclusive of the endpoints. 262 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C To facilitate these modeling requirements, delayed copies of the data and reference signals are generated in the timing checks, and these are used internally for timing check evaluation at run-time. The setup and hold times used internally are adjusted so as to shift the violation window and make it overlap the reference signal. Delayed data and reference signals can be declared within the timing check so they can be used in the model's functional implementation to insure accurate simulation. If no delayed signals are declared in the timing check, and if a negative setup or hold value is present, then implicit delayed signals are created. Since implicit delayed signals can not be used in defining model behavior, such a model can possibly behave incorrectly. Examples: Example 1: $setuphold(posedge CLK, DATA, -10, 20); Implicit delayed signals shall be created for CLK and DATA, but it shall not be possible to access them. The $setuphold check shall be properly evaluated, but functional behavior shall not always be accurate. The old DATA value shall be incorrectly clocked in if DATA transitions between posedge CLK and 10 time units later. Example 2: $setuphold(posedge CLK, DATA1, -10, 20); $setuphold(posedge CLK, DATA2, -15, 18); Implicit delayed signals shall be created for CLK, DATA1 and DATA2, one for each. Even though CLK is referenced in two different timing checks, only one implicit delayed signal is created, and it is used for both timing checks. Example 3: If a given signal has a delayed signal in some timing checks but not in others, the delayed signal shall be used in both cases: $setuphold(posedge CLK, DATA1, -10, 20,,,, del_CLK, del_DATA1); $setuphold(posedge CLK, DATA2, -15, 18); Explicit delayed signals of del_CLK and del_DATA1 are created for CLK and DATA1, while an implicit delayed signal is created for DATA2. In other words, CLK has only one delayed signal created for it, del_CLK, rather than one explicit delayed signal for the first check, and another implicit delayed signal for the second check. The delayed versions of the signals, whether implicit or explicit, shall be used in the $setup, $hold, $setuphold, $recovery, $removal, $recrem, $width, $period and $nochange timing checks, and these checks shall have their limits adjusted accordingly. This ensures the notifier shall be toggled at the proper moment. If the adjusted limit becomes less than or equal to 0, the limit shall be set to 0 and the simulator shall issue a warning. The delayed versions of the signals shall not be used for the $skew, $fullskew and $timeskew timing checks because it can possibly result in the reversal of the order of signal transitions. This causes the notifiers for these timing checks to toggle at the wrong time relative to the rest of the model, perhaps resulting in transitions to X due to a timing check violation being canceled. This issue shall be addressed in the model, possibly by using separate notifiers for these checks. Copyright © 2001 IEEE. All rights reserved. 263 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® It is possible for a set of negative timing check values to be mutually inconsistent and produce no solution for the delay values of delayed signals. In these situations the simulator shall issue a warning message. The inconsistency shall be resolved by changing the smallest negative limit value to 0 and recalculating the delays for the delayed signals, and this shall be repeated until a solution is reached. This procedure shall always produce a solution because in the worst case all negative limit values become 0, and no delayed signals are needed. The delayed timing check signals are only actually delayed when negative limit values are present. If a timing check signal becomes delayed by more than the propagation delay from that signal to an output, that output shall take longer than its propagation delay to change. It shall instead transition at the same time which the delayed timing check signal changes. Thus, the output shall behave as if its specify path delay were equal to the delay applied to the timing check signal. This situation can only arise when unique setup/hold or removal/recovery times are given for each edge of the data signal. Example: (CLK = Q) = 6; $setuphold (posedge CLK, posedge D, -3, 8, , , , dCLK, dD); $setuphold (posedge CLK, negedge D, -7, 13, , , , dCLK, dD); The setup time of -7 (the larger in absolute value of -3 and -7) creates a delay of 7 for dCLK, and so output Q shall not change until 7 time units after a positive edge on CLK, rather than the 6 time units given in the specify path. 15.5.2 Conditions in negative timing checks Conditions can be associated with both the reference and data signals by using the &&& operator, but when either the setup or hold time is negative the conditions need to be paired with reference and data signals in a more flexible way. This example illustrates why. This pair of $setup and $hold checks work together to provide the same check as a single $setuphold: $setup (data, clk&&&cond1, tsetup, ntfr); $hold (clk, data&&&cond1, thold, ntfr); clk is the timecheck event for the $setup check, while data is the timecheck event for the $hold check. This can not be represented in a single $setuphold check, and so additional arguments are provided to make this possible. These arguments are timestamp_cond and timecheck_cond, and they immediately follow the notifier (see 15.2.3). This $setuphold check is equivalent to the separate $setup and $hold checks shown above: $setuphold( clk, data, tsetup, thold, ntfr, , cond1); The timestamp_cond argument is null, while the timecheck_cond argument is cond1. The timestamp_cond and timecheck_cond arguments are associated with either the reference or data signals based on which delayed version of these signals occurs first. timestamp_cond is associated with the delayed signal which transitions first, while timecheck_cond is associated with the delayed signal which transitions second. Delayed signals are only created for the reference and data signals, and not for any condition signals associated with them. Therefore, timestamp_cond and timecheck_cond are not implicitly delayed by the simulator. Delayed condition signals for the timestamp_cond and timecheck_cond fields can be created by making them a function of the delayed signals. 264 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Example: assign TE_cond_D = (dTE !== 1’b1); assign TE_cond_TI = (dTE !== 1’b0); assign DXTI_cond = (dTI !== dD); specify $setuphold(posedge CP, D, -10, 20, notifier, ,TE_cond_D, dCP, dD); $setuphold(posedge CP, TI, 20, -10, notifier, ,TE_cond_TI, dCP, dTI); $setuphold(posedge CP, TE, -4, 8, notifier, ,DXTI_cond, dCP, dTE); endspecify The assign statements create condition signals which are functions of the delayed signals. Creating delayed signal conditions synchronizes the conditions with the delayed versions of the reference and data signals used to perform the checks. The first $setuphold has a negative setup time, and so the timecheck condition TE_cond_D is associated with data signal D. The second $setuphold has a negative hold time, and so the timecheck condition TE_cond_TI is associated with reference signals CP. The third $setuphold has a negative setup time, and so the timecheck condition DXTI_cond is associated with data signal TE. The violation windows for the example are shown in Figure 48. CP 500 D 510 520 TE 504 508 TI 480 490 Figure 48—Timing check violation windows These are the delay values calculated for the delayed signals: dCP 10.01 dD 0.00 dTI 20.02 dTE 2.02 Use of delayed signals in creating the signals for the timestamp_cond and timecheck_cond arguments is not required, but it is usually closer to actual device behavior. Copyright © 2001 IEEE. All rights reserved. 265 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 15.5.3 Notifiers in negative timing checks Because the reference and data signals are delayed internally, the detection of the timing violation is also delayed. Notifier regs in negative timing checks shall be toggled when the timing check detects a timing violation, which occurs when the delayed signals as measured by the adjusted timing check values are in violation, not when the undelayed signals at the model inputs as measured by the original timing check values are in violation. 15.5.4 Option behavior As already mentioned, the ability of Verilog simulators to handle negative values in $setuphold and $recrem timing checks shall be enabled with an invocation option. It is possible models written to accept negative timing check values with delayed reference and/or delayed data signals can be run without this invocation option enabled. In this circumstance the delayed reference and data signals become copies of the original reference and data signals. The same occurs if an invocation option turning off all timing checks is used. 15.6 Enabling timing checks with conditioned events A construct called a conditioned event ties the occurrence of timing checks to the value of a conditioning signal. Syntax 15-16 shows the syntax for controlled timing check event. timing_check_event ::= (From Annex A - A.7.5.3) [timing_check_event_control] specify_terminal_descriptor [ &&& timing_check_condition ] controlled_timing_check_event ::= timing_check_event_control specify_terminal_descriptor [ &&& timing_check_condition ] timing_check_event_control ::= posedge | negedge | edge_control_specifier specify_terminal_descriptor ::= specify_input_terminal_descriptor | specify_output_terminal_descriptor timing_check_condition ::= scalar_timing_check_condition | ( scalar_timing_check_condition ) scalar_timing_check_condition ::= expression | ~ expression | expression == scalar_constant | expression === scalar_constant | expression != scalar_constant | expression !== scalar_constant scalar_constant ::= 1'b0 | 1'b1 | 1'B0 | 1'B1 | 'b0 | 'b1 | 'B0 | 'B1 | 1 | 0 Syntax 15-16—Syntax for controlled timing check event The comparisons used in the condition can be deterministic, as in ===, !==, ~, or no operation, or nondeterministic, as in == or !=. When comparisons are deterministic, an x value on the conditioning signal shall not enable the timing check. For nondeterministic comparisons, an x on the conditioning signal shall enable the timing check. 266 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The conditioning signal shall be a scalar net; if a vector net or an expression resulting in a multi-bit value is used, then the least significant bit of the vector net or the expression value is used. If more than one conditioning signal is required for conditioning timing checks, appropriate logic shall be combined in a separate signal outside the specify block, which can be used as the conditioning signal. Examples: Example 1—To illustrate the difference between conditioned and unconditioned timing check events, consider the following example with unconditioned timing check: $setup( data, posedge clk, 10 ); Here, a setup timing check shall occur every time there is a positive edge on the signal clk. To trigger the setup check on the positive edge on the signal clk only when the signal clr is high, rewrite the command as $setup( data, posedge clk &&& clr, 10 ) ; Example 2—This example shows two ways to trigger the same timing check as in example 1 (on the positive clk edge) only when clr is low. The second method uses the === operator, which makes the comparison deterministic. $setup( data, posedge clk &&& (~clr), 10 ) ; $setup( data, posedge clk &&& (clr===0), 10 ); Example 3—To perform the previous sample setup check on the positive clk edge only when clr and set are high, add the following statement outside the specify block: and new_gate( clr_and_set, clr, set ); Then add the condition to the timing check using the signal clr_and_set as follows: $setup( data, posedge clk &&& clr_and_set, 10 ); 15.7 Vector signals in timing checks Either or both signals in a timing check can be a vector. This shall be interpreted as a single timing check where the transition of one or more bits of a vector is considered a single transition of that vector. Example: module DFF (Q, CLK, DAT); input CLK; input [7:0] DAT; output [7:0] Q; always @(posedge clk) Q = DAT; specify $setup (DAT, posedge CLK, 10); endspecify endmodule If DAT transitions from 'b00101110 to 'b01010011 at time 100, and CLK transitions from 0 to 1 at time 105, then the $setup timing check shall still only report a single timing violation. Copyright © 2001 IEEE. All rights reserved. 267 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Simulators can provide an option causing vectors in timing checks to result in the creation of multiple single-bit timing checks. For timing checks with only a single signal, such as $period or $width, a vector of width N results in N unique timing checks. For timing checks with two signals, such as $setup, $hold, $setuphold, $skew, $timeskew, $fullskew, $recovery, $removal, $recrem and $nochange, where M and N are the widths of the signals, the result is M*N unique timing checks. If there is a notifier, all the timing checks trigger that notifier. With such an option enabled, the above example yields six timing violation because six bits of DAT transitioned. 15.8 Negative timing checks Both the $setuphold and $recrem timing checks can accept negative values when the negative timing check option is enabled. The behavior of these two timing checks is identical with respect to negative values. The descriptions in this section are for the $setuphold timing check, but apply equally to the $recrem timing check. The setup and hold timing check values define a timing violation window with respect to the reference signal edge during which the data shall remain constant. Any change of the data during the specified window causes a timing violation. The timing violation is reported and, through the notifier reg, other actions can take place in the model, such as forcing the output of a flip-flop to X when it detects a timing violation. A positive value for both setup and hold times implies this violation window straddles the reference signal shown in Figure 49. clock data ..........Setup time (+) ..........Hold Time (+) Figure 49—Data constraint interval, positive setup/hold A negative hold or setup time means the violation window is shifted to either before or after the reference edge. This can happen in a real device because of disparate internal device delays between the internal clock and data signal paths. These internal device delays are illustrated in Figure 50 showing how significant differences in these delays can cause negative setup or hold values. 268 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE data clock ASIC Cell D1 Seq. Elem. D2 IEEE Std 1364-2001 Version C output clock data Negative Hold time (D1>D2) ..........Setup time (+) ..........Hold Time (-) Negative Setup time (D2>D1) clock data ..........Setup time (-) ..........Hold Time (+) Figure 50—Data constraint interval, negative setup/hold Copyright © 2001 IEEE. All rights reserved. 269 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 16. Backannotation using the Standard Delay Format (SDF) SDF files contain timing values for specify path delays, specparam values, timing check constraints, and interconnect delays. SDF files can also contain other information in addition to simulation timing, but these need not concern Verilog simulation. The timing values in SDF files usually come from ASIC delay calculation tools that take advantage of connectivity, technology, and layout geometry information. Verilog backannotation is the process by which timing values from the SDF file update specify path delays, specparam values, timing constraint values, and interconnect delays. All this information is covered further in IEEE Std 1497-2001 [B2]. 16.1 The SDF annotator The term SDF Annotator refers to any tool capable of backannotating SDF data to a Verilog simulator. It shall report a warning for any data it is unable to annotate. An SDF file can contain many constructs which are not related to specify path delays, specparam values, timing check constraint values, or interconnect delays. An example is any construct in the TIMINGENV section of the SDF file. All constructs unrelated to Verilog timing shall be ignored without any warnings issued. Any Verilog timing value for which the SDF file does not provide a value shall not be modified during the backannotation process, and its pre-backannotation value shall be unchanged. 16.2 Mapping of SDF constructs to Verilog SDF timing values appear within a CELL declaration, which can contain one or more of DELAY, TIMINGCHECK and LABEL sections. The DELAY section contains propagation delay values for specify paths and interconnect delays. The TIMINGCHECK section contains timing check constraint values. The LABEL section contains new values for specparams. Backannotation into Verilog is done by matching SDF constructs to the corresponding Verilog declarations, then replacing the existing Verilog timing values with those from the SDF file. 16.2.1 Mapping of SDF delay constructs to Verilog declarations When annotating DELAY constructs that are not interconnect delays (covered in 16.2.3), the SDF annotator looks for specify paths where the names and conditions match. When annotating TIMINGCHECK constructs, the SDF annotator looks for timing checks of the same type where the names and conditions match. Table 62 shows which Verilog structures can be annotated by each SDF construct in the DELAY section. 270 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Table 62—Mapping of SDF delay constructs to Verilog declarations SDF Construct (PATHPULSE... (PATHPULSEPERCENT... (IOPATH... (IOPATH (RETAIN... (COND (IOPATH... (COND (IOPATH (RETAIN... (CONDELSE (IOPATH... (CONDELSE (IOPATH (RETAIN... (DEVICE... (DEVICE port_instance... Verilog annotated structure Conditional and non-conditional specify path pulse limits Conditional and non-conditional specify path pulse limits Conditional and non-conditional specify path delays/pulse limits Conditional and non-conditional specify path delays/pulse limits, RETAIN ignored without warning Conditional specify path delays/pulse limits Conditional specify path delays/pulse limits, RETAIN ignored without warning ifnone ifnone, RETAIN ignored without warning All specify paths to module outputs. If no specify paths, all primitives driving module outputs. If port_instance is a module instance, all specify paths to module outputs. If no specify paths, all primitives driving module outputs. If port_instance is a module instance output, all specify paths to that module output. If no specify path, all primitives driving that module output. In this example the source SDF signal sel matches the source Verilog signal, and the destination SDF signal zout also matches the destination Verilog signal, and so the rise/fall times of 1.3 and 1.7 are annotated to the specify path. SDF file: (IOPATH sel zout (1.3) (1.7)) Verilog specify path: (sel => zout) = 0; A conditional IOPATH delay between two ports shall annotate only to Verilog specify paths between those same two ports with the same condition. In this example the rise/fall times of 1.3 and 1.7 are annotated only to the second specify path. SDF file: (COND mode (IOPATH sel zout (1.3) (1.7))) Verilog specify paths: if (!mode) (sel => zout) = 0; if (mode) (sel => zout) = 0; A non-conditional IOPATH delay between two ports shall annotate to all Verilog specify paths between those same two ports. In this example the rise/fall times of 1.3 and 1.7 are annotated to both specify paths. Copyright © 2001 IEEE. All rights reserved. 271 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® SDF file: (IOPATH sel zout (1.3) (1.7)) Verilog specify paths: if (!mode) (sel => zout) = 0; if (mode) (sel => zout) = 0; 16.2.2 Mapping of SDF timing check constructs to Verilog Table 63 shows which Verilog timing checks are annotated to by each type of SDF timing check. v1 is the first value of a timing check, v2 is the second value, while x indicates no value is annotated. Table 63—Mapping of SDF timing check constructs to Verilog SDF Timing Check Annotated Verilog Timing checks (SETUP v1... $setup(v1), $setuphold(v1,x) (HOLD v1... $hold(v1), $setuphold(x,v1) (SETUPHOLD v1 v2... $setup(v1), $hold(v2), $setuphold(v1,v2) (RECOVERY v1... $recovery(v1), $recrem(v1,x) (REMOVAL v1... $removal(v1), $recrem(x,v1) (RECREM v1 v2... $recovery(v1), $removal(v2), $recrem(v1,v2) (SKEW v1... $skew(v1) (TIMESKEW v1...a $timeskew(v1) (FULLSKEW v1 v2... a $fullskew(v1,v2) (WIDTH v1... $width(v1,x) (PERIOD v1... $period(v1) (NOCHANGE v1 v2... $nochange(v1,v2)b aNot part of current SDF standard bNot usually implemented in Verilog simulators The reference and data signals of timing checks can have logical condition expressions and edges associated with them. An SDF timing check with no conditions or edges on any of its signals shall match all corresponding Verilog timing checks regardless of whether conditions are present or not. In this example the SDF timing check shall annotate to all the Verilog timing checks: SDF file: (SETUPHOLD data clk (3) (4)) Verilog timing checks: $setuphold (posedge clk&&& mode, data, 1, 1, ntfr); $setuphold (negedge clk&&&!mode, data, 1, 1, ntfr); 272 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C When conditions and/or edges are associated with the signals in an SDF timing check, then they shall match those in any corresponding Verilog timing check before annotation shall happen. In this example the SDF timing check shall annotate to the first Verilog timing check, but not the second: SDF file: (SETUPHOLD data (posedge clk) (3) (4)) Verilog timing checks: $setuphold (posedge clk&&& mode, data, 1, 1, ntfr); // Annotated $setuphold (negedge clk&&&!mode, data, 1, 1, ntfr); // Not annotated Here, the SDF timing check shall not annotate to any of the Verilog timing checks: SDF file: (SETUPHOLD data (COND !mode (posedge clk)) (3) (4)) Verilog timing checks: $setuphold (posedge clk&&& mode, data, 1, 1, ntfr); // Not annotated $setuphold (negedge clk&&&!mode, data, 1, 1, ntfr); // Not annotated 16.2.3 SDF annotation of specparams The SDF LABEL construct annotates to specparams. Any expression containing one or more specparams is reevaluated when annotated to from an SDF file. This example shows SDF LABEL constructs annotating to specparams in a Verilog module. The specparams are used in procedural delays to control when the clock transitions. The SDF LABEL construct annotates the values of dhigh and dlow, thereby setting the period and duty cycle of the clock. SDF file: (LABEL (ABSOLUTE (dhigh 60) (dlow 40))) Verilog file: module clock(clk); output clk; reg clk; specparam dhigh=0, dlow=0; initial clk = 0; always begin #dhigh clk = 1; // Clock remains low for time dlow // before transitioning to 1 #dlow clk = 0; // Clock remains high for time dhigh // before transitioning to 0 end; endmodule Copyright © 2001 IEEE. All rights reserved. 273 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® This example shows a specparam in an expression of a specify path. The SDF LABEL construct can be used to change the value of the specparam and cause reevaluation of the expression: specparam cap = 0; ... specify (A => Z) = 1.4 * cap + 0.7; endspecify 16.2.4 SDF annotation of interconnect delays SDF interconnect delay annotation differs from annotation of other constructs described above in that there exists no corresponding Verilog declaration to which to annotate. In Verilog simulation, interconnect delays are an abstraction that represents the signal propagation delay from an output or inout module port to an input or inout module port. The INTERCONNECT construct includes a source, a load, and delay values, while the PORT and NETDELAY constructs include only a load and delay values. Interconnect delays can only be annotated between module ports, never between primitive pins. Table 64 shows how the SDF interconnect constructs in the DELAY section are annotated: Table 64—SDF annotation of interconnect delays SDF Construct Verilog annotated structure (PORT... Interconnect delay (NETDELAY a Interconnect delay (INTERCONNECT... Interconnect delay aOnly OVI SDF version 1.0, 2.0, and 2.1, and IEEE SDF version 4.0 Interconnect delays can be annotated to both single source and multi-source nets. When annotating a PORT construct, the SDF annotator shall search for the port and if it exists shall annotate an interconnect delay to that port which shall represent the delay from all sources on the net to that port. When annotating a NETDELAY construct, the SDF annotator shall check to see if it is annotating to a port or a net. If it is a port then the SDF annotator shall annotate an interconnect delay to that port. If it is a net then it shall annotate an interconnect delay to all load ports connected to that net. If the port or net has more than one source then the delay shall represent the delay from all sources. NETDELAY delays can only be annotated to input or inout module ports, or to nets. In the case of multi-source nets, unique delays can be annotated between each source/load pair using the INTERCONNECT construct. When annotating this construct, the SDF annotator shall find the source port and the load port, and if both exist it shall annotate an interconnect delay between the two. If the source port is not found, or if the source port and the load port are not actually on the same net, then a warning message is issued, but the delay to the load port is annotated anyway. If this happens for a load port that is part of a multi-source net, then the delay is treated as if it were the delay from all source ports, which is the same as the annotation behavior for a PORT delay. Source ports shall be output or input ports, while load ports shall be input or inout ports. Interconnect delays share many of the characteristics of specify path delays. The same rules for specify path delays for filling in missing delays and pulse limits also apply for interconnect delays. Interconnect delays have twelve transition delays, and unique reject and error pulse limits are associated with each of the twelve. An unlimited number of future schedules are permitted. 274 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C In a Verilog module, a reference to an annotated port, wherever it occurs, whether in $monitor and $display statements or in expressions, shall provide the delayed signal value. A reference to the source shall yield the undelayed signal value, while a reference to the load shall yield the delayed signal value. In general, references to the signal value hierarchically before the load shall yield the undelayed signal value, while references to the signal at or hierarchically after the load shall yield the delayed signal value. An annotation to a hierarchical port shall affect all connected ports at higher or lower hierarchical levels, depending on the direction of annotation. An annotation from a source port shall be interpreted as being from all sources hierarchically higher or lower than that source port. Up-hierarchy annotations shall be properly handled. This situation arises when the load is hierarchically above the source. The delay to all ports hierarchically above the load or which connect to the net at points hierarchically above the load is the same as the delay to that load. Down-hierarchy annotation shall also be properly handled. This situation arises when the source is hierarchically above the load. The delay to the load is interpreted as being from all ports at or above the source or which connect to the net at points hierarchically above the source. Hierarchically overlapping annotations are permitted. This occurs when annotations to or from the same port take place at different hierarchical levels, and therefore do not correspond to the same hierarchical subset of ports. In this example, the first INTERCONNECT statement annotates to all ports of the net that are at or hierarchically within i53/selmode, while the second annotates to a smaller subset of ports, only those at or hierarchically within i53/u21/in: (INTERCONNECT i14/u5/out i53/selmode (1.43) (2.17)) (INTERCONNECT i14/u5/out i53/u21/in (1.58) (1.92)) Overlapping annotations can occur in many different ways, particularly on multi-source/multi-load nets, and SDF annotation shall properly resolve all the interactions. 16.3 Multiple annotations SDF annotation is an ordered process. The constructs from the SDF file are annotated in their order of occurrence. This means that annotation of an SDF construct can be changed by annotation of a subsequent construct that either modifies (INCREMENT) or overwrites (ABSOLUTE) it. These do not have to be the same construct. This example first annotates pulse limits to an IOPATH, then annotates the entire IOPATH, thereby overwriting the pulse limits that were just annotated: (DELAY (ABSOLUTE (PATHPULSE A Z (2.1) (3.4)) (IOPATH A Z (3.5) (6.1)) Overwriting the pulse limits can be avoided by using empty parentheses to hold the current values of the pulse limits: (DELAY (ABSOLUTE (PATHPULSE A Z (2.1) (3.4)) (IOPATH A Z ((3.5) () ()) ((6.1) () ()) ) The above annotation can be simplified into a single statement like this: (DELAY (ABSOLUTE (IOPATH A Z ((3.5) (2.1) (3.4)) ((6.1) (2.1) (3.4)) ) Copyright © 2001 IEEE. All rights reserved. 275 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® A PORT annotation followed by an INTERCONNECT annotation to the same load shall cause only the delay from the INTERCONNECT source to be affected. For this net with three sources and a single load, the delay from all sources except i13/out remains 6: (DELAY (ABSOLUTE (PORT i15/in (6)) (INTERCONNECT i13/out i15/in (5)) An INTERCONNECT annotation followed by a PORT annotation shall cause the INTERCONNECT annotation to be overwritten. Here, the delays from all sources to the load shall become 6. (DELAY (ABSOLUTE (INTERCONNECT i13/out i15/in (5)) (PORT i15/in (6)) 16.4 Multiple SDF files More than one SDF file can be annotated. Each call to the $sdf_annotate task annotates the design with timing information from an SDF file. Annotated values either modify (INCREMENT) or overwrite (ABSOLUTE) values from earlier SDF files. Different regions of a design can be annotated from different SDF files by specifying the region’s hierarchy scope as the second argument to $sdf_annotate. 16.5 Pulse limit annotation For SDF annotation of delays (not timing constraints), the default values annotated for pulse limits shall be calculated using the percentage settings for the reject and error limits. By default these limits are 100%, but they can be modified through invocation options. For example, assuming invocation options have set the reject limit to 40% and the error limit to 80%, this SDF construct shall annotate a delay of 5, a reject limit of 2, and an error limit of 4: (DELAY (ABSOLUTE (IOPATH A Z (5)) Given that the specify path delay was originally 0, this annotation results in a delay of 5 and pulse limits of 0: (DELAY (ABSOLUTE (IOPATH A Z ((5) () ()) ) Annotations in INCREMENT mode can result in pulse limits less than 0, in which case they shall be adjusted to 0. For example, if the specify path pulse limits were both 3, this annotation results in a 0 value for both pulse limits: (DELAY (INCREMENT (IOPATH A Z (() (-4) (-5)) ) There are two SDF constructs that annotate only to pulse limits, PATHPULSE and PATHPULSEPERCENT. They do not affect the delay. When PATHPULSE sets the pulse limits to values greater than the delay Verilog shall exhibit the same behavior as if the pulse limits had been set equal to the delay. 276 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 16.6 SDF to Verilog delay value mapping Verilog specify paths and interconnects can have unique delays for up to twelve state transitions (see 14.3.1). All other constructs, such as gate primitives and continuous assignments, can have only three state transition delays (see 7.14). For Verilog specify path and interconnect delays, the number of transition delay values provided by SDF might be less than twelve. Table 65 shows how fewer than twelve SDF delays are extended to be twelve delays. The Verilog transition types are shown down the left-hand side, while the number of SDF delays provided is shown across the top. The SDF values are given the names v1 through v12. Table 65—SDF to Verilog delay value mapping Verilog transition 0 -> 1 1 -> 0 0 -> z z -> 1 1 -> z z -> 0 0 -> x x -> 1 1 -> x x -> 0 x -> z z -> x 1 value v1 v1 v1 v1 v1 v1 v1 v1 v1 v1 v1 v1 Number of SDF delay values provided 2 values 3 values 6 values v1 v2 v1 v1 v2 v2 v1 v1 v2 v2 max(v1,v2) min(v1,v2) v1 v2 v3 v1 v3 v2 min(v1,v3) v1 min(v2,v3) v2 v3 min(v1,v2) v1 v2 v3 v4 v5 v6 min(v1,v3) max(v1,v4) min(v2,v5) max(v2,v6) max(v3,v5) min(v4,v6) 12 values v1 v2 v3 v4 v5 v6 v7 v8 v9 v10 v11 v12 For other delays that can have at most three values, the expansion of less than three SDF delays into three Verilog delays is covered in Table 39. More than three SDF delays are reduced to three Verilog delays by simply ignoring the extra delays. The delay to the X-state is created from the minimum of the other three delays. Copyright © 2001 IEEE. All rights reserved. 277 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 17. System tasks and functions This clause describes system tasks and functions that are considered part of the Verilog HDL. These system tasks and functions are divided into ten categories as follows: Display tasks $display $displayb $displayh $displayo $monitor $monitorb $monitorh $monitoro $monitoroff File I/O tasks $fclose $fdisplay $fdisplayb $fdisplayh $fdisplayo $fgetc $fflush $fgets $fmonitor $fmonitorb $fmonitorh $fmonitoro $readmemb $swrite $swriteo $sformat $fscanf $fread $fseek Timescale tasks $printtimescale Simulation control tasks $finish [17.1] $strobe $strobeb $strobeh $strobeo $write $writeb $writeh $writeo $monitoron [17.2] $fopen $fstrobe $fstrobeb $fstrobeh $fstrobeo $ungetc $ferror $rewind $fwrite $fwriteb $fwriteh $fwriteo $readmemh $swriteb $swriteh $sdf_annotate $sscanf $ftell [17.3] $timeformat [17.4] $stop PLA modeling tasks $async$and$array $async$nand$array $async$or$array $async$nor$array $sync$and$array $sync$nand$array $sync$or$array $sync$nor$array [17.5] $async$and$plane $async$nand$plane $async$or$plane $async$nor$plane $sync$and$plane $sync$nand$plane $sync$or$plane $sync$nor$plane Stochastic analysis tasks $q_initialize $q_remove $q_exam $q_add $q_full [17.6] Simulation time functions $realtime $time $stime [17.7] Conversion functions $bitstoreal $itor $signed $realtobits $rtoi $unsigned [17.8] Probabilistic distribution functions [17.9] $dist_chi_square $dist_exponential $dist_poisson $dist_uniform $dist_erlang $dist_normal $dist_t $random Command line input [17.10] $test$plusargs $value$plusargs These utility tasks and functions provide some broadly useful capabilities. The following clauses describe the behavior of these tasks and functions. Additional tasks for value change dump (VCD) are described in Clause 18. 17.1 Display system tasks The display group of system tasks are divided into three categories: the display and write tasks, strobed monitoring tasks, and continuous monitoring tasks. 278 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE 17.1.1 The display and write tasks IEEE Std 1364-2001 Version C display_tasks ::= display_task_name ( list_of_arguments ) ; display_task_name ::= $display | $displayb | $displayo | $displayh | $write | $writeb | $writeo | $writeh Syntax 17-1—Syntax for $display and $write system tasks These are the main system task routines for displaying information. The two sets of tasks are identical except that $display automatically adds a newline character to the end of its output, whereas the $write task does not. The $display and $write tasks display their arguments in the same order as they appear in the argument list. Each argument can be a quoted string, an expression that returns a value, or a null argument. The contents of string arguments are output literally except when certain escape sequences are inserted to display special characters or to specify the display format for a subsequent expression. Escape sequences are inserted into a string in three ways: — The special character \ indicates that the character to follow is a literal or nonprintable character (see Table 66). — The special character % indicates that the next character should be interpreted as a format specification that establishes the display format for a subsequent expression argument (see Table 67). For each % character, with the exception of %m that appears in a string, a corresponding expression argument shall be supplied after the string. — The special character string %% indicates the display of the percent sign character % (see Table 66). Any null argument produces a single space character in the display. (A null argument is characterized by two adjacent commas in the argument list.) The $display task, when invoked without arguments, simply prints a newline character. A $write task supplied without parameters prints nothing at all. 17.1.1.1 Escape sequences for special characters The escape sequences given in Table 66, when included in a string argument, cause special characters to be displayed. Copyright © 2001 IEEE. All rights reserved. 279 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Example: Table 66—Escape sequences for printing special characters Argument Description \n The newline character \t The tab character \\ The \ character \" The " character \ddd A character specified by 1 to 3 octal digits %% The % character module disp; initial begin $display("\\\t\\\n\"\123"); end endmodule Simulating this example shall display the following: \\ "S 17.1.1.2 Format specifications Table 67 shows the escape sequences used for format specifications. Each escape sequence, when included in a string argument, specifies the display format for a subsequent expression. For each % character (except %m) that appears in a string, a corresponding expression shall follow the string in the argument list. The value of the expression replaces the format specification when the string is displayed. Any expression argument that has no corresponding format specification is displayed using the default decimal format in $display and $write, binary format in $displayb and $writeb, octal format in $displayo and $writeo, and hexadecimal format in $displayh and $writeh. Table 67—Escape sequences for format specifications Argument %h or %H %d or %D %o or %O %b or %B %c or %C %l or %L %v or %V Description Display in hexadecimal format Display in decimal format Display in octal format Display in binary format Display in ASCII character format Display library binding information Display net signal strength 280 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Table 67—Escape sequences for format specifications (continued) %m or %M %s or %S %t or %T %u or %U %z or %Z Display hierarchical name Display as a string Display in current time format Unformatted 2 value data Unformatted 4 value data The formatting specification %l (or %L) is defined for displaying the library information of the specific module. This information shall be displayed as "library.cell" corresponding to the library name the current module instance was extracted from and the cell name of the current module instance. See Clause 13 for information on libraries and configuring designs. The formatting specification %u (or %U) is defined for writing data without formatting (binary values). The application shall transfer the 2 value binary representation of the specified data to the output stream. This escape sequence can be used with any of the existing display system tasks, although $fwrite should be the preferred one to use. Any unknown or high-impedance bits in the source shall be treated as zero. This formatting specifier is intended to be used to support transferring data to and from external programs that have no concept of x and z. Applications that require preservation of x and z are encouraged to use the %z I/O format specification. The data shall be written to the file in the native endian format of the underlying system (i.e., in the same endian order as if the PLI was used, and the C language write (2) system call was used). The data shall be written in units of 32 bits with the word containing the LSB written first. NOTE—For POSIX applications: It might be necessary to open files for unformatted I/O with the wb, wb+, or w+b specifiers, to avoid the systems implementation of I/O altering patterns in the unformatted stream that match special characters. The formatting specification %z (or %Z) is defined for writing data without formatting (binary values). The application shall transfer the 4 value binary representation of the specified data to the output stream. This escape sequence can be used with any of the existing display system tasks, although $fwrite should be the preferred one to use. This formatting specifier is intended to be used to support transferring data to and from external programs that recognize and support the concept of x and z. Applications that do not require the preservation of x and z are encouraged to use the %u I/O format specification. The data shall be written to the file in the native endian format of the underlying system (i.e., in the same endian order as if the PLI was used, and the data were in a s_vpi_vecval structure (See 27.14, Figure 179), and the C language write(2) system call was used to write the structure to disk). The data shall be written in units of 32 bits with the structure containing the LSB written first. NOTE—For POSIX applications: It might be necessary to open files for unformatted I/O with the wb, wb+ or w+b specifiers, to avoid the systems implementation of I/O altering patterns in the unformatted stream that match special characters. The format specifications in Table 68 are used with real numbers and have the full formatting capabilities available in the C language. For example, the format specification %10.3g specifies a minimum field width of 10 with 3 fractional digits. Copyright © 2001 IEEE. All rights reserved. 281 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Table 68—Format specifications for real numbers Argument Description %e or %E %f or %F %g or %G Display ‘real’ in an exponential format Display ‘real’ in a decimal format Display ‘real’ in exponential or decimal format, whichever format results in the shorter printed output The net signal strength, hierarchical name, and string format specifications are described in 17.1.1.5 through 17.1.1.7. The %t format specification works with the $timeformat system task to specify a uniform time unit, time precision, and format for reporting timing information from various modules that use different time units and precisions. The $timeformat task is described in 17.3.2. Example: module disp; reg [31:0] rval; pulldown (pd); initial begin rval = 101; $display("rval = %h hex %d decimal",rval,rval); $display("rval = %o octal\nrval = %b bin",rval,rval); $display("rval has %c ascii character value",rval); $display("pd strength value is %v",pd); $display("current scope is %m"); $display("%s is ascii value for 101",101); $display("simulation time is %t", $time); end endmodule Simulating this example shall display the following: rval = 00000065 hex 101 decimal rval = 00000000145 octal rval = 00000000000000000000000001100101 bin rval has e ascii character value pd strength value is StX current scope is disp e is ascii value for 101 17.1.1.3 Size of displayed data For expression arguments, the values written to the output file (or terminal) are sized automatically. For example, the result of a 12-bit expression would be allocated three characters when displayed in hexadecimal format and four characters when displayed in decimal format, since the largest possible value for the expression is FFF (hexadecimal) and 4095 (decimal). 282 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C When displaying decimal values, leading zeros are suppressed and replaced by spaces. In other radices, leading zeros are always displayed. The automatic sizing of displayed data can be overridden by inserting a zero between the % character and the letter that indicates the radix, as shown in the following example. $display("d=%0h a=%0h", data, addr); Example: module printval; reg [11:0] r1; initial begin r1 = 10; $display( "Printing with maximum size - :%d: :%h:", r1,r1 ); $display( "Printing with minimum size - :%0d: :%0h:", r1,r1 ); end endmodule Printing with maximum size - : 10: :00a: Printing with minimum size - :10: :a: In this example, the result of a 12-bit expression is displayed. The first call to $display uses the standard format specifier syntax and produces results requiring four and three columns for the decimal and hexadecimal radices, respectively. The second $display call uses the %0 form of the format specifier syntax and produces results requiring two columns and one column, respectively. 17.1.1.4 Unknown and high impedance values When the result of an expression contains an unknown or high impedance value, the following rules apply to displaying that value. In decimal (%d) format — If all bits are at the unknown value, a single lowercase x character is displayed. — If all bits are at the high impedance value, a single lowercase z character is displayed. — If some, but not all, bits are at the unknown value, the uppercase X character is displayed. — If some, but not all, bits are at the high impedance value, the uppercase Z character is displayed. — Decimal numerals always appear right-justified in a fixed-width field. In hexadecimal (%h) and octal (%o) formats — Each group of 4 bits is represented as a single hexadecimal digit; each group of 3 bits is represented as a single octal digit. — If all bits in a group are at the unknown value, a lowercase x is displayed for that digit. — If all bits in a group are at a high impedance state, a lowercase z is printed for that digit. — If some, but not all, bits in a group are unknown, an uppercase X is displayed for that digit. — If some, but not all, bits in a group are at a high impedance state, then an uppercase Z is displayed for that digit. In binary (%b) format, each bit is printed separately using the characters 0, 1, x, and z. Copyright © 2001 IEEE. All rights reserved. 283 IEEE Std 1364-2001 Version C Example: IEEE STANDARD VERILOG® STATEMENT $display("%d", 1’bx); $display("%h", 14’bx01010); $display("%h %o", 12’b001xxx101x01, 12’b001xxx101x01); RESULT x xxXa XXX 1x5X 17.1.1.5 Strength format The %v format specification is used to display the strength of scalar nets. For each %v specification that appears in a string, a corresponding scalar reference shall follow the string in the argument list. The strength of a scalar net is reported in a three-character format. The first two characters indicate the strength. The third character indicates the current logic value of the scalar and can be any one of the values given in Table 69. Table 69—Logic value component of strength format Argument Description 0 For a logic 0 value 1 For a logic 1 value X For an unknown value Z For a high impedance value L For a logic 0 or high impedance value H For a logic 1 or high impedance value The first two characters—the strength characters—are either a two-letter mnemonic or a pair of decimal digits. Usually, a mnemonic is used to indicate strength information; however, in less typical cases, a pair of decimal digits can be used to indicate a range of strength levels. Table 70 shows the mnemonics used to represent the various strength levels. Table 70—Mnemonics for strength levels Mnemonic Su St Pu La We Me Sm Hi Strength name Supply drive Strong drive Pull drive Large capacitor Weak drive Medium capacitor Small capacitor High impedance Strength level 7 6 5 4 3 2 1 0 284 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Note that there are four driving strengths and three charge storage strengths. The driving strengths are associated with gate outputs and continuous assignment outputs. The charge storage strengths are associated with the trireg type net. (See Clause 7 for strength modeling.) For the logic values 0 and 1, a mnemonic is used when there is no range of strengths in the signal. Otherwise, the logic value is preceded by two decimal digits, which indicate the maximum and minimum strength levels. For the unknown value, a mnemonic is used when both the 0 and 1 strength components are at the same strength level. Otherwise, the unknown value X is preceded by two decimal digits, which indicate the 0 and 1 strength levels respectively. The high impedance strength cannot have a known logic value; the only logic value allowed for this level is Z. For the values L and H, a mnemonic is always used to indicate the strength level. Examples: always #15 $display($time,,"group=%b signals=%v %v %v",{s1,s2,s3},s1,s2,s3); The example below shows the output that might result from such a call, while Table 71 explains the various strength formats that appear in the output. 0 group=111 signals=St1 Pu1 St1 15 group=011 signals=Pu0 Pu1 St1 30 group=0xz signals=520 PuH HiZ 45 group=0xx signals=Pu0 65X StX 60 group=000 signals=Me0 St0 St0 Table 71—Explanation of strength formats Argument Description St1 Means a strong driving 1 value Pu0 Means a pull driving 0 value HiZ Means the high-impedance state Me0 Means a 0 charge storage of medium capacitor strength StX Means a strong driving unknown value PuH Means a pull driving strength of 1 or high-impedance value 65X Means an unknown value with a strong driving 0 component and a pull driving 1 component 520 Means an 0 value with a range of possible strength from pull driving to medium capacitor Copyright © 2001 IEEE. All rights reserved. 285 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 17.1.1.6 Hierarchical name format The %m format specifier does not accept an argument. Instead, it causes the display task to print the hierarchical name of the module, task, function, or named block that invokes the system task containing the format specifier. This is useful when there are many instances of the module that calls the system task. One obvious application is timing check messages in a flip-flop or latch module; the %m format specifier shall pinpoint the module instance responsible for generating the timing check message. 17.1.1.7 String format The %s format specifier is used to print ASCII codes as characters. For each %s specification that appears in a string, a corresponding parameter shall follow the string in the argument list. The associated argument is interpreted as a sequence of 8-bit hexadecimal ASCII codes, with each 8 bits representing a single character. If the argument is a variable, its value should be right-justified so that the rightmost bit of the value is the least-significant bit of the last character in the string. No termination character or value is required at the end of a string, and leading zeros are never printed. 17.1.2 Strobed monitoring strobe_tasks ::= strobe_task_name ( list_of_arguments ) ; strobe_task_name ::= $strobe | $strobeb | $strobeo | $strobeh Syntax 17-2—Syntax for $strobe system tasks The system task $strobe provides the ability to display simulation data at a selected time. That time is the end of the current simulation time, when all the simulation events that have occurred for that simulation time, just before simulation time is advanced. The arguments for this task are specified in exactly the same manner as for the $display system task—including the use of escape sequences for special characters and format specifications (see 17.1.1). Example: forever @(negedge clock) $strobe ("At time %d, data is %h",$time,data); In this example, $strobe writes the time and data information to the standard output and the log file at each negative edge of the clock. The action shall occur just before simulation time is advanced and after all other events at that time have occurred, so that the data written is sure to be the correct data for that simulation time. 286 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE 17.1.3 Continuous monitoring IEEE Std 1364-2001 Version C monitor_tasks ::= monitor_task_name [ ( list_of_arguments ) ] ; | $monitoron ; | $monitoroff ; monitor_task_name ::= $monitor | $monitorb | $monitoro | $monitorh Syntax 17-3—Syntax for $monitor system tasks The $monitor task provides the ability to monitor and display the values of any variables or expressions specified as arguments to the task. The arguments for this task are specified in exactly the same manner as for the $display system task—including the use of escape sequences for special characters and format specifications (see 17.1.1). When a $monitor task is invoked with one or more arguments, the simulator sets up a mechanism whereby each time a variable or an expression in the argument list changes value—with the exception of the $time, $stime or $realtime system functions—the entire argument list is displayed at the end of the time step as if reported by the $display task. If two or more arguments change value at the same time, only one display is produced that shows the new values. Only one $monitor display list can be active at any one time; however, a new $monitor task with a new display list can be issued any number of times during simulation. The $monitoron and $monitoroff tasks control a monitor flag that enables and disables the monitoring. Use $monitoroff to turn off the flag and disable monitoring. The $monitoron system task can be used to turn on the flag so that monitoring is enabled and the most recent call to $monitor can resume its display. A call to $monitoron shall produce a display immediately after it is invoked, regardless of whether a value change has taken place; this is used to establish the initial values at the beginning of a monitoring session. By default, the monitor flag is turned on at the beginning of simulation. 17.2 File input-output system tasks and functions The system tasks and functions for file-based operations are divided into three categories: — Functions and tasks that open and close files — Tasks that output values into files — Tasks that output values into variables — Tasks and functions that read values from files and load into variables or memories 17.2.1 Opening and closing files file_open_function ::= integer multi_channel_descriptor = $fopen ( " file_name " ); | integer fd = $fopen ( " file_name ", type ); file_close_task ::= $fclose ( multi_channel_descriptor ); | $fclose (fd); Syntax 17-4—Syntax for $fopen and $fclose system tasks Copyright © 2001 IEEE. All rights reserved. 287 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® The function $fopen opens the file specified as the filename argument and returns either a 32 bit multi channel descriptor, or a 32 bit file descriptor, determined by the absence or presence of the type argument. filename is a character string, or a reg containing a character string that names the file to be opened. type is a character string, or a reg containing a character string of one of the following forms in the table below, which indicates how the file should be opened. If type is omitted, the file is opened for writing, and a multi channel descriptor mcd is returned. If type is supplied, the file is opened as specified by the value of type, and a file descriptor fd is returned. The multi channel descriptor mcd is a 32 bit reg in which a single bit is set indicating which file is opened. The least significant bit (bit 0) of a mcd always refers to the standard output. Output is directed to two or more files opened with multi channel descriptors by bitwise oring together their mcds and writing to the resultant value. The most significant bit (bit 32) of a multi channel descriptor is reserved, and shall always be cleared, limiting an implementation to at most 31 files opened for output via multi channel descriptors. The file descriptor fd is a 32 bit value. The most significant bit (bit 32) of a fd is reserved, and shall always be set; this allows implementations of the file input and output functions to determine how the file was opened. The remaining bits hold a small number indicating what file is opened. Three file descriptors are pre opened; they are STDIN, STDOUT and STDERR, which have the values 32'h8000_0000, 32'h8000_0001 and 32'h8000_0002, respectively. STDIN is pre opened for reading, and STDOUT and STDERR are pre opened for append. Unlike multi channel descriptors, file descriptors can not be combined via bitwise or in order to direct output to multiple files. Instead, files are opened via file descriptor for input, output, input and output, as well as for append operations, based on the value of type, according to the following table: Table 72—Types for file descriptors Argument Description "r" or "rb" "w" or "wb" "a" or "ab" "r+", "r+b", or "rb+" open for reading truncate to zero length or create for writing append; open for writing at end of file, or create for writing open for update (reading and writing) "w+", "w+b", or "wb+" truncate or create for update "a+", "a+b", or "ab+" append; open or create for update at end-of-file If a file can not be opened (either the file doesn't exist, and the type specified is "r", "rb", "r+", "r+b", or "rb+", or the permissions do not allow the file to be opened at that path, a zero is returned for either the mcd or the fd. Applications can call $ferror to determine the cause of the most recent error (see 17.2.7). The "b" in the above types exists to distinguish binary files from text files. Many systems (such as Unix) make no distinction between binary and text files, and on these systems the "b" is ignored. However, some systems (such as machines running NT or Windows) perform data mappings on certain binary values written to and read from files that are opened for text access. 288 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The $fclose system tasks closes the file specified by fd or closes the file(s) specified by the multi channel descriptor mcd. No further output to or input from any file descriptor(s) closed by $fclose is allowed. Active $fmonitor and/or $fstrobe operations on a file descriptor or multi channel descriptor are implicitly cancelled by an $fclose operation. The $fopen function shall reuse channels that have been closed. NOTE—The number of simultaneous input and output channels that can be open at any one time is dependent on the operating system. Some operating systems do not support opening files for update. 17.2.2 File output system tasks file_output_tasks ::= file_output_task_name ( multi_channel_descriptor , list_of_arguments ) ; | file_output_task_name ( fd , list_of_arguments ) ; file_output_task_name ::= $fdisplay | $fdisplayb | $fdisplayh | $fdisplayo | $fwrite | $fwriteb | $fwriteh | $fwriteo | $fstrobe | $fstrobeb | $fstrobeh | $fstrobeo | $fmonitor | $fmonitorb | $fmonitorh | $fmonitoro Syntax 17-5—Syntax for file output system tasks Each of the four formatted display tasks—$display, $write, $monitor, and $strobe—has a counterpart that writes to specific files as opposed to the standard output. These counterpart tasks—$fdisplay, $fwrite, $fmonitor, and $fstrobe—accept the same type of arguments as the tasks upon which they are based, with one exception: The first parameter shall be either a multi channel descriptor or a file descriptor, which indicates where to direct the file output. Multi channel descriptors are described in detail in 17.2.1. A multichannel descriptor is either a variable or the result of an expression that takes the form of a 32-bit unsigned integer value. The $fstrobe and $fmonitor system tasks work just like their counterparts, $strobe and $monitor, except that they write to files using the multi channel descriptor for control. Unlike $monitor, any number of $fmonitor tasks can be set up to be simultaneously active. However, there is no counterpart to $monitoron and $monitoroff tasks. The task $fclose is used to cancel an active $fstrobe or $fmonitor task. Example: This example shows how to set up multi channel descriptors. In this example, three different channels are opened using the $fopen function. The three multi channel descriptors that are returned by the function are then combined in a bit-wise or operation and assigned to the integer variable messages. The messages variable can then be used as the first parameter in a file output task to direct output to all three channels at once. To create a descriptor that directs output to the standard output as well, the messages variable is a bit-wise logical or with the constant 1, which effectively enables channel 0. Copyright © 2001 IEEE. All rights reserved. 289 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® integer messages, broadcast, cpu_chann, alu_chann, mem_chann; initial begin cpu_chann = $fopen("cpu.dat"); if (cpu_chann == 0) $finish; alu_chann = $fopen("alu.dat"); if (alu_chann == 0) $finish; mem_chann = $fopen("mem.dat"); if (mem_chann == 0) $finish; messages = cpu_chann | alu_chann | mem_chann; // broadcast includes standard output broadcast = 1 | messages; end endmodule The following file output tasks show how the channels opened in the preceding example might be used: $fdisplay( broadcast, "system reset at time %d",$time ); $fdisplay( messages, "Error occurred on address bus", " at time %d, address = %h", $time, address ); forever @(posedge clock) $fdisplay( alu_chann, "acc= %h f=%h a=%h b=%h", acc, f, a, b ); 17.2.3 Formatting data to a string string_output_tasks ::= string_output_task_name ( output_reg, list_of_arguments ); string_output_task_name ::= $swrite | $swriteb | $swriteh | $swriteo variable_format_string_output_task ::= $sformat ( output_reg, format_string, list_of_arguments ); Syntax 17-6—Syntax for formatting data tasks The syntax for the string output system tasks is $swrite(output_reg, list_of_arguments); $sformat(output_reg, format_string, list_of_arguments); length = $sformat(output_reg, format_string, list_of_arguments); The $swrite family of tasks is based on the $fwrite family of tasks, and accept the same type of arguments as the tasks upon which they are based, with one exception: The first parameter to $swrite shall be a reg variable to which the resulting string shall be written, instead of a variable specifying the file to which to write the resulting string. The system task $sformat is similar to the system task $swrite, with one major difference. 290 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Unlike the display and write family of output system tasks, $sformat always interprets its second argument, and only its second argument as a format string. This format argument can be a static string, such as ’"data is %d"’, or can be a reg variable whose content is interpreted as the format string. No other arguments are interpreted as format strings. $sformat supports all the format specifiers supported by $display, as documented in 17.1.1.2. The remaining arguments to $sformat are processed using any format specifiers in the format_string, until all such format specifiers are used up. If not enough arguments are supplied for the format specifiers, or too many are supplied, then the application shall issue a warning, and continue execution. The application, if possible, can statically determine a mismatch in format specifiers and number of arguments, and issue a compile time error message. NOTE—If the format_string is a reg, it might not be possible to determine its value at compile time. The variable output_reg is assigned using the string assignment to variable rules, as specified in 4.2.3. 17.2.4 Reading data from a file Files opened using file descriptors can be read from only if they were opened with either the r or r+ type values. See 17.2.1 for more information about opening files. 17.2.4.1 Reading a character at a time c = $fgetc ( fd ); Read a byte from the file specified by fd. If an error occurs reading from the file, then c is set to EOF (-1). Define the width of the reg to be wider than 8 bits so that a return value from $fgetc of EOF (-1) can be differentiated from the character code 0xFF. Applications can call $ferror to determine the cause of the most recent error (see 17.2.7). code = $ungetc ( c, fd ); Insert the character specified by c into the buffer specified by file descriptor fd. The character c shall be returned by the next $fgetc call on that file descriptor. The file itself is unchanged. Note that the features of the underlying implementation of fileio on the host system limits the number of characters that can be pushed back onto a stream. Note also that operations like $fseek might erase any pushed back characters. If an error occurs pushing a character onto a file descriptor, then code is set to EOF. Otherwise code is set to zero. Applications can call $ferror to determine the cause of the most recent error (see 17.2.7). 17.2.4.2 Reading a line at a time integer code = $fgets ( str, fd ); Read characters from the file specified by fd into the reg str until either str is filled, or a newline character is read and transferred to str, or an end-of-file condition is encountered. If str is not an integral number of bytes in length, the most significant partial byte is not used in order to determine the size. If an error occurs reading from the file, then code is set to zero. Otherwise the number of characters read is returned in code. Applications can call $ferror to determine the cause of the most recent error (see 17.2.7). 17.2.4.3 Reading formatted data integer code = $fscanf ( fd, format, args ); integer code = $sscanf ( str, format, args ); $fscanf reads from the files specified by the file descriptor fd. Copyright © 2001 IEEE. All rights reserved. 291 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® $sscanf reads from the reg str. Both functions read characters, interpret them according to a format, and store the results. Both expect as arguments a control string, format, and a set of arguments specifying where to place the results. If there are insufficient arguments for the format, the behavior is undefined. If the format is exhausted while arguments remain, the excess arguments are ignored. If an argument is too small to hold the converted input, then in general, the least significant bits are transferred. Arguments of any length that is supported by Verilog can be used. However if the destination is a real or realtime then the value +Inf (or -Inf) is transferred. The format can be a string constant or a reg contain- ing a string constant. The string contains conversion specifications, which direct the conversion of input into the arguments. The control string can contain a) White-space characters (blanks, tabs, new-lines, or form-feeds) that, except in one case described below, cause input to be read up to the next non-white-space character. b) An ordinary character (not %) that must match the next character of the input stream. c) Conversion specifications consisting of the character %, an optional assignment suppression character *, a decimal digit string that specifies an optional numerical maximum field width, and a conversion code. A conversion specification directs the conversion of the next input field; the result is placed in the variable specified in the corresponding argument unless assignment suppression was indicated by the character *; in this case no argument shall be supplied. The suppression of assignment provides a way of describing an input field that is to be skipped. An input field is defined as a string of non-space characters; it extends to the next inappropriate character or until the maximum field width, if one is specified, is exhausted. For all descriptors except the character c, white space leading an input field is ignored. % A single % is expected in the input at this point; no assignment is done. b Matches a binary number, consisting of a sequence from the set 0,1,X,x,Z,z,? and _. o Matches a octal number, consisting of a sequence of characters from the set 0,1,2,3,4,5,6,7,X,x,Z,z,? and _. d Matches an optionally signed decimal number, consisting of the optional sign from the set + or -, followed by a sequence of characters from the set 0,1,2,3,4,5,6,7,8,9 and _, or a single value from the set x,X,z,Z,?. h or x Matches a hexadecimal number, consisting of a sequence of characters from the set 0,1,2,3,4,5,6,7,8,9,a,A,b,B,c,C,d,D,e,E,f,F,x,X,z,Z,? and _. f, e or g Matches a floating point number. The format of a floating point number is an optional sign (either + or -), followed by a string of digits from the set 0,1,2,3,4,5,6,7,8,9 optionally containing a decimal point character (.), then an optional exponent part including e or E followed by an optional sign, followed by a string of digits from the set 0,1,2,3,4,5,6,7,8,9. v Matches a net signal strength, consisting of three character sequence as specified in 17.1.1.5. This conversion is not extremely useful, as strength values are really only usefully assigned to nets and $fscanf can only assign values to regs (if assigned to regs, the values are converted to the 4 value equivalent). t Matches a floating point number. The format of a floating point number is an optional sign (either + or -), followed by a string of digits from the set 0,1,2,3,4,5,6,7,8,9 optionally containing a decimal 292 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C point character (.), then an optional exponent part including e or E followed by an optional sign, followed by a string of digits from the set 0,1,2,3,4,5,6,7,8,9. The value matched is then scaled and rounded according to the current time scale as set by $timeformat. For example, if the timescale is `timescale 1ns/100ps and the time format is $timeformat(-3,2," ms",10);, then a value read with $sscanf("10.345", "%t", t) would return 10350000.0. c Matches a single character, whose 8 bit ASCII value is returned. s Matches a string, which is a sequence of non white space characters. u Matches unformatted (binary) data. The application shall transfer sufficient data from the input to fill the target reg. Typically the data is obtained from a matching $fwrite ("%u",data), or from an external application written in another programming language such as C, Perl or FORTRAN. The application shall transfer the 2 value binary data from the input stream to the destination reg, expanding the data to the four value format. This escape sequence can be used with any of the existing input system tasks, although $fscanf should be the preferred one to use. As the input data can not represent x or z, it is not possible to obtain an x or z in the result reg. This formatting specifier is intended to be used to support transferring data to and from external programs that have no concept of x and z. Applications that require preservation of x and z are encouraged to use the %z I/O format specification. The data shall be read from the file in the native endian format of the underlying system (i.e., in the same endian order as if the PLI was used, and the C language read(2) system call was used). For POSIX applications: It might be necessary to open files for unformatted I/O with the "rb", "rb+" or "r+b" specifiers, to avoid the systems implementation of I/O altering patterns in the unformatted stream that match special characters. z The formatting specification %z (or %Z) is defined for reading data without formatting (binary val- ues). The application shall transfer the 4 value binary representation of the specified data from the input stream to the destination reg. This escape sequence can be used with any of the existing input system tasks, although $fscanf should be the preferred one to use. This formatting specifier is intended to be used to support transferring data to and from external programs that recognize and support the concept of x and z. Applications that do not require the preservation of x and z are encouraged to use the %u I/O format specification. The data shall be read from the file in the native endian format of the underlying system (i.e., in the same endian order as if the PLI was used, and the data were in a s_vpi_vecval structure (See 27.14, Figure 179), and the C language read(2) system call was used to read the data from disk). For POSIX applications: It might be necessary to open files for unformatted I/O with the "rb", "rb+" or "r+b" specifiers, to avoid the systems implementation of I/O altering patterns in the unformatted stream that match special characters. m Returns the current hierarchical path as a string. Does not read data from the input file or str argu- ment. If an invalid conversion character follows the %, the results of the operation are implementation dependent. If the format string, or the str argument to $sscanf contains unknown bits (x or z) then the system task shall return EOF. Copyright © 2001 IEEE. All rights reserved. 293 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® If end-of-file is encountered during input, conversion is terminated. If end-of-file occurs before any characters matching the current directive have been read (other than leading white space, where permitted), execution of the current directive terminates with an input failure; otherwise, unless execution of the current directive is terminated with a matching failure, execution of the following directive (if any) is terminated with an input failure. If conversion terminates on a conflicting input character, the offending input character is left unread in the input stream. Trailing white space (including new-line characters) is left unread unless matched by a directive. The success of literal matches and suppressed assignments is not directly determinable. The number of successfully matched and assigned input items is returned in code; this number can be 0 in the event of an early matching failure between an input character and the control string. If the input ends before the first matching failure or conversion, EOF is returned. Applications can call $ferror to determine the cause of the most recent error (see 17.2.7). 17.2.4.4 Reading binary data integer code = $fread( myreg, fd); integer code = $fread( mem, fd); integer code = $fread( mem, fd, start); integer code = $fread( mem, fd, start, count); integer code = $fread( mem, fd, , count); Read a binary data from the file specified by fd into the reg myreg or the memory mem. start is an optional argument. If present, start shall be used as the address of the first element in the memory to be loaded. If not present the lowest numbered location in the memory shall be used. count is an optional argument. If present, count shall be the maximum number of locations in mem that shall be loaded. If not supplied the memory shall be filled with what data is available. start and count are ignored if $fread is loading a reg. If no addressing information is specified within the system task, and no address specifications appear within the data file, then the default start address is the lowest address given in the declaration of the memory. Consecutive words are loaded towards the highest address until either the memory is full or the data file is completely read. If the start address is specified in the task without the finish address, then loading starts at the specified start address and continues towards the highest address given in the declaration of the memory. start is the address in the memory. For start = 12 and the memory up[10:20], the first data would be loaded at up[12]. For the memory down[20:10], the first location loaded would be down[12], then down[13]. The data in the file shall be read byte by byte to fulfill the request. An 8-bit wide memory is loaded using one byte per memory word, while a 9-bit wide memory is loaded using 2 bytes per memory word. The data is read from the file in a big endian manner; the first byte read is used to fill the most significant location in the memory element. If the memory width is not evenly divisible by 8 (8, 16, 24, 32), not all data in the file is loaded into memory because of truncation. The data loaded from the file is taken as "two value" data. A bit set in the data is interpreted as a 1, and bit not set is interpreted as a 0. It is not possible to read a value of x or z using $fread. If an error occurs reading from the file, then code is set to zero. Otherwise the number of characters read is returned in code. Applications can call $ferror to determine the cause of the most recent error (see 17.2.7). Note that there is not a "binary" mode and a "ASCII" mode; one can freely intermingle binary and formatted read commands from the same file. 294 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 17.2.5 File positioning integer pos = $ftell ( fd ); Returns in pos the offset from the beginning of the file of the current byte of the file fd which shall be read or written by a subsequent operation on that file descriptor. This value can be used by subsequent $fseek calls to reposition the file to this point. Note that any repositioning shall cancel any $ungetc operations. If an error occurs EOF is returned. Applications can call $ferror to determine the cause of the most recent error (see 17.2.7). code = $fseek ( fd, offset, operation ); code = $rewind ( fd ); Sets the position of the next input or output operation on the file specified by fd. The new position is at the signed distance offset bytes from the beginning, from the current position, or from the end of the file, according to an operation value of 0, 1 and 2 as follows: — 0 set position equal to offset bytes — 1 set position to current location plus offset — 2 set position to EOF plus offset $rewind is equivalent to $fseek (fd,0,0); Repositioning the current file position with $fseek or $rewind shall cancel any $ungetc operations. $fseek() allows the file position indicator to be set beyond the end of the existing data in the file. If data is later written at this point, subsequent reads of data in the gap shall return zero until data is actually written into the gap. $fseek, by itself, does not extend the size of the file. When a file is opened for append (that is, when type is "a", or "a+"), it is impossible to overwrite information already in the file. $fseek can be used to reposition the file pointer to any position in the file, but when output is written to the file, the current file pointer is disregarded. All output is written at the end of the file and causes the file pointer to be repositioned at the end of the output. If an error occurs repositioning the file, then code is set to -1. Otherwise code is set to 0. Applications can call $ferror to determine the cause of the most recent error (see 17.2.7). 17.2.6 Flushing output $fflush ( mcd ); $fflush ( fd ); $fflush ( ); Writes any buffered output to the file(s) specified by mcd, the file specified by fd or if $fflush is invoked with no arguments, writes any buffered output to all open files. 17.2.7 I/O error status Should any error be detected by one of the fileio routines, an error code is returned. Often this is sufficient for normal operation; (i.e., if the opening of a optional configuration file fails, the application typically would simply continue using default values.) However sometimes it is useful to obtain more information about the error for correct application operation. In this case the $ferror function can be used: integer errno = $ferror ( fd, str ); Copyright © 2001 IEEE. All rights reserved. 295 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® A string description of type of error encountered by the most recent file I/O operation is written into str which should be at least 640 bits wide. The integral value of the error code is returned in errno. If the most recent operation did not result in an error, then the value returned shall be zero, and the reg str shall be cleared. 17.2.8 Loading memory data from a file load_memory_tasks ::= $readmemb ( " file_name " , memory_name [ , start_addr [ , finish_addr ] ] ) ; | $readmemh ( " file_name " , memory_name [ , start_addr [ , finish_addr ] ] ) ; Syntax 17-7—Syntax for memory load system tasks Two system tasks—$readmemb and $readmemh—read and load data from a specified text file into a specified memory. Either task can be executed at any time during simulation. The text file to be read shall contain only the following: — White space (spaces, new lines, tabs, and form-feeds) — Comments (both types of comment are allowed) — Binary or hexadecimal numbers The numbers shall have neither the length nor the base format specified. For $readmemb, each number shall be binary. For $readmemh, the numbers shall be hexadecimal. The unknown value (x or X), the high impedance value (z or Z), and the underscore (_) can be used in specifying a number as in a Verilog HDL source description. White space and/or comments shall be used to separate the numbers. In the following discussion, the term “address” refers to an index into the array that models the memory. As the file is read, each number encountered is assigned to a successive word element of the memory. Addressing is controlled both by specifying start and/or finish addresses in the system task invocation and by specifying addresses in the data file. When addresses appear in the data file, the format is an “at” character (@) followed by a hexadecimal number as follows: @hh...h Both uppercase and lowercase digits are allowed in the number. No white space is allowed between the @ and the number. As many address specifications as needed within the data file can be used. When the system task encounters an address specification, it loads subsequent data starting at that memory address. If no addressing information is specified within the system task, and no address specifications appear within the data file, then the default start address shall be the left-hand address given in the declaration of the memory. Consecutive words shall be loaded until either the memory is full or the data file is completely read. If the start address is specified in the task without the finish address, then loading shall start at the specified start address and shall continue towards the right-hand address given in the declaration of the memory. If both start and finish addresses are specified as parameters to the task, then loading shall begin at the start address and shall continue toward the finish address, regardless of how the addresses are specified in the memory declaration. When addressing information is specified both in the system task and in the data file, the addresses in the data file shall be within the address range specified by the system task parameters; otherwise, an error message is issued and the load operation is terminated. 296 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C A warning message shall be issued if the number of data words in the file differs from the number of words in the range implied by the start through finish addresses. Example: reg [7:0] mem[1:256]; Given this declaration, each of the following statements load data into mem in a different manner: initial $readmemh("mem.data", mem); initial $readmemh("mem.data", mem, 16); initial $readmemh("mem.data", mem, 128, 1); The first statement loads up the memory at simulation time 0 starting at the memory address 1. The second statement begins loading at address 16 and continue on towards address 256. For the third and final statement, loading begins at address 128 and continue down towards address 1. In the third case, when loading is complete, a final check is performed to ensure that exactly 128 numbers are contained in the file. If the check fails, a warning message is issued. 17.2.9 Loading timing data from an SDF file The syntax for the $sdf_annotate system task is shown in Syntax 17-8. sdf_annotate_task ::= $sdf_annotate ("sdf_file" [ , [ module_instance ] [ , [ "config_file" ] [ , [ "log_file" ] [ , [ "mtm_spec" ] [ , [ "scale_factors" ] [ , [ "scale_type" ] ] ] ] ] ] ] ); Syntax 17-8—Syntax for $sdf_annotate system task The $sdf_annotate system task reads timing data from an SDF file into a specified region of the design. sdf_file is a character string, or a reg containing a character string naming the file to be opened. module_instance is an optional argument specifying the scope to which to annotate the information in the SDF file. The SDF annotator uses the hierarchy level of the specified instance for running the annotation. Array indices are permitted. If the module_instance not specified, the SDF Annotator uses the module containing the call to the $sdf_annotate system task as the module_instance for annotation. config_file is an optional character string argument providing the name of a configuration file. Information in this file can be used to provide detailed control over many aspects of annotation. log_file is an optional character string argument providing the name of the log file used during SDF annotation. Each individual annotation of timing data from the SDF file results in an entry in the log file. mtm_spec is an optional character string argument specifying which member of the min/typ/max triples shall be annotated. The legal values for this string are described in Table 73. This overrides any MTM_SPEC keywords in the configuration file. Copyright © 2001 IEEE. All rights reserved. 297 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Table 73—mtm spec argument Keyword Description MAXIMUM Annotate the maximum value MINIMUM Annotate the minimum value TOOL_CONTROL (default) TYPICAL Annotate the value as selected by the simulator Annotate the typical value scale_factors scale_type is an optional character string argument specifying the scale factors to be used while annotating timing values. For example, "1.6:1.4:1.2" causes minimum values to be multiplied by 1.6, typical values by 1.4, and maximum values by 1.2. The default values are 1.0:1.0:1.0. The scale_factors argument overrides any SCALE_FACTORS keywords in the configuration file. is an optional character string argument specifying how the scale factors should be applied to the min/typ/max triples. The legal values for this string are shown in Table 74. This overrides any SCALE_TYPE keywords in the configuration file. Table 74—scale type argument Keyword Description FROM_MAXIMUM FROM_MINIMUM FROM_MTM (default) FROM_TYPICAL Apply scale factors to maximum value Apply scale factors to minimum value Apply scale factors to min/typ/max values Apply scale factors to typical value 17.3 Timescale system tasks The following system tasks display and set timescale information: a) $printtimescale b) $timeformat 17.3.1 $printtimescale The $printtimescale system task displays the time unit and precision for a particular module. The syntax for the system task is shown in Syntax 17-9. 298 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C printtimescale_task ::= $printtimescale [ ( hierarchical_identifier ) ] ; Syntax 17-9—Syntax for $printtimescale This system task can be specified with or without an argument. — When no argument is specified, $printtimescale displays the time unit and precision of the module that is the current scope. — When an argument is specified, $printtimescale displays the time unit and precision of the module passed to it. The timescale information shall appear in the following format: Time scale of (module_name) is unit / precision Example: `timescale 1 ms / 1 us module a_dat; initial $printtimescale(b_dat.c1); endmodule `timescale 10 fs / 1 fs module b_dat; c_dat c1 (); endmodule `timescale 1 ns / 1 ns module c_dat; . . . endmodule In this example, module a_dat invokes the $printtimescale system task to display timescale information about another module c_dat, which is instantiated in module b_dat. The information about c_dat shall be displayed in the following format: Time scale of (b_dat.c1) is 1ns / 1ns 17.3.2 $timeformat The syntax for $timeformat system task is shown in Syntax 17-10. Copyright © 2001 IEEE. All rights reserved. 299 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® timeformat_task ::= $timeformat [ ( units_number , precision_number , suffix_string , minimum_field_width ) ] ; Syntax 17-10—Syntax for $timeformat The $timeformat system task performs the following two functions: — It specifies how the %t format specification reports time information for the $write, $display, $strobe, $monitor, $fwrite, $fdisplay, $fstrobe, and $fmonitor group of system tasks. — It specifies the time unit for delays entered interactively. The units number argument shall be an integer in the range from 0 to -15. This argument represents the time unit as shown in Table 75. Table 75—$timeformat units_number arguments Unit number Time unit Unit number Time unit 0 1s -8 10 ns -1 100 ms -9 1 ns -2 10 ms -10 100 ps -3 1 ms -11 10 ps -4 100 us -12 1 ps -5 10 us -13 100 fs -6 1 us -14 10 fs -7 100 ns -15 1 fs The $timeformat system task performs the following two operations: — It sets the time unit for all later-entered delays entered interactively. — It sets the time unit, precision number, suffix string, and minimum field width for all %t formats specified in all modules that follow in the source description until another $timeformat system task is invoked. The default $timeformat system task arguments are given in Table 76. Table 76—$timeformat default value for arguments Argument Default units_number The smallest time precision argument of all the `timescale compiler directives in the source description precision_number 0 suffix_string A null character string minimum_field_width 20 300 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Example: The following example shows the use of %t with the $timeformat system task to specify a uniform time unit, time precision, and format for timing information. `timescale 1 ms / 1 ns module cntrl; initial $timeformat(-9, 5, " ns", 10); endmodule `timescale 1 fs / 1 fs module a1_dat; reg in1; integer file; buf #10000000 (o1,in1); initial begin file = $fopen("a1.dat"); #00000000 $fmonitor(file,"%m: %t in1=%d o1=%h", $realtime,in1,o1); #10000000 in1 = 0; #10000000 in1 = 1; end endmodule `timescale 1 ps / 1 ps module a2_dat; reg in2; integer file2; buf #10000 (o2,in2); initial begin file2=$fopen("a2.dat"); #00000 $fmonitor(file2,"%m: %t in2=%d o2=%h",$realtime,in2,o2); #10000 in2 = 0; #10000 in2 = 1; end endmodule The contents of file a1.dat are as follows: a1_dat: 0.00000 ns in1= x o1=x a1_dat: 10.00000 ns in1= 0 o1=x a1_dat: 20.00000 ns in1= 1 o1=0 a1_dat: 30.00000 ns in1= 1 o1=1 The contents of file a2.dat are as follows: a2_dat: 0.00000 ns in2=x o2=x a2_dat: 10.00000 ns in2=0 o2=x a2_dat: 20.00000 ns in2=1 o2=0 a2_dat: 30.00000 ns in2=1 o2=1 Copyright © 2001 IEEE. All rights reserved. 301 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® In this example, the times of events written to the files by the $fmonitor system task in modules a1_dat and a2_dat are reported as multiples of 1 ns—even though the time units for these modules are 1 fs and 1 ps respectively—because the first argument of the $timeformat system task is -9 and the %t format specification is included in the arguments to $fmonitor. This time information is reported after the module names with five fractional digits, followed by an ns character string in a space wide enough for 10 ASCII characters. 17.4 Simulation control system tasks There are two simulation control system tasks: a) $finish b) $stop 17.4.1 $finish Syntax 17-11 shows the syntax for $finish system task. finish_task ::= $finish [ ( n ) ] ; Syntax 17-11—Syntax for $finish The $finish system task simply makes the simulator exit and pass control back to the host operating system. If an expression is supplied to this task, then its value (0, 1, or 2) determines the diagnostic messages that are printed before the prompt is issued. If no argument is supplied, then a value of 1 is taken as the default. Table 77—Diagnostics for $finish Parameter value 0 1 2 Diagnostic message Prints nothing Prints simulation time and location Prints simulation time, location, and statistics about the memory and CPU time used in simulation 17.4.2 $stop The syntax for $stop system task is shown in Syntax 17-12. stop_task ::= $stop [ ( n ) ] ; Syntax 17-12—Syntax for $stop 302 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The $stop system task causes simulation to be suspended. This task takes an optional expression argument (0, 1, or 2) that determines what type of diagnostic message is printed. The amount of diagnostic messages output increases with the value of the optional argument passed to $stop. 17.5 PLA modeling system tasks The modeling of PLA devices is provided in the Verilog HDL by a group of system tasks. This clause describes the syntax and use of these system tasks and the formats of the logic array personality file.The syntax for PLA modeling system task is shown in Syntax 17-13. pla_system_task ::= $array_type$logic$format ( memory_type , input_terms , output_terms ) ; array_type ::= sync | async logic ::= and | or | nand | nor format ::= array | plane input_terms ::= expression output_terms ::= variable_lvalue Syntax 17-13 —Syntax for PLA modeling system task NOTE—The input terms can be nets or variables whereas the output terms shall only be variables. The PLA syntax allows for the system tasks as shown in Table 78. Table 78—PLA modeling system tasks $async$and$array $sync$and$array $async$nand$array $sync$nand$array $async$or$array $sync$or$array $async$nor$array $sync$nor$array $async$and$plane $async$nand$plane $async$or$plane $async$nor$plane $sync$and$plane $sync$nand$plane $sync$or$plane $sync$nor$plane 17.5.1 Array types The modeling of both synchronous and asynchronous arrays is provided by the PLA system tasks. The synchronous forms control the time at which the logic array shall be evaluated and the outputs shall be updated. For the asynchronous forms, the evaluations are automatically performed whenever an input term changes value or any word in the personality memory is changed. For both the synchronous and asynchronous forms, the output terms are updated without any delay. Copyright © 2001 IEEE. All rights reserved. 303 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Examples: An example of an asynchronous system call is as follows: wire a1, a2, a3, a4, a5, a6, a7; reg b1, b2, b3; wire [1:7] awire; reg [1:3] breg; $async$and$array(mem,{a1,a2,a3,a4,a5,a6,a7},{b1,b2,b3}); or $async$and$array(mem,awire, breg); An example of a synchronous system call is as follows: $sync$or$plane(mem,{a1,a2,a3,a4,a5,a6,a7}, {b1,b2,b3}); 17.5.2 Array logic types The logic arrays are modeled with and, or, nand, and nor logic planes. This applies to all array types and formats. Examples: An example of a nor plane system call is as follows: $async$nor$plane(mem,{a1,a2,a3,a4,a5,a6,a7},{b1,b2,b3}); An example of a nand plane system call is as follows: $sync$nand$plane(mem,{a1,a2,a3,a4,a5,a6,a7}, {b1,b2,b3}); 17.5.3 Logic array personality declaration and loading The logic array personality is declared as an array of regs that is as wide as the number of input terms and as deep as the number of output terms. The personality of the logic array is normally loaded into the memory from a text data file using the system tasks $readmemb or $readmemh. Alternatively, the personality data can be written directly into the memory using the procedural assignment statements. PLA personalities can be changed dynamically at any time during simulation simply by changing the contents of the memory. The new personality shall be reflected on the outputs of the logic array at the next evaluation. Example: The following example shows a logic array with n input terms and m output terms. reg [1:n] mem[1:m]; NOTE—Put PLA input terms, output terms, and memory in ascending order, as shown in examples in this clause. 17.5.4 Logic array personality formats Two separate personality formats are supported by the Verilog HDL and are differentiated by using either an array system call or a plane system call. The array system call allows for a 1 or 0 in the memory that has been declared. A 1 means take the input value and a 0 means do not take the input value. 304 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The plane system call complies with the University of California at Berkeley format for Espresso. Each bit of the data stored in the array has the following meaning: 0 Take the complemented input value 1 Take the true input value x Take the “worst case” of the input value z Don’t-care; the input value is of no significance ? Same as z Examples: Example 1—The following example illustrates an array with logic equations: b1 = a1 & a2 b2 = a3 & a4 & a5 b3 = a5 & a6 & a7 The PLA personality is as follows: 1100000 in mem[1] 0011100 in mem[2] 0000111 in mem[3] The module for the PLA is as follows: module async_array(a1,a2,a3,a4,a5,a6,a7,b1,b2,b3); input a1, a2, a3, a4, a5, a6, a7 ; output b1, b2, b3; reg [1:7] mem[1:3]; // memory declaration for array personality reg b1, b2, b3; initial begin // setup the personality from the file array.dat $readmemb("array.dat", mem); // setup an asynchronous logic array with the input // and output terms expressed as concatenations $async$and$array(mem,{a1,a2,a3,a4,a5,a6,a7},{b1,b2,b3}); end endmodule Where the file array.dat contains the binary data for the PLA personality: 1100000 0011100 0000111 A synchronous version of this example has the following description: Copyright © 2001 IEEE. All rights reserved. 305 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® module sync_array(a1,a2,a3,a4,a5,a6,a7,b1,b2,b3,clk); input a1, a2, a3, a4, a5, a6, a7, clk; output b1, b2, b3; reg [1:7] mem[1:3]; // memory declaration reg b1, b2, b3; initial begin // setup the personality $readmemb("array.dat", mem); // setup a synchronous logic array to be evaluated // when a positive edge on the clock occurs forever @(posedge clk) $async$and$array(mem,{a1,a2,a3,a4,a5,a6,a7},{b1,b2,b3}); end endmodule Example 2—An example of the usage of the plane format tasks follows. The logical function of this PLA is shown first, followed by the PLA personality in the new format, the Verilog HDL description using the $async$and$plane system task, and finally the result of running the simulation. The logical function of the PLA is as follows: b[1] = a[1] & ~a[2]; b[2] = a[3]; b[3] = ~a[1] & ~a[3]; b[4] = 1; The PLA personality is as follows: 3’b10? 3’b??1 3’b0?0 3’b??? 306 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C module pla; `define rows 4 `define cols 3 reg [1:`cols] a, mem[1:`rows]; reg [1:`rows] b; initial begin // PLA system call $async$and$plane(mem,a[1:3],b[1:4]); mem[1] = 3’b10?; mem[2] = 3’b??1; mem[3] = 3’b0?0; mem[4] = 3’b???; // stimulus and display #10 a = 3’b111; #10 $displayb(a, " -> ", b); #10 a = 3’b000; #10 $displayb(a, " -> ", b); #10 a = 3’bxxx; #10 $displayb(a, " -> ", b); #10 a = 3’b101; #10 $displayb(a, " -> ", b); end endmodule The output is as follows: 111 -> 0101 000 -> 0011 xxx -> xxx1 101 -> 1101 17.6 Stochastic analysis tasks This clause describes a set of system tasks and functions that manage queues and generate random numbers with specific distributions. These tasks facilitate implementation of stochastic queueing models. The set of tasks and functions that create and manage queues follow: $q_initialize (q_id, q_type, max_length, status) ; $q_add (q_id, job_id, inform_id, status) ; $q_remove (q_id, job_id, inform_id, status) ; $q_full (q_id, status) ; $q_exam (q_id, q_stat_code, q_stat_value, status) ; 17.6.1 $q_initialize The $q_initialize system task creates new queues. The q_id parameter is an integer input that shall uniquely identify the new queue. The q_type parameter is an integer input. The value of the q_type parameter specifies the type of the queue as shown in Table 79. Copyright © 2001 IEEE. All rights reserved. 307 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Table 79—Types of queues of $q_type values q_type value Type of queue 1 first-in, first-out 2 last-in, first-out The maximum length parameter is an integer input that specifies the maximum number of entries allowed on the queue. The success or failure of the creation of the queue is returned as an integer value in status. The error conditions and corresponding values of status are described in Table 81. 17.6.2 $q_add The $q_add system task places an entry on a queue. The q_id parameter is an integer input that indicates to which queue to add the entry. The job_id parameter is an integer input that identifies the job. The inform_id parameter is an integer input that is associated with the queue entry. Its meaning is userdefined. For example, inform_id parameter can represent execution time for an entry in a CPU model. The status parameter reports on the success of the operation or error conditions as described in Table 81. 17.6.3 $q_remove The $q_remove system task receives an entry from a queue. The q_id parameter is an integer input that indicates from which queue to remove. The job_id parameter is an integer output that identifies the entry being removed. The inform_id parameter is an integer output that the queue manager stored during $q_add. Its meaning is user-defined. The status parameter reports on the success of the operation or error conditions as described in Table 81. 17.6.4 $q_full The $q_full system function checks whether there is room for another entry on a queue. It returns 0 when the queue is not full and 1 when the queue is full. 17.6.5 $q_exam The $q_exam system task provides statistical information about activity at the queue q_id. It returns a value in q_stat_value depending on the information requested in q_stat_code. The values of q_stat_code and the corresponding information returned in q_stat_value are described in Table 80. Table 80—Parameter values for $q_exam system task Value requested in q_stat_code Information received back from q_stat_value 1 Current queue length 2 Mean interarrival time 3 Maximum queue length 4 Shortest wait time ever 5 Longest wait time for jobs still in the queue 6 Average wait time in the queue 308 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 17.6.6 Status codes All of the queue management tasks and functions return an output status parameter. The status parameter values and corresponding information are described in Table 81. Table 81—Status parameter values Status parameter values What it means 0 OK 1 Queue full, cannot add 2 Undefined q_id 3 Queue empty, cannot remove 4 Unsupported queue type, cannot create queue 5 Specified length <= 0, cannot create queue 6 Duplicate q_id, cannot create queue 7 Not enough memory, cannot create queue 17.7 Simulation time system functions The following system functions provide access to current simulation time: $time $stime $realtime 17.7.1 $time The syntax for $time system function is shown in Syntax 17-14. time_function ::= $time Syntax 17-14—Syntax for $time The $time system function returns an integer that is a 64-bit time, scaled to the timescale unit of the module that invoked it. Copyright © 2001 IEEE. All rights reserved. 309 IEEE Std 1364-2001 Version C Example: IEEE STANDARD VERILOG® `timescale 10 ns / 1 ns module test; reg set; parameter p = 1.55; initial begin $monitor($time,,"set=",set); #p set = 0; #p set = 1; end endmodule // The output from this example is as follows: // 0 set=x // 2 set=0 // 3 set=1 In this example, the reg set is assigned the value 0 at simulation time 16 ns, and the value 1 at simulation time 32 ns. Note that these times do not match the times reported by $time. The time values returned by the $time system function are determined by the following steps: a) The simulation times 16ns and 32 ns are scaled to 1.6 and 3.2 because the time unit for the module is 10 ns, so time values reported by this module are multiples of 10 ns. b) The value 1.6 is rounded to 2, and 3.2 is rounded to 3 because the $time system function returns an integer. The time precision does not cause rounding of these values. 17.7.2 $stime The syntax for $stime system function is shown in Syntax 17-15. stime_function ::= $stime Syntax 17-15—Syntax for $stime The $stime system function returns an unsigned integer that is a 32-bit time, scaled to the timescale unit of the module that invoked it. If the actual simulation time does not fit in 32 bits, the low order 32 bits of the current simulation time are returned. 17.7.3 $realtime The syntax for $realtime system function is shown in Syntax 17-16. realtime_function ::= $realtime Syntax 17-16—Syntax for $realtime 310 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The $realtime system function returns a real number time that, like $time, is scaled to the time unit of the module that invoked it. Example: `timescale 10 ns / 1 ns module test; reg set; parameter p = 1.55; initial begin $monitor($realtime,,"set=",set); #p set = 0; #p set = 1; end endmodule // The output from this example is as follows: // 0 set=x // 1.6 set=0 // 3.2 set=1 In this example, the event times in the reg set are multiples of 10 ns because 10 ns is the time unit of the module. They are real numbers because $realtime returns a real number. 17.8 Conversion functions The following functions handle real values: integer real [63:0] real $rtoi(real_val) ; $itor(int_val) ; $realtobits(real_val) ; $bitstoreal(bit_val) ; $rtoi converts real values to integers by truncating the real value (for example, 123.45 becomes 123) $itor converts integers to real values (for example, 123 becomes 123.0) $realtobits passes bit patterns across module ports; converts from a real number to the 64-bit representation (vector) of that real number $bitstoreal is the reverse of $realtobits; converts from the bit pattern to a real number. The real numbers accepted or generated by these functions shall conform to the IEEE Std 754-1985 [B1] representation of the real number. The conversion shall round the result to the nearest valid representation. Example: The following example shows how the $realtobits and $bitstoreal functions are used in port connections: Copyright © 2001 IEEE. All rights reserved. 311 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® module driver (net_r); output net_r; real r; wire [64:1] net_r = $realtobits(r); endmodule module receiver (net_r); input net_r; wire [64:1] net_r; real r; initial assign r = $bitstoreal(net_r); endmodule See 4.5 for a description of $signed and $unsigned. 17.9 Probabilistic distribution functions There are a set of random number generators that return integer values distributed according to standard probabilistic functions. 17.9.1 $random function The syntax for the system function $random is shown in Syntax 17-17. random_function ::= $random [ ( seed ) ] ; Syntax 17-17—Syntax for $random The system function $random provides a mechanism for generating random numbers. The function returns a new 32-bit random number each time it is called. The random number is a signed integer; it can be positive or negative. For further information on probabilistic random number generators, see 17.9.2. The seed parameter controls the numbers that $random returns such that different seeds generate different random streams. The seed parameter shall be either a reg, an integer, or a time variable. The seed value should be assigned to this variable prior to calling $random. Examples: Example 1—Where b is greater than 0, the expression ($random % b) gives a number in the following range: [(-b+1): (b-1)]. The following code fragment shows an example of random number generation between -59 and 59: reg [23:0] rand; rand = $random % 60; 312 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Example 2—The following example shows how adding the concatenation operator to the preceding example gives rand a positive value from 0 to 59. reg [23:0] rand; rand = {$random} % 60; 17.9.2 $dist_ functions dist_functions ::= $dist_uniform ( seed , start , end ) ; | $dist_normal ( seed , mean , standard_deviation ) ; | $dist_exponential ( seed , mean ) ; | $dist_poisson ( seed , mean ) ; | $dist_chi_square ( seed , degree_of_freedom ) ; | $dist_t ( seed , degree_of_freedom ) ; | $dist_erlang ( seed , k_stage , mean ) ; Syntax 17-18—Syntax for the probabilistic distribution functions All parameters to the system functions are integer values. For the exponential, poisson, chisquare, t, and erlang functions, the parameters mean, degree of freedom, and k_stage shall be greater than 0. Each of these functions returns a pseudo-random number whose characteristics are described by the function name. That is, $dist_uniform returns random numbers uniformly distributed in the interval specified by its parameters. For each system function, the seed parameter is an in-out parameter; that is, a value is passed to the function and a different value is returned. The system functions shall always return the same value given the same seed. This facilitates debugging by making the operation of the system repeatable. The argument for the seed parameter should be an integer variable that is initialized by the user and only updated by the system function. This ensures the desired distribution is achieved. In the $dist_uniform function, the start and end parameters are integer inputs that bound the values returned. The start value should be smaller than the end value. The mean parameter, used by $dist_normal, $dist_exponential, $dist_poisson, and $dist_erlang, is an integer input that causes the average value returned by the function to approach the value specified. The standard deviation parameter used with the $dist_normal function is an integer input that helps determine the shape of the density function. Larger numbers for standard deviation spread the returned values over a wider range. The degree of freedom parameter used with the $dist_chi_square and $dist_t functions is an integer input that helps determine the shape of the density function. Larger numbers spread the returned values over a wider range. Copyright © 2001 IEEE. All rights reserved. 313 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 17.9.3 Algorithm for probabilistic distribution functions Table 82 shows the Verilog probabilistic distribution functions listed with their corresponding C functions. Table 82—Verilog to C function cross-listing Verilog function $dist_uniform $dist_normal $dist_exponential $dist_poisson $dist_chi_square $dist_t $dist_erlang $random C function rtl_dist_uniform rtl_dist_normal rtl_dist_exponential rtl_dist_poisson rtl_dist_chi_square rtl_dist_t rtl_dist_erlang rtl_dist_uniform (seed, LONG_MIN, LONG_MAX) The algorithm for these functions is defined by the following C code. /* * Algorithm for probabilistic distribution functions. * * IEEE Std 1364-2001 Verilog Hardware Description Language (HDL). */ #include static double uniform( long *seed, long start, long end ); static double normal( long *seed, long mean, long deviation); static double exponential( long *seed, long mean); static long poisson( long *seed, long mean); static double chi_square( long *seed, long deg_of_free); static double t( long *seed, long deg_of_free); static double erlangian( long *seed, long k, long mean); long rtl_dist_chi_square( seed, df ) long *seed; long df; { double r; long i; if(df>0) { r=chi_square(seed,df); if(r>=0) { i=(long)(r+0.5); } 314 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C else { r = -r; i=(long)(r+0.5); i = -i; } } else { print_error("WARNING: Chi_square distribution must ", "have positive degree of freedom\n"); i=0; } return (i); } long rtl_dist_erlang( seed, k, mean ) long *seed; long k, mean; { double r; long i; if(k>0) { r=erlangian(seed,k,mean); if(r>=0) { i=(long)(r+0.5); } else { r = -r; i=(long)(r+0.5); i = -i; } } else { print_error("WARNING: k-stage erlangian distribution ", "must have positive k\n"); i=0; } return (i); } long rtl_dist_exponential( seed, mean ) long *seed; long mean; { double r; Copyright © 2001 IEEE. All rights reserved. 315 IEEE Std 1364-2001 Version C long i; if(mean>0) { r=exponential(seed,mean); if(r>=0) IEEE STANDARD VERILOG® { i=(long)(r+0.5); } else { r = -r; i=(long)(r+0.5); i = -i; } } else { print_error("WARNING: Exponential distribution must ", "have a positive mean\n"); i=0; } return (i); } long rtl_dist_normal( seed, mean, sd ) long *seed; long mean, sd; { double r; long i; r=normal(seed,mean,sd); if(r>=0) { i=(long)(r+0.5); } else { r = -r; i=(long)(r+0.5); i = -i; } return (i); } long rtl_dist_poisson( seed, mean ) long *seed; long mean; 316 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C { long i; if(mean>0) { i=poisson(seed,mean); } else { print_error("WARNING: Poisson distribution must have a ", "positive mean\n"); i=0; } return (i); } long rtl_dist_t( seed, df ) long *seed; long df; { double r; long i; if(df>0) { r=t(seed,df); if(r>=0) { i=(long)(r+0.5); } else { r = -r; i=(long)(r+0.5); i = -i; } } else { print_error("WARNING: t distribution must have positive ", "degree of freedom\n"); i=0; } return (i); } long rtl_dist_uniform(seed, start, end) long *seed; long start, end; { double r; long i; if (start >= end) return(start); Copyright © 2001 IEEE. All rights reserved. 317 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® if (end != LONG_MAX) { end++; r = uniform( seed, start, end ); if (r >= 0) { i = (long) r; } else { i = (long) (r-1); } if (i=end) i = end-1; } else if (start!=LONG_MIN) { start--; r = uniform( seed, start, end) + 1.0; if (r>=0) { i = (long) r; } else { i = (long) (r-1); } if (i<=start) i = start+1; if (i>end) i = end; } else { r =(uniform(seed,start,end)+ 2147483648.0)/4294967295.0); r = r*4294967296.0-2147483648.0; if (r>=0) { i = (long) r; } else { i = (long) (r-1); } } return (i); } static double uniform( seed, start, end ) long *seed, start, end; { union u_s 318 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE { } u; float s; unsigned stemp; double d = 0.00000011920928955078125; double a,b,c; IEEE Std 1364-2001 Version C if ((*seed) == 0) *seed = 259341593; if (start >= end) { a = 0.0; b = 2147483647.0; } else { a = (double) start; b = (double) end; } *seed = 69069 * (*seed) + 1; u.stemp = *seed; /* * This relies on IEEE floating point format */ u.stemp = (u.stemp >> 9) | 0x3f800000; c = (double) u.s; c = c+(c*d); c = ((b - a) * (c - 1.0)) + a; return (c); } static double normal(seed,mean,deviation) long *seed,mean,deviation; { double v1,v2,s; double log(), sqrt(); s = 1.0; while((s >= 1.0) || (s == 0.0)) { v1 = uniform(seed,-1,1); v2 = uniform(seed,-1,1); s = v1 * v1 + v2 * v2; } s = v1 * sqrt(-2.0 * log(s) / s); v1 = (double) deviation; v2 = (double) mean; Copyright © 2001 IEEE. All rights reserved. 319 IEEE Std 1364-2001 Version C return(s * v1 + v2); } static double exponential(seed,mean) long *seed,mean; { double log(),n; n = uniform(seed,0,1); if(n != 0) { n = -log(n) * mean; } return(n); } static long poisson(seed,mean) long *seed,mean; { long n; double p,q; double exp(); n = 0; q = -(double)mean; p = exp(q); q = uniform(seed,0,1); while(p < q) { n++; q = uniform(seed,0,1) * q; } return(n); } static double chi_square(seed,deg_of_free) long *seed,deg_of_free; { double x; long k; if(deg_of_free % 2) { x = normal(seed,0,1); x = x * x; } else { x = 0.0; } double log(),n; n = uniform(seed,0,1); if(n != 0) { 320 IEEE STANDARD VERILOG® Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE n = -log(n) * mean; } return(n); } static double t(seed,deg_of_free) long *seed,deg_of_free; { double sqrt(),x; double chi2 = chi_square(seed,deg_of_free); double div = chi2 / (double)deg_of_free; double root = sqrt(div); x = normal(seed,0,1) / root; return(x); } static double erlangian(seed,k,mean) long *seed,k,mean; { double x,log(),a,b; long i; x=1.0; for(i=1;i<=k;i++) IEEE Std 1364-2001 Version C { x = x * uniform(seed,0,1); } a=(double)mean; b=(double)k; x= -a*log(x)/b; return(x); } 17.10 Command line input An alternative to reading a file to obtain information for use in the simulation is specifying information with the command to invoke the simulator. This information is in the form of a optional argument provided to the simulation. These arguments are visually distinguished from other simulator arguments by their starting with the plus (+) character. These arguments, referred to below as plusargs, are accessible through the system functions described in 17.10.1 and 17.10.2. Copyright © 2001 IEEE. All rights reserved. 321 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 17.10.1 $test$plusargs (string) This system function searches the list of plusargs for the provided string. The plusargs present on the command line are searched in the order provided. If the prefix of one of the supplied plusargs matches all characters in the provided string, a non-zero integer is returned. If no plusarg from the command line matches the string provided, the integer value zero (0) is returned. Examples: Run simulator with command: +HELLO The Verilog code is: initial begin if ($test$plusargs("HELLO")) $display("Hello argument found.") if ($test$plusargs("HE")) $display("The HE subset string is detected."); if ($test$plusargs("H")) $display("Argument starting with H found."); if ($test$plusargs("HELLO_HERE"))$display("Long argument."); if ($test$plusargs("HI")) $display("Simple greeting."); if ($test$plusargs("LO")) $display("Does not match."); end This would produce the following output: Hello argument found. The HE subset string is detected. Argument starting with H found. 17.10.2 $value$plusargs (user_string, variable) This system function searches the list of plusargs (like the $test$plusargs system function) for a user specified plusarg string. The string is specified in the first argument to the system function as either a string or a register which is interpreted as a string. If the string is found, the remainder of the string is converted to the type specified in the user_string and the resulting value stored in the variable provided. If a string is found, the function returns a non-zero integer. If no string is found matching, the function returns the integer value zero and the variable provided is not modified. No warnings shall be generated when the function returns zero (0). The user_string shall be of the form: "plusarg_string format_string". The format strings are the same as the $display system tasks. These are the only valid ones (upper and lower case as well as leading 0 forms are valid): %d decimal conversion %o octal conversion %h hexadecimal conversion %b binary conversion %e real exponential conversion %f real decimal conversion %g real decimal or exponential conversion %s string (no conversion) 322 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The first string, from the list of plusargs provided to the simulator, which matches the plusarg_string portion of the user_string specified shall be the plusarg string available for conversion. The remainder string of the matching plusarg (the remainder is the part of the plusarg string after the portion which matches the user’s plusarg_string) shall be converted from a string into the format indicated by the format string and stored in the variable provided. If there is no remaining string, the value stored into the variable shall either be a zero (0) or an empty string value. If the size of the variable is larger than the value after conversion, the value stored is zero-padded to the width of the variable. If the variable can not contain the value after conversion, the value shall be truncated. If the value is negative, the value shall be considered larger than the variable provided. If characters exist in the string available for conversion, which are illegal for the specified conversion, the variable shall be written with the value ’bx. Examples: +FINISH=10000 +TESTNAME=this_test +FREQ+5.6666 +FREQUENCY +TEST12 // Get clock to terminate simulation if specified. real frequency; reg [8*32:1] testname; integer stop_clock; if ($value$plusargs("FINISH=%d", stop_clock)) begin repeat (stop_clock) @(posedge clk); $finish; end // Get testname from plusarg. if ($value$plusargs("TESTNAME=%s", testname)) begin $display("Running test %0s.", testname); startTest(); end // Get frequency from command line; set default if not specified. if (!$value$plusargs("FREQ+%0F", frequency)) frequency = 8.33333; // 166MHz; forever begin #frequency clk = 0; #frequency clk = 1; end reg [64*8:1] pstring; pstring = "+TEST%d"; if ($value$plusargs(pstring, test[31:0)) begin $display("Running test number %0d.", test); startTest(); end This code would have the following effects: — The variable test would get the value ’d12. — The variable stop_clock obtains the value 10000. — The variable testname obtains the value this_test. Copyright © 2001 IEEE. All rights reserved. 323 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® — The variable frequency obtains the value 5.6666; note the final plusarg +FREQUENCY does not affect the value of the variable frequency. The output is: Running test this_test. Running test number 12. 324 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 18. Value change dump (VCD) files A value change dump (VCD) file contains information about value changes on selected variables in the design stored by value change dump system tasks. Two types of VCD files exist: a) Four state: to represent variable changes in 0, 1, x, and z with no strength information. b) Extended: to represent variable changes in all states and strength information. This clause describes how to generate both types of VCD files and their format. 18.1 Creating the four state value change dump file The steps involved in creating the four state VCD file are listed below and illustrated in Figure 51. a) Insert the VCD system tasks in the Verilog source file to define the dump file name and to specify the variables to be dumped. b) Run the simulation. Verilog Source File initial $dumpfile(“dump1.dump”); . . . $dumpvars(...) . . . Four State VCD File dump1.dump simulation (Header Information) (Node Information) User Postprocessing (Value Changes) Figure 51—Creating the four state VCD file A VCD file is an ASCII file which contains header information, variable definitions, and the value changes for all variables specified in the task calls. Several system tasks can be inserted in the source description to create and control the VCD file. 18.1.1 Specifying the name of the dump file ($dumpfile) The $dumpfile task shall be used to specify the name of the VCD file. The syntax for the task is given in Syntax 18-1. Copyright © 2001 IEEE. All rights reserved. 325 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® dumpfile_task ::= $dumpfile ( filename ) ; Syntax 18-1—Syntax for $dumpfile task The filename syntax is given in Syntax 18-2. filename ::= literal_string | variable | expression Syntax 18-2—Syntax for filename The filename is optional and defaults to the literal string dump.vcd if not specified. Example: initial $dumpfile ("module1.dump") ; 18.1.2 Specifying the variables to be dumped ($dumpvars) The $dumpvars task shall be used to list which variables to dump into the file specified by $dumpfile. The $dumpvars task can be invoked as often as desired throughout the model (for example, within various blocks), but the execution of all the $dumpvars tasks shall be at the same simulation time. The $dumpvars task can be used with or without arguments. The syntax for the $dumpvars task is given in Syntax 18-3. dumpvars_task ::= (Not in the Annex A BNF) $dumpvars ; | $dumpvars ( levels [ , list_of_modules_or_variables ] ) ; list_of_modules_or_variables ::= (Not in the Annex A BNF) module_or_variable { , module_or_variable } module_or_variable ::= module_identifier | variable_identifier Syntax 18-3—Syntax for $dumpvars task When invoked with no arguments, $dumpvars dumps all the variables in the model to the VCD file. 326 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C When the $dumpvars task is specified with arguments, the first argument indicates how many levels of the hierarchy below each specified module instance to dump to the VCD file. Subsequent arguments specify which scopes of the model to dump to the VCD file. These arguments can specify entire modules or individual variables within a module. Setting the first argument to 0 causes a dump of all variables in the specified module and in all module instances below the specified module. The argument 0 applies only to subsequent arguments which specify module instances, and not to individual variables. Examples: Example 1 $dumpvars (1, top); Because the first argument is a 1, this invocation dumps all variables within the module top; it does not dump variables in any of the modules instantiated by module top. Example 2 $dumpvars (0, top); In this example, the $dumpvars task shall dump all variables in the module top and in all module instances below module top in the hierarchy. Example 3—This example shows how the $dumpvars task can specify both modules and individual variables: $dumpvars (0, top.mod1, top.mod2.net1); This call shall dump all variables in module mod1 and in all module instances below mod1, along with variable net1 in module mod2. The argument 0 applies only to the module instance top.mod1 and not to the individual variable top.mod2.net1. 18.1.3 Stopping and resuming the dump ($dumpoff/$dumpon) Executing the $dumpvars task causes the value change dumping to start at the end of the current simulation time unit. To suspend the dump, the $dumpoff task can be invoked. To resume the dump, the $dumpon task can be invoked. The syntax of these two tasks is given in Syntax 18-4. dumpoff_task ::= $dumpoff ; dumpon_task ::= $dumpon ; Syntax 18-4—Syntax for $dumpoff and $dumpon tasks When the $dumpoff task is executed, a checkpoint is made in which every selected variable is dumped as an x value. When the $dumpon task is later executed, each variable is dumped with its value at that time. In the interval between $dumpoff and $dumpon, no value changes are dumped. Copyright © 2001 IEEE. All rights reserved. 327 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® The $dumpoff and $dumpon tasks provide the mechanism to control the simulation period during which the dump shall take place. Example: initial begin #10 $dumpvars( . . . ); #200 $dumpoff; #800 $dumpon; #900 $dumpoff; end This example starts the value change dump after 10 time units, stops it 200 time units later (at time 210), restarts it again 800 time units later (at time 1010), and stops it 900 time units later (at time 1910). 18.1.4 Generating a checkpoint ($dumpall) The $dumpall task creates a checkpoint in the VCD file which shows the current value of all selected variables. The syntax is given in Syntax 18-5. dumpall_task ::= $dumpall ; Syntax 18-5—Syntax for $dumpall task When dumping is enabled, the value change dumper records the values of the variables which change during each time increment. Values of variables which do not change during a time increment are not dumped. 18.1.5 Limiting the size of the dump file ($dumplimit) The $dumplimit task can be used to set the size of the VCD file. The syntax for this task is given in Syntax 18-6. dumplimit_task ::= $dumplimit ( filesize ) ; Syntax 18-6—Syntax fro $dumplimit task The filesize argument which specifies the maximum size of the VCD file in bytes. When the size of VCD file reaches this number of bytes, the dumping stops and a comment is inserted in the VCD file indicating the dump limit was reached. 328 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 18.1.6 Reading the dump file during simulation ($dumpflush) The $dumpflush task can be used to empty the VCD file buffer of the operating system to ensure all the data in that buffer is stored in the VCD file. After executing a $dumpflush task, dumping is resumed as before so no value changes are lost. The syntax for the task is given in Syntax 18-7. dumpflush_task ::= $dumpflush ; Syntax 18-7—Syntax for $dumpflush task A common application is to call $dumpflush to update the dump file so an application program can read the VCD file during a simulation. Examples: Example 1—This example shows how the $dumpflush task can be used in a Verilog HDL source file: initial begin $dumpvars ; . . . $dumpflush ; $(applications program) ; end Example 2—The following is a simple source description example to produce a VCD file. In this example, the name of the dump file is verilog.dump. It dumps value changes for all variables in the model. Dumping begins when an event do_dump occurs. The dumping continues for 500 clock cycles, then stops and waits for the event do_dump to be triggered again. At every 10000 time steps, the current values of all VCD variables are dumped. Copyright © 2001 IEEE. All rights reserved. 329 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® module dump; event do_dump; initial $dumpfile("verilog.dump"); initial @do_dump $dumpvars; //dump variables in the design always @do_dump //to begin the dump at event do_dump begin $dumpon; //no effect the first time through repeat (500) @(posedge clock); //dump for 500 cycles $dumpoff; //stop the dump end initial @(do_dump) forever #10000 $dumpall; //checkpoint all variables endmodule 18.2 Format of the four state VCD file The dump file is structured in a free format. White space is used to separate commands and to make the file easily readable by a text editor. 18.2.1 Syntax of the four state VCD file The syntax of the four state VCD file is given in Syntax 18-8. 330 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C value_change_dump_definitions ::= { declaration_command }{ simulation_command } declaration_command ::= declaration_keyword [ command_text ] $end simulation_command ::= simulation_keyword { value_change } $end | $comment [ comment_text ] $end | simulation_time | value_change declaration_keyword ::= $comment | $date | $enddefinitions | $scope | $timescale | $upscope | $var | $version simulation_keyword ::= $dumpall | $dumpoff | $dumpon | $dumpvars simulation_time ::= # decimal_number value_change ::= scalar_value_change | vector_value_change scalar_value_change ::= value identifier_code value ::= 0|1|x|X|z|Z vector_value_change ::= b binary_number identifier_code | B binary_number identifier_code | r real_number identifier_code | R real_number identifier_code identifier_code ::= { ASCII character } Syntax 18-8—Syntax of the output four state VCD file The VCD file starts with header information giving the date, the version number of the simulator used for the simulation, and the timescale used. Next, the file contains definitions of the scope and type of variables being dumped, followed by the actual value changes at each simulation time increment. Only the variables which change value during a time increment are listed. The simulation time recorded in VCD file is the absolute value of the simulation time for the changes in variable values which follow. Value changes for real variables are specified by real numbers.Value changes for all other variables are specified in binary format by 0, 1, x, or z values. Strength information and memories are not dumped. A real number is dumped using a %.16g printf() format. This preserves the precision of that number by outputting all 53 bits in the mantissa of a 64-bit IEEE Std 754-1985 [B1] double-precision number. Application programs can read a real number using a %g format to scanf(). Copyright © 2001 IEEE. All rights reserved. 331 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® The value change dumper generates character identifier codes to represent variables. The identifier code is a code composed of the printable characters which are in the ASCII character set from ! to ~ (decimal 33 to 126). NOTES: 1) The VCD format does not support a mechanism to dump part of a vector. For example, bits 8 to 15 (8:15) of a 16-bit vector cannot be dumped in VCD file; instead, the entire vector (0:15) has to be dumped. In addition, expressions, such as a + b, cannot be dumped in the VCD file. 2) Data in the VCD file is case sensitive. 18.2.2 Formats of variable values Variables can be either scalars or vectors. Each type is dumped in its own format. Dumps of value changes to scalar variables shall not have any white space between the value and the identifier code. Dumps of value changes to vectors shall not have any white space between the base letter and the value digits, but they shall have one white space between the value digits and the identifier code. The output format for each value is right-justified. Vector values appear in the shortest form possible: redundant bit values which result from left-extending values to fill a particular vector size are eliminated. The rules for left-extending vector values are given in Table 83. Table 83—Rules for left-extending vector values When the value is VCD left-extends with 1 0 0 0 Z Z X X Table 84 shows how the VCD can shorten values. Table 84—How the VCD can shorten values The binary value Extends to fill a 4-bit reg as Appears in the VCD file as 10 0010 b10 X10 XX10 bX10 ZX0 ZZX0 bZX0 0X10 0X10 b0X10 Events are dumped in the same format as scalars; for example, 1*%. For events, however, the value (1 in this example) is irrelevant. Only the identifier code (*% in this example) is significant. It appears in the VCD file as a marker to indicate the event was triggered during the time step. 332 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Examples: 1*@ No space between the value 1 and the identifier code *@ b1100x01z (k No space between the b and 1100x01z, but a space between b1100x01z and (k 18.2.3 Description of keyword commands The general information in the VCD file is presented as a series of sections surrounded by keywords. Keyword commands provide a means of inserting information in the VCD file. Keyword commands can be inserted either by the dumper or manually. This sub clause deals with the keyword commands given in Table 85. Table 85—Keyword commands Declaration keywords $comment $timescale $date $upscope $enddefinitions $scope $var $version Simulation keywords $dumpall $dumpoff $dumpon $dumpvars 18.2.3.1 $comment The $comment section provides a means of inserting a comment in the VCD file. The syntax for the section is given in Syntax 18-9. vcd_declaration_comment ::= $comment comment_text $end Syntax 18-9—Syntax for $comment section Examples: $comment This is a single-line comment $comment This is a multiple-line comment $end $end 18.2.3.2 $date The $date section indicates the date on which the VCD file was generated.The syntax for the section is given in Syntax 18-10. Copyright © 2001 IEEE. All rights reserved. 333 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® vcd_declaration_date ::= $date date_text $end Syntax 18-10—Syntax for $date section Example: $date June 25, 1989 09:24:35 $end 18.2.3.3 $enddefinitions The $enddefinitions section marks the end of the header information and definitions.The syntax for the section is given in Syntax 18-11. . vcd_declaration_enddefinitions ::= $enddefinitions $end Syntax 18-11—Syntax for $enddefinitions section 18.2.3.4 $scope The $scope section defines the scope of the variables being dumped.The syntax for the section is given in Syntax 18-12. vcd_declaration_scope ::= $scope scope_type scope_identifier $end scope_type ::= begin | fork | function | module | task Syntax 18-12—Syntax for $scope section The scope type indicates one of the following scopes: module task function begin fork Top-level module and module instances Tasks Functions Named sequential blocks Named parallel blocks 334 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Example: $scope module top $end 18.2.3.5 $timescale The $timescale keyword specifies what timescale was used for the simulation.The syntax for the keyword is given in Syntax 18-13. vcd_declaration_timescale ::= $timescale time_number time_unit $end time_number ::= 1 | 10 | 100 time_unit ::= s | ms | us | ns | ps | fs Syntax 18-13—Syntax for $timescale Example: $timescale 10 ns $end 18.2.3.6 $upscope The $upscope section indicates a change of scope to the next higher level in the design hierarchy. The syntax for the section is given in Syntax 18-14. vcd_declaration_upscope ::= $upscope $end Syntax 18-14—Syntax for $upscope section 18.2.3.7 $version The $version section indicates which version of the VCD writer was used to produce the VCD file and the $dumpfile system task used to create the file. If a variable or an expression was used to specify the filename within $dumpfile, the unevaluated variable or expression literal shall appear in the $version string. The syntax for the $version section is given in Syntax 18-15. vcd_declaration_version ::= $version version_text system_task $end Syntax 18-15—Syntax for $version section Copyright © 2001 IEEE. All rights reserved. 335 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Example: $version VERILOG-SIMULATOR 1.0a $dumpfile(“dump1.dump”) $end 18.2.3.8 $var The $var section prints the names and identifier codes of the variables being dumped. The syntax for the section is given in Syntax 18-16. vcd_declaration_vars ::= $var var_type size identifier_code reference $end var_type ::= event | integer | parameter | real | reg | supply0 | supply1 | time | tri | triand | trior | trireg | tri0 | tri1 | wand | wire | wor size ::= decimal_number reference ::= identifier | identifier [ bit_select_index ] | identifier [ msb_index : lsb_index ] index ::= decimal_number Syntax 18-16—Syntax for $var section Size specifies how many bits are in the variable. The identifier code specifies the name of the variable using printable ASCII characters, as previously described. a) The msb index indicates the most significant index; the lsb index indicates the least significant index. b) More than one reference name can be mapped to the same identifier code. For example, net10 and net15 can be interconnected in the circuit and therefore have the same identifier code. c) The individual bits of vector nets can be dumped individually. d) The identifier is the name of the variable being dumped in the model. Example: $var integer 32 (2 index $end 336 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 18.2.3.9 $dumpall The $dumpall keyword specifies current values of all variables dumped. The syntax for the keyword is given in Syntax 18-17. vcd_simulation_dumpall ::= $dumpall { value_changes } $end Syntax 18-17—Syntax for $dumpall keyword Example: $dumpall 1*@ x*# 0*$ bx (k $end 18.2.3.10 $dumpoff The $dumpoff keyword indicates all variables dumped with X values. The syntax for the keyword is given in Syntax 18-18. vcd_simulation_dumpoff ::= $dumpoff { value_changes } $end Syntax 18-18—Syntax for $dumpoff keyword Example: $dumpoff x*@ x*# x*$ bx (k $end 18.2.3.11 $dumpon The $dumpon keyword indicates resumption of dumping and lists current values of all variables dumped. The syntax for the keyword is given in Syntax 18-19. vcd_simulation_dumpon ::= $dumpon { value_changes } $end Syntax 18-19—Syntax for $dumpon keyword Example: $dumpon x*@ 0*# x*$ b1 (k $end Copyright © 2001 IEEE. All rights reserved. 337 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 18.2.3.12 $dumpvars The section beginning with $dumpvars keyword lists initial values of all variables dumped. The syntax for the keyword is given in Syntax 18-20. vcd_simulation_dumpvars ::= $dumpvars { value_changes } $end Syntax 18-20—Syntax for $dumpvars keyword Example: $dumpvars x*@ z*$ b0 (k $end 338 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 18.2.4 Four state VCD file format example The following example illustrates the format of the four state VCD file. $date June 26, 1989 10:05:41 $end $version VERILOG-SIMULATOR 1.0a $end $timescale 1 ns $end $scope module top $end $scope module m1 $end $var trireg 1 *@ net1 $end $var trireg 1 *# net2 $end $var trireg 1 *$ net3 $end $upscope $end $scope task t1 $end $var reg 32 (k accumulator[31:0] $end $var integer 32 {2 index $end $upscope $end $upscope $end $enddefinitions $end $comment Note: $dumpvars was executed at time ’#500’. All initial values are dumped at this time. $end #500 $dumpvars x*@ x*# x*$ bx (k bx {2 $end #505 0*@ 1*# 1*$ b10zx1110x11100 (k b1111000101z01x {2 #510 0*$ #520 1*$ #530 0*$ bz (k #535 $dumpall 0*@ 1*# 0*$ ( Continued in right column ) (Continued from left column) bz (k b1111000101z01x {2 $end #540 1*$ #1000 $dumpoff x*@ x*# x*$ bx (k bx {2 $end #2000 $dumpon z*@ 1*# 0*$ b0 (k bx {2 $end #2010 1*$ Copyright © 2001 IEEE. All rights reserved. 339 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 18.3 Creating the extended value change dump file The steps involved in creating the extended VCD file are listed below and illustrated in Figure 52. Verilog Source File initial $dumpports(“dump2.dump”); . . . . . . Extended VCD File dump2.dump simulation (Header Information) (Node Information) User Postprocessing (Value Changes) Figure 52—Creating the extended VCD file a) Insert the extended VCD system tasks in the Verilog source file to define the dump file name and to specify the variables to be dumped. b) Run the simulation. The four state VCD file rules and syntax apply to the extended VCD file unless otherwise stated in this section. 18.3.1 Specifying the dumpfile name and the ports to be dumped ($dumpports) The $dumpports task shall be used to specify the name of the VCD file and the ports to be dumped. The syntax for the task is given in Syntax 18-21. dumpports_task ::= $dumpports ( scope_list , file_pathname ) ; scope_list ::= module_identifier { , module_identfier } file_pathname ::= literal_string | variable | expression Syntax 18-21—Syntax for $dumpports task Where the arguments are optional and are defined as: scope_list one or more module identifiers. Only modules are allowed (not variables). If more than one module_identifier is specified, they shall be separated by a comma. Pathnames to modules are allowed, using the period hierarchy separator. Literal strings are not allowed for the module_identifier. 340 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C file_pathname If no scope_list value is provided, the scope shall be the module from which $dumpports is called. can be a double quoted pathname (literal string), a reg type variable, or an expression which denotes the file which shall contain the port VCD information. If no file_pathname is provided, the file shall be written to the current working directory with the name dumpports.vcd. If that file already exists, it shall be silently overwritten. All file writing checks shall be made by the simulator (write rights, correct pathname, etc.) and appropriate errors or warnings issued. The following rules apply to the use of the $dumpports system task: — All the ports in the model from the point of the $dumpports call are considered primary I/O pins and shall be included in the VCD file. However, any ports which exist in instantiations below scope_list are not dumped. — If no arguments are specified for the task, $dumpports; and $dumpports(); are allowed. In both of these cases, the default values for the arguments shall be used. — If the first argument is null, a comma shall be used before specifying the second argument in the argument list. — Each scope specified in the scope_list shall be unique. If multiple calls to $dumpports are specified, the scope_list values in these calls shall also be unique. — The $dumpports task can be used in source code which also contains the $dumpvars task. — When $dumpports executes, the associated value change dumping shall start at the end of the cur- rent simulation time unit. — The $dumpports task can be invoked multiple times throughout the model, but the execution of all $dumpports tasks shall be at the same simulation time. Specifying the same file_pathname multiple times is not allowed. 18.3.2 Stopping and resuming the dump ($dumpportsoff/$dumpportson) The $dumpportsoff and $dumpportson system tasks provide a means to control the simulation period for dumping port values. The syntax for these system tasks is given in Syntax 18-22. dumpportsoff_task ::= $dumpportsoff ( file_pathname ) ; dumpportson_task ::= $dumpportson ( file_pathname ) ; file_pathname ::= literal_string | variable | expression Syntax 18-22—Syntax for $dumpportsoff and $dumpportson system tasks The file_pathname argument can be a double quoted pathname (literal string), a reg type variable, or an expression which denotes the file_pathname specified in the $dumpports system task. $dumpportsoff. When this task is executed, a checkpoint is made in the file_pathname where each specified port is dumped with an X value. Port values are no longer dumped from that simulation time forward. If file_pathname is not specified, all dumping to files opened by $dumpports calls shall be suspended. Copyright © 2001 IEEE. All rights reserved. 341 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® $dumpportson. When this task is executed, all ports specified by the associated $dumpports call shall have their values dumped. This system task is typically used to resume dumping after the execution of $dumpportsoff. If file_pathname is not specified, dumping shall resume for all files specified by $dumpports calls, if dumping to those files was stopped. If $dumpportson is executed while ports are already being dumped to file_pathname, the system task is ignored. If $dumpportsoff is executed while port dumping is already suspended for file_pathname, the system task is ignored. 18.3.3 Generating a checkpoint ($dumpportsall) The $dumpportsall system task creates a checkpoint in the VCD file which shows the value of all selected ports at that time in the simulation, regardless of whether the port values have changed since the last timestep. The syntax for this system task is given in Syntax 18-23. dumpportsall_task ::= $dumpportsall ( file_pathname ) ; file_pathname ::= literal_string | variable | expression Syntax 18-23—Syntax for $dumpportsall system task The file_pathname argument can be a double quoted pathname (literal string), a reg type variable, or an expression which denotes the file_pathname specified in the $dumpports system task. If the file_pathname is not specified, checkpointing occurs for all files opened by calls to $dumpports. 18.3.4 Limiting the size of the dump file ($dumpportslimit) The $dumpportslimit system task allows control of the VCD file size. The syntax for this system task is given in Syntax 18-24. dumpportslimit_task ::= $dumpportslimit ( filesize , file_pathname ) ; file_size ::= integer file_pathname ::= literal_string | variable | expression Syntax 18-24—Syntax for $dumpportslimit system task 342 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The filesize argument is required and it specifies the maximum size in bytes for the associated file_pathname. When this filesize is reached, the dumping stops and a comment is inserted into file_pathname indicating the size limit was attained. The file_pathname argument can be a double quoted pathname (literal string), a reg type variable, or an expression which denotes the file_pathname specified in the $dumpports system task. If the file_pathname is not specified, the filesize limit applies to all files opened for dumping due to calls to $dumpports. 18.3.5 Reading the dump file during simulation ($dumpportsflush) To facilitate performance, simulators often buffer VCD output and write to the file at intervals, instead of line by line. The $dumpportsflush system task writes all port values to the associated file, clearing a simulator’s VCD buffer. The syntax for this system task is given in Syntax 18-25. dumpportsflush_task ::= $dumpportsflush ( file_pathname ) ; file_pathname ::= literal_string | variable | expression Syntax 18-25—Syntax for $dumpportsflush system task The file_pathname argument can be a double quoted pathname (literal string), a reg type variable, or an expression which denotes the file_pathname specified in the $dumpports system task. If the file_pathname is not specified, the VCD buffers shall be flushed for all files opened by calls to $dumpports. 18.3.6 Description of keyword commands The general information in the extended VCD file is presented as a series of sections surrounded by keywords. Keyword commands provide a means of inserting information in the extended VCD file. Keyword commands can be inserted either by the dumper or manually. Extended VCD provides one additional keyword command to that of the four state VCD. 18.3.6.1 $vcdclose The $vcdclose keyword indicates the final simulation time at the time the extended VCD file is closed. This allows accurate recording of the end simulation time, regardless of the state of signal changes, in order to assist parsers which require this information. The syntax for the keyword is given in Syntax 18-26. Copyright © 2001 IEEE. All rights reserved. 343 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® vcdclose_task ::= $vcdclose final_simulation_time $end Syntax 18-26—Syntax for $vcdclose keyword Example: $vcdclose #13000 $end 18.3.7 General rules for extended VCD system tasks For each extended VCD system task, the following rules apply: — If a file_pathname is specified which does not match a file_pathname specified in a $dumpports call, the control task shall be ignored. — If no arguments are specified for the tasks which have only optional arguments, the system task name can be used with no arguments or the name followed by () can be specified. For example: $dumpportsflush; or $dumpportsflush(). In both of these cases, the default actions for the arguments shall be executed. 18.4 Format of the extended VCD file The format of the extended VCD file is similar to that of the four state VCD file, as it is also structured in a free format. White space is used to separate commands and to make the file easily readable by a text editor. 18.4.1 Syntax of the extended VCD file The syntax of the extended VCD file is given in Syntax 18-27. A four state VCD construct name which matches an extended VCD construct shall be considered equivalent, except if preceded by an *. 344 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C value_change_dump_definitions ::={declaration_command} {simulation_command} declaration_command ::= declaration_keyword [command_text] $end simulation_command ::= (Not in the Annex A BNF) simulation_keyword { value_change } $end | $comment [comment_text] $end | simulation_time | value_change * declaration_keyword ::= $comment | $date | $enddefinitions | $scope | $timescale | $upscope | $var | $vcdclose | $version command_text ::= comment_text | close_text | date_section | scope_section | timescale_section | var_section | version_section * simulation_keyword ::= $dumpports | $dumpportsoff | $dumpportson | $dumpportsall simulation_time ::= #decimal_number value_change ::= value identifier_code value ::= pport_value 0_strength_component 1_strength_component port_value ::= input_value | output_value | unknown_direction_value input_value ::= D | U | N | Z | d | u output_value ::= L | H | X | T | l | h unknown_direction_value ::= 0 | 1 | ? | F | A | a | B | b | C | c | f strength_component ::= 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 * identifier_code ::= <{integer} comment_text ::= {ASCII_character} close_text ::= final_simulation_time date_section ::= date_text date_text :: = day month date time year scope_section ::= scope_type scope_identifier * scope_type ::= module timescale_section ::= number time_unit number ::= 1 | 10 | 100 time_unit ::= fs | ps | ns | us | ms | s var_section ::= var_type size identifier_code reference * var_type ::= port * size ::= 1 | vector_index vector_index ::= [ msb_index : lsb_index ] index ::= decimal_number * reference ::= port_identifier identifier ::= {printable_ASCII_character} version_section ::= version_text * version_text ::= version_identifier {dumpports_command} dumpports_command ::= $dumpports (scope_identifier , string_literal | variable | expression ) Syntax 18-27—Syntax of the output extended VCD file Copyright © 2001 IEEE. All rights reserved. 345 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® The extended VCD file starts with header information giving the date, the version number of the simulator used for the simulation, and the timescale used. Next, the file contains definitions of the scope of the ports being dumped, followed by the actual value changes at each simulation time increment. Only the ports which change value during a time increment are listed. The simulation time recorded in the extended VCD file is the absolute value of the simulation time for the changes in port values which follow. Value changes for all ports are specified in binary format by 0, 1, x, or z values and include strength information. A real number is dumped using a %.16g printf() format. This preserves the precision of that number by outputting all 53 bits in the mantissa of a 64-bit IEEE Std 754-1985 [B1] double-precision number. Application programs can read a real number using a %g format to scanf(). NOTES: 1) The extended VCD format does not support a mechanism to dump part of a vector. For example, bits 8 to 15 (8:15) of a 16-bit vector cannot be dumped in VCD file; instead, the entire vector (0:15) has to be dumped. In addition, expressions, such as a + b, cannot be dumped in the VCD file. 2) Data in the extended VCD file is case sensitive. 18.4.2 Extended VCD node information The node information section (also referred to as the variable definitions section) is affected by the $dumpports task as Syntax 18-28 shows. $var var_type size < identifier_code reference $end var_type ::= port size ::= 1 | vector_index vector_index ::= [msb_index : lsb_index] index ::= decimal_number identifier_code ::= integer reference ::= port_identifier Syntax 18-28—Syntax of extended VCD node information 346 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The constructs are defined as: var_type the keyword port. No other keyword is allowed. size a decimal number indicating the number of bits in the port. If the port is a single bit, the value shall be 1. If the port is a bus, the actual index is printed. The msb indicates the most significant index; lsb the least significant index. identifier_code an integer preceded by < which starts at zero and ascends in one unit increments for each port, in the order found in the module declaration. reference identifier indicating the port name. Example: module test_device(count_out, carry, data, reset) output count_out, carry ; input [0:3] data; input reset; ... initial begin $dumpports(testbench.DUT, "testoutput.vcd"); ... end This example produces the following node information in the VCD file: $scope module testbench.DUT $end $var port 1 <0 $var port 1 <1 $var port [0:3] <2 $var port 1 <3 $upscope $end count_out carry data reset $end $end $end $end At least one space shall separate each syntactical element. However, the formatting of the information is the choice of the simulator vendor. All four state VCD syntax rules for the vector_index apply. If the vector_index appears in the port declaration, this shall be the index dumped. If the vector_index is not in the port declaration, the vector_index in the net or reg declaration matching the port name shall be dumped. If no vector_index is found, the port is considered scalar (1 bit wide). Concatenated ports shall appear in the extended VCD file as separate entries. Example: module addbit ({A, b}, ci, sum, co); input A, b, ci; output sum, co; ... Copyright © 2001 IEEE. All rights reserved. 347 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® The VCD file output looks like: $scope module addbit $end $var port 1 <0 A $end $var port 1 <1 b $end $var port 1 <2 ci $end $enddefinitions $end ... 18.4.3 Value changes The value change section of the VCD file is also affected by $dumpports, as Syntax 18-29 shows. pport_value 0_strength_component 1_strength_component identifier_code Syntax 18-29—Syntax of value change section Where the constructs are defined as: p port_value 0_strength_component 1_strength_component key character which indicates a port. There is no space between the p and the port_value. state character (described below). one of the 8 Verilog strengths which indicates the strength0 specification for the port. one of the 8 Verilog strengths which indicates the strength1 specification for the port. The Verilog strength values are (append keyword with 0 or 1 as appropriate for the strength component): 0 1 2 3 4 5 6 7 identifier_code highz small medium weak large pull strong supply the integer preceded by the < character as defined in the $var construct for the port. 18.4.3.1 State characters The following state information is listed in terms of input values from a test fixture, the output values of the device under test (DUT), and the states representing unknown direction: INPUT (TESTFIXTURE) D low U high N unknown Z three-state d low (two or more drivers active) u high (two or more drivers active) 348 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C OUTPUT (DUT) L low H high X unknown (don't care) T three-state l low (two or more drivers active) h high (two or more drivers active) UNKNOWN DIRECTION 0 low (both input and output are active with 0 value) 1 high (both input and output are active with 1 value) ? unknown F three-state (input and output unconnected) A unknown (input 0 and output 1) a unknown (input 0 and output X) B unknown (input 1 and output 0) b unknown (input 1 and output X) C unknown (input X and output 0) c unknown (input X and output 1) f unknown (input and output three-stated) 18.4.3.2 Drivers Where drivers are considered only in terms of primitives, continuous assignments, and procedural continuous assignments. Value 0/1 means both input and output are active with value 0/1. 0 and 1 are conflict states. The following rules apply to conflicts: — If both input and output are driving the same value with the same range of strength, then this is a conflict. The resolved value is 0/1 and the strength is the stronger of the two. — If the input is driving a strong strength (range) and the output is driving a weak strength (range), the resolved value is d/u and the strength is the strength of the input. — If the input is driving a weak strength (range) and the output is driving a strong strength (range), then the resolved value is l/h and the strength is the strength of the output. Where range is: — Strength supply 7 to 5 (large) - strong strength — Strength 4 to 1 - weak strength 18.4.4 Extended VCD file format example The following example illustrates the format of the extended VCD file. A module declaration: module adder(data0, data1, data2, data3, carry, as, rdn, reset, test, write); inout data0, data1, data2, data3; output carry; input as, rdn, reset, test, write; ... Copyright © 2001 IEEE. All rights reserved. 349 IEEE Std 1364-2001 Version C And the resulting VCD fragment: $scope module testbench.adder_instance $end $var port 1 <0 data0 $end $var port 1 <1 data1 $end $var port 1 <2 data2 $end $var port 1 <3 data3 $end $var port 1 <4 carry $end $var port 1 <5 as $end $var port 1 <6 rdn $end $var port 1 <7 reset $end $var port 1 <8 test $end $var port 1 <9 write $end $upscope $end $enddefinitions $end #0 $dumpports pX 6 6 <0 pX 6 6 <1 pX 6 6 <2 pX 6 6 <3 pX 6 6 <4 pN 6 6 <5 pN 6 6 <6 pU 0 6 <7 pD 6 0 <8 pN 6 6 <9 $end #180 pH 0 6 <4 #200000 pD 6 0 <5 pU 0 6 <6 pD 6 0 <9 #200500 pf 0 0 <0 pf 0 0 <1 pf 0 0 <2 pf 0 0 <3 IEEE STANDARD VERILOG® 350 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 19. Compiler directives All Verilog compiler directives are preceded by the (` ) character. This character is called accent grave. It is different from the character (’), which is the single quote character. The scope of compiler directives extends from the point where it is processed, across all files processed, to the point where another compiler directive supersedes it or the processing completes. This clause describes the following compiler directives: `celldefine `default_nettype `define `else `elsif `endcelldefine `endif `ifdef `ifndef `include `line `nounconnected_drive `resetall `timescale `unconnected_drive `undef [19.1] [19.2] [19.3] [19.4] [19.4] [19.1] [19.4] [19.4] [19.4] [19.5] [19.7] [19.9] [19.6] [19.8] [19.9] [19.3] 19.1 `celldefine and `endcelldefine The directives `celldefine and `endcelldefine tag modules as cell modules. Cells are used by certain PLI routines for applications, such as delay calculations. It is advisable to pair each `celldefine with an `endcelldefine. More than one of these pairs may appear in a single source description. These directives may appear anywhere in the source description, but it is recommended that the directives be specified outside the module definition. The `resetall directive includes the effects of a `endcelldefine directive. 19.2 `default_nettype The directive `default_nettype controls the net type created for implicit net declarations (see 3.5). It can be used only outside of module definitions. It affects all modules that follow the directive, even across source file boundaries. Multiple `default_nettype directives are allowed. The latest occurrence of this directive in the source controls the type of nets that will be implicitly declared. Syntax 19-1 contains the syntax of the directive. default_nettype_compiler_directive ::= `default_nettype net_type net_type ::= wire | tri | tri0 | wand | triand | wor | trior | trireg | none Syntax 19-1—Syntax for default nettype compiler directive Copyright © 2001 IEEE. All rights reserved. 351 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® When no `default_nettype directive is present or if the `resetall directive is specified, implicit nets are of type wire. When the `default_nettype is set to none, all nets must be explicitly declared. If a net is not explicitly declared, an error is generated. 19.3 `define and `undef A text macro substitution facility has been provided so that meaningful names can be used to represent commonly used pieces of text. For example, in the situation where a constant number is repetitively used throughout a description, a text macro would be useful in that only one place in the source description would need to be altered if the value of the constant needed to be changed. The text macro facility is not affected by the compiler directive `resetall. 19.3.1 `define The directive `define creates a macro for text substitution. This directive can be used both inside and outside module definitions. After a text macro is defined, it can be used in the source description by using the (`) character, followed by the macro name. The compiler shall substitute the text of the macro for the string `macro_name. All compiler directives shall be considered predefined macro names; it shall be illegal to redefine a compiler directive as a macro name. A text macro can be defined with arguments. This allows the macro to be customized for each use individually. The syntax for text macro definitions is given in Syntax 19-2. text_macro_definition ::= `define text_macro_name macro_text text_macro_name ::= text_macro_identifier [ ( list_of_formal_arguments ) ] list_of_formal_arguments ::= formal_argument_identifier { , formal_argument_identifier } text_macro_identifier ::= (From Annex A - A.9.3) simple_identifier Syntax 19-2—Syntax for text macro definition The macro text can be any arbitrary text specified on the same line as the text macro name. If more than one line is necessary to specify the text, the newline shall be preceded by a backslash (\). The first newline not preceded by a backslash shall end the macro text. The newline preceded by a backslash shall be replaced in the expanded macro with a newline (but without the preceding backslash character). When formal arguments are used to define a text macro, the scope of the formal argument shall extend up to the end of the macro text. A formal argument can be used in the macro text in the same manner as an identifier. If a one-line comment (that is, a comment specified with the characters //) is included in the text, then the comment shall not become part of the substituted text. The macro text can be blank, in which case the text macro is defined to be empty, and no text is substituted when the macro is used. The syntax for using a text macro is given in Syntax 19-3. 352 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C text_macro_usage ::= `text_macro_identifier [ ( list_of_actual_arguments ) ] list_of_actual_arguments ::= actual_argument { , actual_argument } actual_argument ::= expression Syntax 19-3—Syntax for text macro usage For a macro without arguments, the text shall be substituted “as is” for every occurrence of `text_macro_name. However, a text macro with one or more arguments shall be expanded by substituting each formal argument with the expression used as the actual argument in the macro usage. Once a text macro name has been defined, it can be used anywhere in a source description; that is, there are no scope restrictions. Text macros can be defined and used interactively. The text macro name shall be a simple identifier. The text specified for macro text shall not be split across the following lexical tokens: — Comments — Numbers — Strings — Identifiers — Keywords — Operators Examples: `define wordsize 8 reg [1:`wordsize] data; //define a nand with variable delay `define var_nand(dly) nand #dly `var_nand(2) g121 (q21, n10, n11); `var_nand(5) g122 (q22, n10, n11); The following is illegal syntax because it is split across a string: `define first_half "start of string $display(`first_half end of string"); NOTES: 1) Each actual argument is substituted for the corresponding formal argument literally. Therefore, when an expression is used as an actual argument, the expression will be substituted in its entirety. This may cause an expression to be evaluated more than once if the formal argument was used more than once in the macro text. For example, `define max(a,b)((a) > (b) ? (a) : (b)) n = `max(p+q, r+s) ; will expand as Copyright © 2001 IEEE. All rights reserved. 353 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® n = ((p+q) > (r+s)) ? (p+q) : (r+s) ; Here, the larger of the two expressions p + q and r + s will be evaluated twice. 2) The word define is known as a compiler directive keyword, and it is not part of the normal set of keywords. Thus, normal identifiers in a Verilog HDL source description can be the same as compiler directive keywords (although this is not recommended). The following problems should be considered: a) Text macro names may not be the same as compiler directive keywords. b) Text macro names can re-use names being used as ordinary identifiers. For example, signal_name and `signal_name are different. c) Redefinition of text macros is allowed; the latest definition of a particular text macro read by the compiler prevails when the macro name is encountered in the source text. 3) The macro text can contain usages of other text macros. Such usages shall be substituted after the original macro is substituted, not when it is defined. It shall be an error for a macro to expand directly or indirectly to text containing another usage of itself (a recursive macro). 19.3.2 `undef The directive `undef shall undefine a previously defined text macro. An attempt to undefine a text macro that was not previously defined using a `define compiler directive can result in a warning. The syntax for `undef compiler directive is given in Syntax 19-4. undefine_compiler_directive ::= `undef text_macro_identifier Syntax 19-4—Syntax for undef compiler directive An undefined text macro has no value, just as if it had never been defined. 19.4 `ifdef, `else, `elsif, `endif, `ifndef These conditional compilation compiler directives are used to include optionally lines of a Verilog HDL source description during compilation. The `ifdef compiler directive checks for the definition of a text_macro_name. If the text_macro_name is defined, then the lines following the `ifdef directive are included. If the text_macro_name is not defined and an `else directive exists, then this source is compiled. The `ifndef compiler directive checks for the definition of a text_macro_name. If the text_macro_name is not defined, then the lines following the `ifndef directive are included. If the text_macro_name is defined and an `else directive exists, then this source is compiled. If the `elsif directive exists (instead of the `else) the compiler checks for the definition of the text_macro_name. If the name exists the lines following the `elsif directive are included. The `elsif directive is equivalent to the compiler directive sequence `else `ifdef ... `endif. This directive does not need a corresponding `endif directive. This directive must be preceded by an `ifdef or `ifndef directive. These directives may appear anywhere in the source description. 354 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Situations where the `ifdef, `else, `elsif, `endif, and `ifndef compiler directives may be useful include: — Selecting different representations of a module such as behavioral, structural, or switch level — Choosing different timing or structural information — Selecting different stimulus for a given run The `ifdef, `else, `elsif, `endif, and `ifndef compiler directives have the syntax shown in Syntax 19-5. conditional_compilation_directive ::= ifdef_directive | ifndef_directive ifdef_directive ::= `ifdef text_macro_identifier ifdef_group_of_lines { `elsif text_macro_identifier elsif_group_of_lines } [ `else else_group_of_lines ] `endif ifndef_directive ::= `ifndef text_macro_identifier ifndef_group_of_lines { `elsif text_macro_identifier elsif_group_of_lines } [ `else else_group_of_lines ] `endif Syntax 19-5—Syntax for conditional compilation directives The text_macro_identifier is a Verilog HDL simple_identifier. The ifdef_group_of_lines, ifndef_group_of_lines, elsif_group_of_lines and the else_group_of_lines are parts of a Verilog HDL source description. The `else and `elsif compiler directives and all of the groups of lines are optional. The `ifdef, `else, `elsif, and `endif compiler directives work together in the following manner: — When an `ifdef is encountered, the `ifdef text macro identifier is tested to see if it is defined as a text macro name using `define within the Verilog HDL source description. — If the `ifdef text macro identifier is defined, the `ifdef group of lines is compiled as part of the description and if there are `else or `elsif compiler directives, these compiler directives and corresponding groups of lines are ignored. — If the `ifdef text macro identifier has not been defined, the `ifdef group of lines is ignored. — If there is an `elsif compiler directive, the `elsif text macro identifier is tested to see if it is defined as a text macro name using `define within the Verilog HDL source description. — If the `elsif text macro identifier is defined, the `elsif group of lines is compiled as part of the descrip- tion and if there are other `elsif or `else compiler directives, the other `elsif or `else directives and corresponding groups of lines are ignored. — If the first `elsif text macro identifier has not been defined, the first `elsif group of lines is ignored. — If there are multiple `elsif compiler directives, they are evaluated like the first `elsif compiler directive in the order they are written in the Verilog HDL source description. — If there is an `else compiler directive, the `else group of lines is compiled as part of the description. Copyright © 2001 IEEE. All rights reserved. 355 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® The `ifndef, `else, `elsif, and `endif compiler directives work together in the following manner: — When an `ifndef is encountered, the `ifndef text macro identifier is tested to see if it is defined as a text macro name using `define within the Verilog HDL source description. — If the `ifndef text macro identifier is not defined, the `ifndef group of lines is compiled as part of the description and if there are `else or `elsif compiler directives, these compiler directives and corresponding groups of lines are ignored. — If the `ifndef text macro identifier is defined, the `ifndef group of lines is ignored. — If there is an `elsif compiler directive, the `elsif text macro identifier is tested to see if it is defined as a text macro name using `define within the Verilog HDL source description. — If the `elsif text macro identifier is defined, the `elsif group of lines is compiled as part of the descrip- tion and if there are other `elsif or `else compiler directives, the other `elsif or `else directives and corresponding groups of lines are ignored. — If the first `elsif text macro identifier has not been defined, the first `elsif group of lines is ignored. — If there are multiple `elsif compiler directives, they are evaluated like the first `elsif compiler directive in the order they are written in the Verilog HDL source description. — If there is an `else compiler directive, the `else group of lines is compiled as part of the description. Although the names of compiler directives are contained in the same name space as text macro names, the names of compiler directives are considered not to be defined by `ifdef, `ifndef, and `elseif. Nesting of `ifdef, `ifndef, `else, `elsif, and `endif compiler directives shall be permitted. NOTE—Any group of lines that the compiler ignores still has to follow the Verilog HDL lexical conventions for white space, comments, numbers, strings, identifiers, keywords, and operators. Examples: Example 1—The example below shows a simple usage of an `ifdef directive for conditional compilation. If the identifier behavioral is defined, a continuous net assignment will be compiled in; otherwise, an and gate will be instantiated. module and_op (a, b, c); output a; input b, c; `ifdef behavioral wire a = b & c; `else and a1 (a,b,c); `endif endmodule 356 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Example 2—The following example shows usage of nested conditional compilation directives. module test(out); output out; `define wow `define nest_one `define second_nest `define nest_two `ifdef wow initial $display(“wow is defined”); `ifdef nest_one initial $display(“nest_one is defined”); `ifdef nest_two initial $display(“nest_two is defined”); `else initial $display(“nest_two is not defined”); `endif `else initial $display(“nest_one is not defined”); `endif `else initial $display(“wow is not defined”); `ifdef second_nest initial $display(“second_nest is defined”); `else initial $display(“second_nest is not defined”); `endif `endif endmodule Example 3—The following example shows usage of chained nested conditional compilation directives. module test; `ifdef first_block `ifndef second_nest initial $display(“first_block is defined”); `else initial $display(“first_block and second_nest defined”); `endif `elsif second_block initial $display(“second_block defined, first_block is not”); `else `ifndef last_result initial $display(“first_block, second_block,” “ last_result not defined.”); `elsif real_last initial $display(“first_block, second_block not defined,” “ last_result and real_last defined.”); `else initial $display(“Only last_result defined!”); `endif `endif endmodule Copyright © 2001 IEEE. All rights reserved. 357 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 19.5 `include The file inclusion (`include) compiler directive is used to insert the entire contents of a source file in another file during compilation. The result is as though the contents of the included source file appear in place of the `include compiler directive. The `include compiler directive can be used to include global or commonly used definitions and tasks without encapsulating repeated code within module boundaries. Advantages of using the `include compiler directive include the following: — Providing an integral part of configuration management — Improving the organization of Verilog HDL source descriptions — Facilitating the maintenance of Verilog HDL source descriptions The syntax for the `include compiler directive is given in Syntax 19-6. include_compiler_directive ::= `include "filename" Syntax 19-6—Syntax for include compiler directive The compiler directive `include can be specified anywhere within the Verilog HDL description. The filename is the name of the file to be included in the source file. The filename can be a full or relative path name. Only white space or a comment may appear on the same line as the `include compiler directive. A file included in the source using the `include compiler directive may contain other `include compiler directives. The number of nesting levels for include files shall be finite. Examples: Examples of `include compiler directives are as follows: `include "parts/count.v" `include "fileB" `include "fileB" // including fileB NOTE—Implementations may limit the maximum number of levels to which include files can be nested, but the limit shall be at least 15. 19.6 `resetall When `resetall compiler directive is encountered during compilation, all compiler directives are set to the default values. This is useful for ensuring that only those directives that are desired in compiling a particular source file are active. The recommended usage is to place `resetall at the beginning of each source text file, followed immediately by the directives desired in the file. 19.7 `line The compiler is expected to maintain the current line and the filename of the file being compiled. The line 358 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C number (`line) compiler directive is used to reset the current line number and filename of the current file to the line number and filename presented. This can be used to reflect the location in an original file; if the actual source file has been modified by addition or reduction of lines. After specifying the new line number or file name, the compiler can correctly refer to the original source file location. For example error messages, source code debugging, etc. can direct the user to the actual original line. The syntax for the `line compiler directive is given in Syntax 19-7. line_compiler_directive ::= `line number "filename" level Syntax 19-7—Syntax for line compiler directive The directive can be specified anywhere within the Verilog HDL source description. The number parameter is the new line number of the next line. The filename parameter is the new name of the file. The filename can be a full or relative path name. The level parameter indicates whether an include file has been entered (value is 1), an include file is exited (value is 2), or neither has been done (value is 0). The results of this directive are not affected by the compiler directive `resetall. As the compiler processes the remainder of the file and new files, the line number shall be incremented as each line is read and the filename shall be updated to the new current file being processed. When beginning to read include files, the current line and filename shall be stored for restoration at the termination of the include file. The updated line number and filename information shall be available for PLI access. The mechanism of library searching is not affected by the effects of the `line compiler directive. 19.8 `timescale This directive specifies the time unit and time precision of the modules that follow it. The time unit is the unit of measurement for time values such as the simulation time and delay values. To use modules with different time units in the same design, the following timescale constructs are useful: — The `timescale compiler directive to specify the unit of measurement for time and precision of time in the modules in the design — The $printtimescale system task to display the time unit and precision of a module — The $time and $realtime system functions, the $timeformat system task, and the %t format speci- fication to specify how time information is reported The `timescale compiler directive specifies the unit of measurement for time and delay values and the degree of accuracy for delays in all modules that follow this directive until another `timescale compiler directive is read. If there is no `timescale specified or it has been reset by a `resetall directive, the time unit and precision are simulator specific. It shall be an error if some modules have a `timescale specified and others do not. The syntax for the `timescale directive is given in Syntax 19-8. Copyright © 2001 IEEE. All rights reserved. 359 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® timescale_compiler_directive ::= `timescale time_unit / time_precision Syntax 19-8—Syntax for timescale compiler directive The time_unit argument specifies the unit of measurement for times and delays. The time_precision argument specifies how delay values are rounded before being used in simulation. The values used are accurate to within the unit of time specified here, even if there is a smaller time_precision argument elsewhere in the design. The smallest time_precision argument of all the `timescale compiler directives in the design determines the precision of the time unit of the simulation. The time_precision argument shall be at least as precise as the time_unit argument; it cannot specify a longer unit of time than time_unit. The integers in these arguments specify an order of magnitude for the size of the value; the valid integers are 1, 10, and 100. The character strings represent units of measurement; the valid character strings are s, ms, us, ns, ps, and fs. The units of measurement specified by these character strings are given in Table 86. Table 86—Arguments of time_precision Character string Unit of measurement s seconds ms milliseconds us microseconds ns nanoseconds ps picoseconds fs femtoseconds Examples: The following example shows how this directive is used: `timescale 1 ns / 1 ps Here, all time values in the modules that follow the directive are multiples of 1 ns because the time_unit argument is “1 ns”. Delays are rounded to real numbers with three decimal places—or precise to within one thousandth of a nanosecond—because the time_precision argument is “1 ps,” or one thousandth of a nanosecond. Consider the following example: `timescale 10 us / 100 ns 360 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The time values in the modules that follow this directive are multiples of 10 us because the time_unit argument is “10 us”. Delays are rounded to within one tenth of a microsecond because the time_precision argument is “100 ns,” or one tenth of a microsecond. The following example shows a `timescale directive in the context of a module: `timescale 10 ns / 1 ns module test; reg set; parameter d = 1.55; initial begin #d set = 0; #d set = 1; end endmodule The `timescale 10 ns / 1 ns compiler directive specifies that the time unit for module test is 10 ns. As a result, the time values in the module are multiples of 10 ns, rounded to the nearest 1 ns and, therefore, the value stored in parameter d is scaled to a delay of 16 ns. This means that the value 0 is assigned to reg set at simulation time 16 ns (1.6 × 10 ns), and the value 1 at simulation time 32 ns. Parameter d retains its value no matter what timescale is in effect. These simulation times are determined as follows: a) The value of parameter d is rounded from 1.55 to 1.6 according to the time precision. b) The time unit of the module is 10 ns, and the precision is 1 ns, so the delay of parameter d is scaled from 1.6 to 16. c) The assignment of 0 to reg set is scheduled at simulation time 16 ns and the assignment of 1 at simulation time 32 ns. The time values are not rounded when the assignments are scheduled. 19.9 `unconnected_drive and `nounconnected_drive All unconnected input ports of a module appearing between the directives `unconnected_drive and `nounconnected_drive are pulled up or pulled down instead of the normal default. The directive `unconnected_drive takes one of two arguments—pull1 or pull0. When pull1 is specified, all unconnected input ports are automatically pulled up. When pull0 is specified, unconnected ports are pulled down. These directives shall be specified in pairs, and outside of the module declarations. The `resetall directive includes the effects of a `nounconnected_drive directive. Copyright © 2001 IEEE. All rights reserved. 361 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 20. PLI overview 20.1 PLI purpose and history (informative) Clause 20 through Clause 27 and Annex E through Annex G describe the C language procedural interface standard and interface mechanisms that are part of the Verilog HDL. This procedural interface, known as the Programming Language Interface, or PLI, provides a means for Verilog HDL users to access and modify data in an instantiated Verilog HDL data structure dynamically. An instantiated Verilog HDL data structure is the result of compiling Verilog HDL source descriptions and generating the hierarchy modeled by module instances, primitive instances, and other Verilog HDL constructs that represent scope. The PLI procedural interface provides a library of C language functions that can directly access data within an instantiated Verilog HDL data structure. A few of the many possible applications for the PLI procedural interface are: — C language delay calculators for Verilog model libraries that can dynamically scan the data structure of a Verilog software product and then dynamically modify the delays of each instance of models from the library — C language applications that dynamically read test vectors or other data from a file and pass the data into a Verilog software product — Custom graphical waveform and debugging environments for Verilog software products — Source code decompilers that can generate Verilog HDL source code from the compiled data struc- ture of a Verilog software product — Simulation models written in the C language and dynamically linked into Verilog HDL simulations — Interfaces to actual hardware, such as a hardware modeler, that dynamically interact with simulations This document standardizes the Verilog PLI that has been in use since the mid-1980s. This standard comprises three primary generations of the Verilog PLI. a) Task/function routines, called TF routines, make up the first generation of the PLI. These routines, most of which start with the characters tf_, are primarily used for operations involving user-defined task/function arguments, along with utility functions, such as setting up call-back mechanisms and writing data to output devices. The TF routines are sometimes referred to as utility routines b) Access routines, called ACC routines, form the second generation of the PLI. These routines, which all start with the characters acc_, provide an object-oriented access directly into a Verilog HDL structural description. ACC routines are used to access and modify information, such as delay values and logic values on a wide variety of objects that exist in a Verilog HDL description. There is some overlap in functionality between ACC routines and TF routines. c) Verilog Procedural Interface routines, called VPI routines, are the third generation of the PLI. These routines, all of which start with the characters vpi_, provide an object-oriented access for both Verilog HDL structural and behavioral objects. The VPI routines are a superset of the functionality of the TF routines and ACC routines. 20.2 User-defined system task or function names A user-defined system task or function name is the name that will be used within a Verilog HDL source file to invoke specific PLI applications. The name shall adhere to the following rules: — The first character of the name shall be the dollar sign character ( $ ) — The remaining characters shall be letters, digits, the underscore character ( _ ) or the dollar character ($) — Uppercase and lowercase letters shall be considered to be unique—the name is case sensitive — The name can be any size, and all characters are significant 362 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 20.3 User-defined system task or function types The type of a user-defined system task or function determines how a PLI application is called from the Verilog HDL source code. The types are: — A user task can be used in the same places a Verilog HDL task can be used (refer to 10.2). A userdefined system task can read and modify the arguments of the task, but does not return any value. — A user function can be used in the same places a Verilog HDL function can be used (refer to 10.3). A user-defined system function can read and modify the arguments of the function, and shall return a scalar or vector value. The bit width of the return value shall be determined by a user-supplied sizetf application (see 21.1.1). — A user real-function can be used in the same places a Verilog HDL function can be used (refer to 10.3). A user-defined system real-function can read and modify the arguments of the function, and will return a double-precision floating point value. 20.4 Overriding built-in system task and function names Clause 17 defines a number of built-in system tasks and functions that are part of the Verilog language. In addition, software products can include other built-in system tasks and functions specific to the product. These built-in system task and function names begin with the dollar sign character ( $ ) just as user-defined system task and function names. If a user-provided PLI application is associated with the same name as a built-in system task or function (using the PLI interface mechanism), the user-provided C application shall override the built-in system task/ function, replacing its functionality with that of the user-provided C application. For example, a user could write a random number generator as a PLI application and then associate the application with the name $random, thereby overriding the built-in $random function with the user’s application. Verilog timing checks, such as $setup, are not system tasks, and cannot be overridden. The system functions $signed and $unsigned can be overridden. These system functions are unique in the Verilog HDL, in that the return width is based on the width of their argument. If overridden, the PLI version shall have the same return width for all instances of the system function. The PLI return width is defined by the PLI sizetf routine. 20.5 User-supplied PLI applications User-supplied PLI applications are C language functions that utilize the library of PLI C functions to access and interact dynamically with Verilog HDL software implementations as the Verilog HDL source code is executed. These PLI applications are not independent C programs. They are C functions, which are linked into a software product, and become part of the product. This allows the PLI application to be called when the userdefined system task or function $ name is compiled or executed in the Verilog HDL source code. 20.6 PLI interface mechanism The PLI interface mechanism provides a means to have PLI applications called for various reasons when the associated system task or function $ name is encountered in the Verilog HDL source description. For example, when a Verilog HDL simulator first compiles the Verilog HDL source description, a specific PLI application can be called that performs syntax checking to ensure the user-defined system task or function is being used correctly. Then, as simulation is executing, a different PLI application can be called to perform the operations required by the PLI application. Other PLI applications can be automatically called by the simulator for miscellaneous reasons, such as the end of a simulation time step or a logic value change on a specific signal. Copyright © 2001 IEEE. All rights reserved. 363 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® The PLI provides two interface mechanisms: — The TF and ACC interface mechanism is an older interface, which can be used to associate PLI applications which use routines from the ACC and TF function libraries. This interface mechanism is described in Clause 21. — The VPI interface mechanism is a newer interface, which can be used to associate PLI applications which use routines from the VPI function libraries. This interface mechanism is described in Clause 26. Instances of system tasks and functions which are defined using the TF and ACC interface mechanism can only be accessed using the TF and ACC function libraries. Instances of system tasks and functions which are defined using the VPI interface mechanism can only be accessed using the VPI function library. 20.7 User-defined system task and function arguments When a user-defined system task or function is used in a Verilog HDL source file, it can have arguments that can be used by the PLI applications associated with the system task or function. In the following example, the user-defined system task $get_vector has two arguments: $get_vector(“test_vector.pat”, input_bus); The arguments to a system task or function are referred to as task/function arguments (often abbreviated as tfargs). These arguments are not the same as C language arguments. When the PLI applications associated with a user-defined system task or function are called, the task/function arguments are not passed to the PLI application. Instead, a number of PLI routines are provided that allow the PLI applications to read and write to the task/function arguments. Refer to the sections on ACC routines, TF routines and VPI routines for information on specific routines that work with task/function arguments. 20.8 PLI include files The libraries of PLI functions are defined in C include files, which are a normative part of the 1364 standard. These files also define constants, structures, and other data used by the library of PLI routines and the interface mechanisms. The files are acc_user.h (listed in Annex E), veriuser.h (listed in Annex F) and vpi_user.h (listed in Annex G). — PLI applications that use the TF routines shall include the file veriuser.h. — PLI applications that use the ACC routines shall include the file acc_user.h. — PLI applications that use the VPI routines shall include the file vpi_user.h. 20.9 PLI Memory Restrictions Memory allocated by the PLI routines is not to be modified by the user, with the exception of the value storage returned by the PLI routines tf_exprinfo() and tf_nodeinfo(), as defined in 25.15 and 25.35. 364 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 21. PLI TF and ACC interface mechanism The interface mechanism described in this section provides a means for users to link applications based on PLI task/function (TF) routines and access (ACC) routines to Verilog software products. Through the interface mechanism, a user can: — Specify a user-defined system task or function name that can be included in Verilog HDL source descriptions; the user-defined system task or function name shall begin with a dollar sign ($), such as $get_vector — Provide one or more PLI C applications to be called by a software product (such as a logic simulator) — Define which PLI C applications are to be called—and when the applications should be called— when the user-defined system task or function name is encountered in the Verilog HDL source description — Define whether the PLI applications should be treated as functions (which return a value) or tasks (analogous to subroutines in other programming languages) — Define a data argument to be passed to the PLI applications each time they are called NOTE—The PLI interface mechanism described in this section does not apply to applications that use the Verilog Procedural Interface (VPI) routines; these routines use the VPI registry mechanism described in Clause 26 and Clause 27. 21.1 User-supplied PLI applications User-supplied PLI applications are C language functions that utilize the library of PLI C functions to access and interact dynamically with Verilog HDL software implementations as the Verilog HDL source code is executed. These PLI applications are not independent C programs. They are C functions, which are linked into a software product, and become part of the product. This allows the PLI application to be called when the userdefined system task or function $ name is compiled or executed in the Verilog HDL source code. The PLI interface mechanism for TF and ACC routines provides five classes of user-supplied PLI applications: checktf applications, sizetf applications, calltf applications, misctf applications, and consumer applications. The sizetf, checktf, calltf, and misctf routines are called during specific periods during processing. The purpose of each of the PLI application classes is explained in the following subsections. 21.1.1 The sizetf class of PLI applications A sizetf PLI application can be used in conjunction with user-defined system functions. A function shall return a value, and software products that execute the system function may need to determine how many bits wide that return value shall be. The sizetf application may be called early in the process, prior to a complete instantiation of the design. As a result, access to objects may be limited at this time. Each sizetf function shall be called at most once. It shall be called if its associated system function appears in the design. The value returned by the sizetf function shall be the number of bits that the calltf routine shall provide as the return value for the system function. If no sizetf application is specified for a user-defined system function, the function shall return 32-bits. The sizetf application shall not be called for user-defined system tasks or real-functions. 21.1.2 The checktf class of PLI applications A checktf PLI application shall be called when the user-defined system task or function name is encountered during parsing or compiling the Verilog HDL source code. This application is typically used to check the correctness of any arguments used with the system task in the Verilog HDL source code. The checktf PLI application shall be called one time for each instance of a system task or function in the source description. Providing a checktf application is optional, but it is recommended that any arguments used with the system task or function be checked for correctness to avoid problems when the calltf or other PLI applications read and perform operations on the arguments. The checktf shall be called at the earliest possible time after all Copyright © 2001 IEEE. All rights reserved. 365 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® simulation data structures required by the PLI are available. Generally this means after the design is fully instantiated, but no simulation events have occurred. By the time the checktf application is called, all PLI routines can be used without concern for the state of the process, with the exception of setting the return value of user defined system functions. The return value of a user defined system function can only be set during a calltf application. 21.1.3 The calltf class of PLI applications A calltf PLI application shall be called each time the associated user-defined system task or function is executed within the Verilog HDL source code. For example, the following Verilog loop would call the PLI calltf application that is associated with the $get_vector user-defined system task name 1024 times: for (i = 1; i <= 1024; i = i + 1) @(posedge clk) $get_vector(“test_vector.pat”, input_bus); In this example, the user-supplied PLI calltf application might read a test vector from a file called test_vector.pat (the first task/function argument), perhaps manipulate the vector to put it in a proper format for Verilog, and then assign the vector value to the second task/function argument called input_bus. 21.1.4 The misctf class of PLI applications A misctf PLI application shall be called by a Verilog software product for miscellaneous reasons while the Verilog HDL source description is being executed. Among these reasons can be the end of a simulation time step, a logic value change on a user-defined system task/function argument, or the execution of the $stop and $finish built-in system tasks. When the software product calls the misctf PLI application, it shall pass in a reason argument, which can be used within the misctf application to determine why the application was called. The reason argument shall be a predefined integer constant. Table 87 and Table 88 list the reasons for which the misctf application can be called. For most reasons, the misctf routine will not be called until the instance of the system task has been executed (at which point the calltf routine is called). The following reasons are exceptions, and will be called for each instance of the system task in the design regardless of whether or not it has been executed: — reason_endofcompile — reason_save — reason_startofsave — reason_restart — reason_endofreset — reason_reset 21.1.5 The consumer class of PLI applications A consumer PLI application shall be called through a PLI callback mechanism referred to as the Value Change Link (VCL). Using the VCL, another PLI application, typically the calltf application, can place VCL flags on objects within the Verilog HDL data structure, such as a specific net. Whenever an object with a VCL flag changes value during a simulation, the consumer PLI application shall be called and passed information about the change. 21.2 Associating PLI applications to a class and system task/function name Each user-provided PLI application is a standard C language function that makes use of the library of PLI functions. These user-provided PLI applications shall be associated with both the class of application (such as calltf or checktf) and the user-defined system task or function $ name. In addition, the user-defined name shall be declared as either a system task or a system function. 366 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C For the TF and ACC interface mechanism, the method of associating PLI applications with a class and system task/function name is not defined as part of this standard. Each software product vendor shall define an association mechanism specific to their product. Refer to the documentation provided by the vendor for instructions on associating PLI applications to classes and system task/function names and then linking the PLI applications into the software products of the vendor. 21.3 PLI application arguments When the calltf, checktf, and sizetf PLI applications are called by a Verilog software implementation, they shall be passed two C arguments, data and reason, in that order. When the misctf application is called, it shall be passed three C arguments, data, reason, and paramvc, in that order. These arguments are defined in more detail in the following subsections. 21.3.1 The data C argument The data C argument shall be an integer value. The value is defined by the user at the time the PLI applications are associated with a user-defined system task/function name. This value can be used to allow several different system task/function names to use the same calltf, checktf, sizetf, or misctf applications. To do this, each system task/function name would be associated with the same PLI applications, but each would have a different value for the user-defined data argument. When a PLI application is called, it can then check the value of the data argument to determine which system task/function name was used to call the application. 21.3.2 The reason C argument The reason C argument shall be a predefined integer constant that is passed to the calltf, checktf, sizetf, and misctf applications each time the applications are called. Generally, the calltf, checktf, and sizetf applications do not need to check the reason argument, since these applications can only be called under specific circumstances. The misctf application, however, can be called for a wide variety of reasons, and therefore it should always examine the reason argument to determine why the application was called. The value for the reason argument is defined in the PLI include file veriuser.h. The reason constant that is passed is based on the class of the PLI application, as follows: — The calltf application is passed the reason constant reason_calltf. — The checktf application is passed the reason constant reason_checktf. — The sizetf application is passed the reason constant reason_sizetf. — The misctf application is passed one of the constants listed in Table 87. Software implementations can define additional reason constants to be passed to the misctf application. Table 88 lists some common reason constants that can be available in some software implementations. Table 87—(normative) Predefined misctf reason constants Integer constant reason_endofcompile reason_paramvc reason_synch reason_rosynch reason_reactivate reason_finish Reason end of Verilog source compilation/start of execution a change of value on a user-defined system task or function argument end of a time step flagged by tf_synchronize() end of a time step flagged by tf_rosynchronize() a simulation event scheduled by tf_setdelay() the $finish() built-in system task executed Copyright © 2001 IEEE. All rights reserved. 367 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Table 88—(informative) Additional misctf reason constants Integer constant reason_paramdrc reason_force reason_release reason_disable reason_interactive reason_scope reason_startofsave reason_save reason_restart reason_reset reason_endofreset Reason a value change on the driver of a user-defined system task or function argument execution of a procedural force or procedural continuous assignment on any net, reg, integer variable, time variable or real variable execution of a procedural release or procedural deassign on any net, reg, integer variable, time variable or real variable execution of a procedural disable statement execution of the $stop() built-in system task execution of the $scope() built-in system task start of execution of the $save() built-in system task completion of execution of the $save() built-in system task execution of the $restart() built-in system task start of execution of the $reset() built-in system task completion of execution of the $reset() built-in system task 21.3.3 The paramvc C argument The paramvc C argument shall be an integer value passed to the misctf application. The value of paramvc shall indicate which task/function argument changed value when the misctf application was called back after activating the utility routine tf_asynchon(). This routine shall cause the misctf application to be called with a reason argument of reason_paramvc or reason_paramdrc. Task/function argument index numbering shall proceed from left to right, with the left-most argument being number 1. 368 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 22. Using ACC routines This clause presents a general discussion of how and why to use PLI ACC routines. Clause 23 defines the ACC routine syntax, listed in alphabetical order. 22.1 ACC routine definition ACC routines are C programming language functions that provide procedural access to information within the Verilog HDL. ACC routines perform one of two functions: a) Read data about particular objects in the Verilog HDL description directly from internal data structures. b) Write new information about certain objects in the Verilog HDL description into the internal data structures. ACC routines shall read information about the following objects: — Module instances — Module ports — Module or data paths — Intermodule paths — Top-level modules — Primitive instances — Primitive terminals — Nets — Reg variables — Parameters — Specparams — Timing checks — Named events — Integer, real, and time variables ACC routines shall read and write information on the following objects: — Intermodule path delays — Module path delays — Module input port delays (MIPDs) — Primitive instance delays — Timing check limits — Reg variable logic values — Net logic values (force/release only) — Integer, real and time variable values — Sequential UDP logic values 22.2 The handle data type A handle is a predefined data type that is a pointer to a specific object in the design hierarchy. Each handle conveys information to ACC routines about a unique instance of an accessible object—information about the type of the object, plus how and where to find data about the object. Copyright © 2001 IEEE. All rights reserved. 369 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Most ACC routines require a handle argument to indicate the objects about which they need to read or write information. The PLI provides two categories of ACC routines that return handles for objects: handle routines, which begin with the prefix acc_handle_, and next routines, which begin with the prefix acc_next_. Refer to 22.4.2 for a discussion of handle routines and 22.4.3 for more information about next routines. Handles shall be passed to and from ACC routines through handle variables. To declare a handle variable, the keyword handle (all lowercase) shall be used, followed by the variable name, as in this example: handle net_handle; After declaring a handle variable, it can be passed to any ACC routine that requires a handle argument or be used to receive a handle returned by an ACC routine. The following C language code fragment uses the variable net_handle to store the handle returned by the ACC routine acc_handle_object(): handle net_handle; net_handle = acc_handle_object(“top.mod1.w3”); 22.3 Using ACC routines 22.3.1 Header files The header file acc_user.h shall be included in any C language source file containing an application program that calls ACC routines. The acc_user.h file is listed in Annex E. 22.3.2 Initializing ACC routines The ACC routine acc_initialize() shall initialize the environment for ACC routines and shall be called from the C language application program before the program invokes any other ACC routines. 22.3.3 Exiting ACC routines Before exiting a C language application program that calls ACC routines, the ACC routine acc_close() should be called. This routine shall reset ACC routine configuration parameters back to their defaults, and it shall also free memory allocated by the ACC routines. 22.4 List of ACC routines by major category The ACC routines are divided into the following major categories: — Fetch routines — Handle routines — Next routines — Modify routines — VCL routines — Miscellaneous routines This subclause contains a summary list of each major category. The ACC routines sorted by the types of objects they work with are listed in 22.5. Clause 23 presents an alphabetical list of all ACC routines, with their functions, syntax, and usage. 370 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 22.4.1 Fetch routines Fetch routines shall return a variety of information about different objects in the design hierarchy. The name of each routine begins with the prefix acc_fetch_ and indicates the type of information desired. For example, acc_fetch_fullname() retrieves the full hierarchical path name for any named object, while acc_fetch_paramval() retrieves the value of a parameter or specparam. Table 89—List of fetch routines ACC routine acc_fetch_argc() acc_fetch_argv() acc_fetch_attribute() acc_fetch_attribute_int() acc_fetch_attribute_str() acc_fetch_defname() acc_fetch_delay_mode() acc_fetch_delays() acc_fetch_direction() acc_fetch_edge() acc_fetch_fullname() acc_fetch_fulltype() acc_fetch_index() acc_fetch_itfarg() acc_fetch_itfarg_int() acc_fetch_itfarg_str() acc_fetch_location() acc_fetch_name() acc_fetch_paramtype() acc_fetch_paramval() acc_fetch_polarity() acc_fetch_precision() acc_fetch_pulsere() acc_fetch_range() acc_fetch_size() acc_fetch_tfarg() acc_fetch_tfarg_int() acc_fetch_tfarg_str() acc_fetch_timescale_info() acc_fetch_type() acc_fetch_type_str() acc_fetch_value() Description Get the number of invocation command line arguments Get the invocation command line arguments Get the value of a Verilog parameter or specparam as a double Get the value of a Verilog parameter or specparam as an integer Get the value of a Verilog parameter or specparam as a string Get the definition name of a module or primitive Get the delay mode of a module instance Get the existing delays for a primitive, module path, timing check, intermodule path, or module input port Get the direction of a module port or primitive terminal Get the edge specifier of a module path input terminal Get the full hierarchical name of an object Get the full type description of an object as a predefined integer constant Get the index number of a port or terminal Get the value of an instance of a system task/function argument as a double Get the value of an instance of a system task/function argument as an integer Get the value of an instance of a system task/function argument as a string Get the location of an object in a Verilog source file Get the local name of an object Get the data type of a parameter or specparam Get the value of a parameter or specparam Get the polarity of a module path or data path Get the simulation time precision Get the current pulse handling values of a module path, intermodule path or module input port Get the range of a vector Get the bit size of a vector or port Get the value of a system task/function argument as a double Get the value of a system task/function argument as an integer Get the value of a system task/function argument as a string Get the timescale information for an object Get the general type classification of an object as an integer constant Get the string representation of a type or fulltype integer constant Get the logic or strength value of a net, reg, integer variable, time variable or real variable Copyright © 2001 IEEE. All rights reserved. 371 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 22.4.2 Handle routines Handle routines can return handles to a variety of objects in the design hierarchy. The name of each routine begins with the prefix acc_handle_ and indicates the type of handle desired. For example, acc_handle_object() retrieves a handle for a named object, while acc_handle_conn() retrieves a handle for a net connected to a particular terminal. Each handle routine shall return a handle to an object. This handle can, in turn, be passed as an argument to other ACC routines. Table 90—List of handle routines ACC routine Description acc_handle_by_name() acc_handle_condition() acc_handle_conn() acc_handle_datapath() acc_handle_hiconn() acc_handle_interactive_scope() acc_handle_itfarg() acc_handle_loconn() acc_handle_modpath() acc_handle_notifier() acc_handle_object() acc_handle_parent() acc_handle_path() acc_handle_pathin() acc_handle_pathout() acc_handle_port() acc_handle_scope() acc_handle_simulated_net() acc_handle_tchk() acc_handle_tchkarg1() acc_handle_tchkarg2() acc_handle_terminal() acc_handle_tfarg() acc_handle_tfinst() Get the handle to any named object Get the handle to the condition of a module path, data path, or timing check Get the handle to the net connected to a primitive, path, or timing check terminal Get the handle to a data path Get the handle to the hierarchically higher net connected to a module port bit Get the handle to the current simulation interactive scope Get the handle to an argument of a specific system task/function instance Get the handle to the hierarchically lower net connected to a module port bit Get the handle to a module path Get the handle to the notifier argument of a timing check Get the handle to any named object Get the handle to the parent of an object Get the handle to an intermodule path Get the handle to the first net connected to a module path source Get the handle to the first net connected to a module path destination Get the handle to a module port based on the port index Get the handle to the scope containing an object Get the handle to the net associated with a collapsed net Get the handle to a timing check Get the handle to the first argument of a timing check Get the handle to the second argument of a timing check Get the handle to a terminal of a primitive based on the terminal index Get the handle to the object named in a system task/function argument Get the handle to the current instance of a system task/function 372 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 22.4.3 Next routines When used inside a C loop construct, next routines shall find each object of a given type that is related to a particular reference object in the design hierarchy. The name of each routine begins with the prefix acc_next_ and indicates the type of object desired, known as the target object. For example, acc_next_net() retrieves each net in a module, while acc_next_driver() retrieves each terminal driving a net. Each call to a next routine returns a handle to the object it finds. Most next routines require two arguments: — The first argument shall be a handle to a reference object. — The second argument shall be a handle that indicates whether to retrieve the first or next target object. The reference object shall indicate where the next routine shall look for the target object. The target object is the type of object to be returned by a next routine. Table 91 summarizes how next routines shall find each target object associated with a given reference object. Table 91—How next routines use the target object argument When A next routine shall return the target object is null a handle to the first target object related to the reference object the target object is a handle to the last target object returned a handle to the next target object related to the reference object no target objects remain for the reference object a null handle no target objects are found initially for the reference object a null handle an error occurs a null handle Each call to a next routine shall return only one handle. Therefore, to retrieve all target objects for a particular reference object, the following process can be used: a) Chose an appropriate ACC routine to retrieve the handle of the desired reference object. b) Set the target object handle variable to null. When a next routine is called with a null target handle, it shall return the first target associated with the reference. c) Call the next routine, assigning the return value to the same variable as the target object argument. This automatically updates the target object argument to point to the last object found. d) Place the next routine call inside a C while loop that terminates when the loop control value is null. When a next routine cannot access any more target objects, it shall return a null. NOTE—Most next routines can return objects in an arbitrary order. However, certain next routines shall return objects in a defined order, as noted in the description of the routine in Clause 23. The following example, display_net_names, uses a next routine to display the names of all nets in a module. Copyright © 2001 IEEE. All rights reserved. 373 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® #include "acc_user.h" display_net_names() { handle module_handle; handle net_handle; /*initialize environment for access routines*/ acc_initialize(); /*get handle for module*/ module_handle = acc_handle_tfarg(1); /*display names of all nets in the module*/ net_handle = null; while( net_handle = acc_next_net( module_handle, net_handle ) ) io_printf("Net name is: %s\n", acc_fetch_fullname(net_handle) ); acc_close(); } ACC routine acc_next() acc_next_bit() acc_next_cell() acc_next_cell_load() acc_next_child() acc_next_driver() acc_next_hiconn() acc_next_input() acc_next_load() acc_next_loconn() acc_next_modpath() acc_next_net() acc_next_output() acc_next_parameter() acc_next_port() acc_next_portout() acc_next_primitive() acc_next_scope() acc_next_specparam() acc_next_tchk() acc_next_terminal() acc_next_topmod() Table 92—List of next routines Description Get handles to all objects of a set of types Get handles to all bits of a port or vector Get handles to all cell modules in the current hierarchy and below Get handles to all cell loads on a net Get handles to all module instances within a module Get handles to all primitive terminals that drive a net Get handles to all nets connected hierarchically higher to a module port Get handles to all input terminals of a module path or data path Get handles to all primitive terminals driven by a net Get handles to all nets connected hierarchically lower to a module port Get handles to all paths in a module Get handles to all nets in a module Get handles to all output terminals of a module path or data path Get handles to all parameters in a module Get handles to all ports of a module or connected to a net Get handles to all output ports of a module Get handles to all primitive instances in a module Get handles to all hierarchy scopes within a scope Get handles to all specify block parameters in a module Get handles to all timing checks in a module Get handles to all terminals of a primitive Get handles to all top-level modules 374 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 22.4.4 Modify routines Modify routines shall alter the values of a variety of objects in the design hierarchy. Table 93 lists the types of values that shall be modified for particular objects. Table 93—Values that can be modified Modify routines alter Delay values Logic values Pulse handling values For these objects Primitives Module paths Intermodule paths Module input ports Timing checks Variable data types Net data types Sequential UDPs Module paths Intermodule paths Module input ports ACC routine acc_append_delays() acc_append_pulsere() acc_replace_delays() acc_replace_pulsere() acc_set_pulsere() acc_set_value() Table 94—List of modify routines Description Add delays to existing delays on primitives, module paths, timing checks, intermodule paths, and module input ports Add to existing pulse control values of module paths, intermodule paths and module input ports Replace existing delays on primitives, module paths, timing checks, intermodule paths and module input ports Replace existing values on pulse control values of module paths, intermodule paths and module input ports Set the pulse control values for a module path, intermodule path or module input port as a percentage of the delay Set and propagate a logic value onto a reg, integer variable, time variable, real variable or sequential UDP; continuously assign/deassign a reg or variable; force/release a net or reg or variable More details on using the acc_append_delays() and acc_replace_delays() ACC routines are provided in 22.8. 22.4.5 Miscellaneous routines Miscellaneous routines shall perform a variety of operations, such as initializing and configuring the ACC routine environment. Copyright © 2001 IEEE. All rights reserved. 375 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Table 95—List of miscellaneous routines ACC routine acc_close() acc_collect() acc_compare_handles() acc_configure() acc_count() acc_free() acc_initialize() acc_object_in_typelist() acc_object_of_type() acc_product_type() acc_product_version() acc_release_object() acc_reset_buffer() acc_set_interactive_scope() acc_set_scope() acc_version() Description Close ACC routine environment Collect an array of handles for a reference object Determine if two handles are for the same object Set the ACC routine environment parameters Count the number of objects related to a reference object Free up memory allocated by acc_collect() Initialize the ACC routine environment Determine if an object matches a set of types, fulltypes, or special properties Determine if an object matches a specific type, fulltype, or special property Get the type of software product being used Get the version of software product being used Release memory allocated by acc_next_input() or acc_next_output() Reset the string buffer Set the interactive scope of a software implementation Set the scope used by acc_handle_object() Get the version of the ACC routines being used 22.4.6 VCL routines The VCL shall allow a PLI application to monitor simulation value changes of selected objects. It consists of two ACC routines that instruct a Verilog simulator to start or stop informing an application when an object changes value. How the VCL routine is used is discussed in 22.10. Table 96—List of VCL routines ACC routine acc_vcl_add() acc_vcl_delete() Description Add a value change callback on an object Remove a value change callback 22.5 Accessible objects ACC routines shall access information about the following objects: — Module instances — Module ports — Individual bits of a port — Module or data paths — Intermodule paths 376 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C — Top-level modules — Primitive instances — Primitive terminals — Nets (scalars, vectors, and bit- or part-selects of vectors) — Regs (scalars, vectors, and bit- or part-selects of vectors) — Integer variables (and bit- or part-selects of integers) — Real and time variables — Named events — Parameters — Specparams — Timing checks — Timing check terminals — User-Defined system task/function arguments The following tables summarize the operations that can be performed for each of the above object types. Copyright © 2001 IEEE. All rights reserved. 377 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 22.5.1 ACC routines that operate on module instances Table 97—Operations on module instances To Obtain handles for module instances tagged as cells within a hierarchical scope and below Obtain handles for module instances within a particular module instance Obtain a handle to the parent (the module that contains the instance) Get the instance name Get the full hierarchical name Get the module definition name Get the fulltype of a module instance (cell instance, module instance, or top-level module) Get the delay mode of a module instance (none, zero, unit, distributed, or path) Get timescale information for a module instance Use acc_next_cell() acc_next_child() acc_handle_parent() acc_fetch_name() acc_fetch_fullname() acc_fetch_defname() acc_fetch_fulltype() acc_fetch_delay_mode() acc_fetch_timescale_info() 22.5.2 ACC routines that operate on module ports Table 98—Operations on module ports To Obtain handles for ports of a module instance Obtain handles for output ports of a module instance Obtain a handle for a particular port Obtain a handle to the parent (the module instance that contains the port) Obtain handles to hierarchically higher-connected nets Obtain handles to hierarchically lower-connected nets Obtain a handle to the hierarchically higher-connected net of a scalar module port or bit of a vector port Obtain a handle to the hierarchically lower-connected net of a scalar module port or bit of a vector port Get the instance name Get the full hierarchical name Get the port direction Get the port index number Get the fulltype of a module port Add VCL value change callback monitors Delete VCL value change callback monitors Read Module Input Port Delay (MIPD) Append to existing MIPD Replace existing MIPD Use acc_next_port() acc_next_portout() acc_handle_port() acc_handle_parent() acc_next_hiconn() acc_next_loconn() acc_handle_hiconn() acc_handle_loconn() acc_fetch_name() acc_fetch_fullname() acc_fetch_direction() acc_fetch_index() acc_fetch_fulltype() acc_vcl_add() acc_vcl_delete() acc_fetch_delays() acc_append_delays() acc_replace_delays() 378 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 22.5.3 ACC routines that operate on bits of a port Table 99—Operations on bits of a port To Obtain handles for bits of a module port Obtain a handle to the port from a port bit Get the port name Get the full hierarchical name Get the fulltype of a port’s bit Read Module Input Port Delay (MIPD) Append to existing MIPD Replace existing MIPD Add VCL value change callback monitors Delete VCL value change callback monitors Use acc_next_bit() acc_handle_parent() acc_fetch_name() acc_fetch_fullname() acc_fetch_fulltype() acc_fetch_delays() acc_append_delays() acc_replace_delays() acc_vcl_add() acc_vcl_delete() 22.5.4 ACC routines that operate on module paths or data paths Table 100—Operations on module paths and data paths To Obtain handles for module paths within a scope Obtain a handle to the first connected net Obtain a handle to a module path Obtain a handle to a datapath Obtain a handle to a conditional expression for a path Obtain handles for input terminals of a module path or data path Obtain handles for output terminals of a module path or data path Get the path name Get the full hierarchical name Get the polarity of a path Get the edge specified for a path terminal Read path delays Append to existing path delays Replace existing path delays Read path pulse handling Append to existing path pulse control values Replace existing path pulse control values Specify path pulse control values Free memory allocated by acc_next_input() or acc_next_output() Use acc_next_modpath() acc_handle_pathin() acc_handle_pathout() acc_handle_modpath() acc_handle_datapath() acc_handle_condition() acc_next_input() acc_next_output() acc_fetch_name() acc_fetch_fullname() acc_fetch_polarity() acc_fetch_edge() acc_fetch_delays() acc_append_delays() acc_replace_delays() acc_fetch_pulsere() acc_append_pulsere() acc_replace_pulsere() acc_set_pulsere() acc_release_object() Copyright © 2001 IEEE. All rights reserved. 379 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 22.5.5 ACC routines that operate on intermodule paths Table 101—Operations on intermodule paths To Obtain a handle for an intermodule path Get the fulltype of an intermodule path Read intermodule path delays Modify intermodule path delays Read intermodule path pulse control values Append to existing intermodule path pulse control values Replace existing intermodule path pulse control values Specify intermodule path pulse control values Use acc_handle_path() acc_fetch_fulltype() acc_fetch_delays() acc_replace_delays() acc_fetch_pulsere() acc_append_pulsere() acc_replace_pulsere() acc_set_pulsere() 22.5.6 ACC routines that operate on top-level modules Table 102—Operations on top-level modules To Obtain handles for top-level modules in a design Get the module name Use acc_next_topmod() acc_next_child() acc_fetch_name() acc_fetch_fullname() acc_fetch_defname() 22.5.7 ACC routines that operate on primitive instances Table 103—Operations on primitive instances To Obtain handles for primitive instances within a module instance Obtain a handle to the parent (the module that contains the primitive) Get the instance name Get the full hierarchical name Get the definition name Get the primitive fulltype Read delays Append to existing primitive delays Replace existing primitive delays Use acc_next_primitive() acc_handle_parent() acc_fetch_name() acc_fetch_fullname() acc_fetch_defname() acc_fetch_fulltype() acc_fetch_delays() acc_append_delays() acc_replace_delays() 380 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 22.5.8 ACC routines that operate on primitive terminals Table 104—Operations on primitive terminals To Obtain handles for terminals of a primitive instance Obtain a handle to the net connected to the terminal Obtain a handle to the parent (primitive instance containing the terminal) Get the direction (input, output, inout) Get the terminal index number Get the fulltype Add VCL value change callback monitors Delete VCL value change callback monitors Use acc_next_terminal() acc_handle_conn() acc_handle_parent() acc_fetch_direction() acc_fetch_index() acc_fetch_fulltype() acc_vcl_add() acc_vcl_delete() 22.5.9 ACC routines that operate on nets Table 105—Operations on nets To Obtain handles for nets within a module instance Obtain handles for nets within a module instance Obtain a handle to the parent (the module instance that contains the net) Determine if net is scalar, vector, collapsed, or expanded Obtain handles to bits of a vector net Obtain handles to driving terminals of the net Obtain handles to load terminals of the net Obtain handles to connected load terminals; only one per driven cell port Obtain a handle to the simulated net of a collapsed net Get the net name Get the full hierarchical name Get the net vector size Get the msb and lsb vector range Get the net fulltype Get the net logic or strength value Force or release the net logic value Add VCL value change callback monitors Delete VCL value change callback monitors Use acc_next_net() acc_next() acc_handle_parent() acc_object_of_type() acc_next_bit() acc_next_driver() acc_next_load() acc_next_cell_load() acc_handle_simulated_net() acc_fetch_name() acc_fetch_fullname() acc_fetch_size() acc_fetch_range() acc_fetch_fulltype() acc_fetch_value() acc_set_value() acc_vcl_add() acc_vcl_delete() Copyright © 2001 IEEE. All rights reserved. 381 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 22.5.10 ACC routines that operate on reg types Table 106—Operations on reg types To Obtain handles to regs within a given scope Obtain handles to bits of a vector reg Obtain a handle to the parent (module instance containing the reg) Obtain handles to load terminals of the reg Determine if reg is a scalar or a vector Get the reg name Get the full hierarchical name Get the reg size Get the msb and lsb vector range Get the reg value Set the reg value Add VCL value change callback monitors Delete VCL value change callback monitors Use acc_next() acc_next_bit() acc_handle_parent() acc_next_load() acc_object_of_type() acc_fetch_name() acc_fetch_fullname() acc_fetch_size() acc_fetch_range() acc_fetch_value() acc_set_value() acc_vcl_add() acc_vcl_delete() 22.5.11 ACC routines that operate on integer, real, and time variables Table 107—Operations on integer, real, and time variables To Obtain handles to variables within a given scope Obtain a handle to the parent (module instance containing the variable) Get the variable name Get the full hierarchical name Get the variable value Set the variable value Add VCL value change callback monitors Delete VCL value change callback monitors Use acc_next() acc_handle_parent() acc_fetch_name() acc_fetch_fullname() acc_fetch_value() acc_set_value() acc_vcl_add() acc_vcl_delete() 22.5.12 ACC routines that operate on named events Table 108—Operations on named events To Obtain handles to named events within a given scope Obtain a handle to the parent (module instance containing the named event) Get the named-event name Get the full hierarchical name Add VCL value change callback monitors Delete VCL value change callback monitors Use acc_next() acc_handle_parent() acc_fetch_name() acc_fetch_fullname() acc_vcl_add() acc_vcl_delete() 382 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 22.5.13 ACC routines that operate on parameters and specparams Table 109—Operations on parameters and specparams To Obtain handles for parameters within a module instance Obtain handles for specparams within a module instance Obtain a handle to the parent (the module instance that contains the parameter) Get the parameter or specparam name Get the full hierarchical name Get the parameter value data type (integer, floating point, string) Get the value of a parameter Get the attribute value of a parameter defined with an attribute name Use acc_next_parameter() acc_next_specparam() acc_handle_parent() acc_fetch_name() acc_fetch_fullname() acc_fetch_paramtype() acc_fetch_fulltype() acc_fetch_paramval() acc_fetch_attribute() acc_fetch_attribute_int() acc_fetch_attribute_str() 22.5.14 ACC routines that operate on timing checks Table 110—Operations on timing checks To Obtain handles for timing checks within a module instance Obtain a handle to a specific timing check Obtain handles to all timing check terminals Free memory allocated by acc_next_input() Obtain a handle to a timing check terminal Get the timing check fulltype Get a timing check limit Append to an existing timing check limit Replace to an existing timing check limit Use acc_next_tchk() acc_handle_tchk() acc_next_input() acc_release_object() acc_handle_tchkarg1() acc_handle_tchkarg2() acc_fetch_fulltype() acc_fetch_delays() acc_append_delays() acc_replace_delays() 22.5.15 ACC routines that operate on timing check terminals Table 111—Operations on timing check terminals To Obtain a handle to the net attached to timing check terminals Obtain a handle to the condition on a timing check terminal Get edge information on a timing check terminal Use acc_handle_conn() acc_handle_condition() acc_fetch_edge() Copyright © 2001 IEEE. All rights reserved. 383 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 22.5.16 ACC routines that operate on user-defined system task/function arguments Table 112—Operations on user-defined system task/function arguments To Obtain a handle for an object named in a task/function argument Get the value of a task/function argument as a double Get the value of a task/function argument as an integer Get the value of a task/function argument as a string pointer Use acc_handle_tfarg() acc_handle_itfarg() acc_fetch_tfarg() acc_fetch_itfarg() acc_fetch_tfarg_int() acc_fetch_itfarg_int() acc_fetch_tfarg_str() acc_fetch_itfarg_str() 22.6 ACC routine types and fulltypes Many objects in the Verilog HDL can have both a type and a fulltype associated with them. A type shall be a general classification of an object, whereas a fulltype shall be a specific classification. The type and fulltype for a given object can be different constants, or they can be the same constant. For example, an and logic gate has a type of accPrimitive and a fulltype of accAndPrimitive. The type and fulltype are predefined integer constants in the file acc_user.h. Several ACC routines either return a type or fulltype value, or use a type or fulltype value as an argument. Table 113 lists all type and fulltype constants that shall be supported by ACC routines, listed alphabetically by the type name. Table 113—List of all predefined type and fulltype constants type constant accConstant accDataPath accFunction accIntegerVar accModPath accModule accNamedEvent fulltype constant accConstant accDataPath accFunction accIntegerVar accModPath accModuleInstance accCellInstance accTopModule accNamedEvent Description Object is a constant Object is a data path in a path delay Object is a Verilog HDL function Object is declared as an integer data type Object is a module path Object is a module instance Object is a module instance that has been defined as a cell Object is a top-level module Object is declared as an event data type 384 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Table 113—List of all predefined type and fulltype constants (continued) type constant accNet accNetBit accOperator accParameter accPartSelect accPathTerminal accPort accPortBit fulltype constant accSupply0 accSupply1 accTri accTriand accTrior accTrireg accTri0 accTri1 accWand accWire accWor accNetBit accOperator accIntegerParam accRealParam accStringParam accPartSelect accPathInput accPathOutput accConcatPort accScalarPort accBitSelectPort accPartSelectPort accVectorPort accPortBit Description Object is declared as a supply0 net data type Object is declared as a supply1 net data type Object is declared as a tri net data type Object is declared as a triand net data type Object is declared as a trior net data type Object is declared as a trireg net data type Object is declared as a tri0 net data type Object is declared as a tri1 net data type Object is declared as a wand net data type Object is declared as a wire net data type Object is declared as a wor net data type Object is a bit-select of a net data type Object is a Verilog HDL operator Object is a parameter with an integer value Object is a parameter with a real value Object is a parameter with a string value Object is a part-select of a vector Object is an input terminal of a module path Object is an output terminal of a module path Object is a module port concatenation Object is a scalar module port Object is a bit-select of a module port (e.g.: module (.a[1], .a[0], ...); input [1:0] a; Object is a part-select of a module port (e.g.: module (.a[3:2], .a[1:0], ...); input [3:0] a; Object is a vector module port Object is a bit of a module port Copyright © 2001 IEEE. All rights reserved. 385 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Table 113—List of all predefined type and fulltype constants (continued) type constant accPrimitive accRealVar accReg accRegBit accSpecparam fulltype constant accAndGate accBufGate accBufif0Gate accBufif1Gate accCmosGate accCombPrim accNandGate accNmosGate accNorGate accNotGate accNotif0Gate accNotif1Gate accOrGate accPmosGate accPulldownGate accPullupGate accRcmosGate accRnmosGate accRpmosGate accRtranGate accRtranif0Gate accRtranif1Gate accSeqPrim accTranGate accTranif0Gate accTranif1Gate accXnorGate accXorGate accRealVar accReg accRegBit accIntegerParam accRealParam accStringParam Description Object is an and primitive Object is a buf primitive Object is a bufif0 primitive Object is a bufif1 primitive Object is a cmos primitive Object is a combinational logic UDP Object is a nand primitive Object is an nmos primitive Object is a nor primitive Object is a not primitive Object is a notif0 primitive Object is a notif1 primitive Object is an or primitive Object is a pmos primitive Object is a pulldown primitive Object is a pullup primitive Object is an rcmos primitive Object is an rnmos primitive Object is an rpmos primitive Object is an rtran primitive Object is an rtranif0 primitive Object is an rtranif1 primitive Object is a sequential logic UDP Object is a tran primitive Object is a tranif0 primitive Object is a tranif1 primitive Object is an xnor primitive Object is an xor primitive Object is declared as a real data type Object is declared as a reg data type Object is a bit-select of a reg data type Object is a specparam with an integer value Object is a specparam with a real value Object is a specparam with a string value 386 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Table 113—List of all predefined type and fulltype constants (continued) type constant accStatement accSystemTask accSystemFunction accSystemRealFunction accTask accTchk accTchkTerminal accTerminal accTimeVar accUserTask accUserFunction accUserRealFunction accWirePath fulltype constant accStatement accNamedBeginStat accNamedForkStat accSystemTask accSystemFunction accSystemRealFunction accTask accHold accNochange accPeriod accRecovery accSetup accSetuphold accSkew accWidth accTchkTerminal accInputTerminal accOutputTerminal accInoutTerminal accTimeVar accUserTask accUserFunction accUserRealFunction accIntermodPath Description Object is a procedural statement Object is a named begin statement Object is a named fork statement Object is a built-in system task Object is a built-in system function with a scalar or vector return Object is a built-in system function with a real value return Object is a Verilog HDL task Object is a $hold timing check Object is a $nochange timing check Object is a $period timing check Object is a $recovery timing check Object is a $setup timing check Object is a $setuphold timing check Object is a $skew timing check Object is a $width timing check Object is a timing check terminal Object is a primitive input terminal Object is a primitive output terminal Object is a primitive inout terminal Object is declared as a time data type Object is a user-defined system task Object is a user-defined system function with a scalar or vector return Object is a user-defined system function with a real value return Object is an intermodule path (from a module output to a module input) 22.7 Error handling When an ACC routine detects an error, it shall perform the following operations: a) Set the global error flag acc_error_flag to non-zero b) Display an error message at run time to the output channel of the software product which invoked the PLI application c) Return an exception value Copyright © 2001 IEEE. All rights reserved. 387 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® When an ACC routine is called, it automatically resets acc_error_flag to 0. 22.7.1 Suppressing error messages By default, ACC routines shall display error messages. Error messages can be suppressed using the ACC routine acc_configure() to set the configuration parameter accDisplayErrors to “false”. 22.7.2 Enabling warnings By default, ACC routines shall not display warning messages. To enable warning messages, use the ACC routine acc_configure() to set the configuration parameter accDisplayWarnings to “true”. 22.7.3 Testing for errors If automatic error reporting is suppressed, error handling can be performed by checking the acc_error_flag explicitly after calling an ACC routine. This procedure is described in Figure 53. call access routine is yes acc_error_flag set? no continue normal processing perform error processing Figure 53—Using acc_error_flag to detect errors 22.7.4 Example The following example shows a C language application that performs error checking for ACC routines. This example uses acc_configure() to suppress automatic error reporting. Instead, it checks acc_error_flag explicitly and displays its own specialized error message. 388 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C #include "acc_user.h" check_new_timing() { handle gate_handle; /* initialize and configure access routines */ acc_initialize(); /* suppress error reporting by access routines */ acc_configure( accDisplayErrors, "false" ); /* check type of first argument, the object */ gate_handle = acc_handle_tfarg( 1 ); /* check for valid argument */ if (acc_error_flag) tf_error("Cannot derive handle from argument\n"); else /* argument is valid */ /* make sure it is a primitive */ if ( acc_fetch_type(gate_handle) != accPrimitive ) tf_error("Invalid argument type:not a primitive\n"); acc_close(); } 22.7.5 Exception values ACC routines shall return one of three exception values when an error occurs, unless specified differently in the syntax of a specific ACC routine. Table 114—Exception values returned by ACC routines on errors When routine returns PLI_INT32 double values pointers or handles bool (boolean) values The exception value shall be 0 0.0 null false Because ACC routines can return valid values that are the same as exception values, the only definitive way to detect errors explicitly is to check acc_error_flag. Note that null and false are predefined constants, declared in acc_user.h. 22.8 Reading and writing delay values This section explains how ACC routines that read and modify delays are used. The ACC routines acc_fetch_delays(), acc_replace_delays(), and acc_append_delays() can read or modify delay values in a Verilog software implementation data structure. Refer to Clause 23 for the complete syntax of each of these routines. Copyright © 2001 IEEE. All rights reserved. 389 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 22.8.1 Number of delays for Verilog HDL objects There are a variety of objects in a Verilog HDL source description that can model delays. These objects can have a single delay that represents all possible logic transitions, or multiple delays that represent different logic transitions. Table 115 lists the objects that can have delays and the number of different delays for each object. Table 115—Number of possible delays for Verilog HDL objects Verilog HDL Objects 2-state primitives 3-state primitives Module paths Intermodule paths Module ports Module port bits Timing checks Number of delays Description 1 One delay for: all transitions 2 Separate delays for: rise, fall 1 One delay for: all transitions 2 Separate delays for: rise, fall 3 Separate delays for: rise, fall, toZ 1 One delay for: all transitions 2 Separate delays for: rise, fall 3 Separate delays for: rise, fall, toZ 6 Separate delays for: 0->1, 1->0, 0->Z, Z->1, 1->Z, Z->0 12 Separate delays for: 0->1, 1->0, 0->Z, Z->1, 1->Z, Z->0, 0->X, X->1, 1->X, X->0, X->Z, Z->X 1 One delay for: timing limit In addition to the number of delays, each delay can be represented as a single delay for each transition or as a minimum:typical:maximum delay set for each transition. Thus, a module path, intermodule path and module input port with 1 delay might have one value or three values, and a module path, intermodule path and module input port with 12 delays can have 12 delay values or 36 delay values. 22.8.2 ACC routine configuration The PLI shall use configuration parameters to set up the delay ACC routines to work with the variations of Verilog objects and the number of possible delays. These parameters shall be set using the routine acc_configure(). The parameters that configure the delay ACC routines are summarized in Table 116. How these configuration parameters are used is presented in 22.8.3. Refer to 23.6 for details on using acc_configure(). 390 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Table 116—Configuration parameters for delay ACC routines Configuration parameter accMinTypMaxDelays accToHiZDelay accPathDelayCount Description When “false”, each delay shall be represented by one value. When “true”, each delay shall be represented by three delay values, representing minimum, typical, maximum, respectively. The default shall be “false”. When set to “average”, “max” or “min”, the delay modify ACC routines shall calculate the toZ delay for 3-state primitives, or for path and input port objects when accPathDelayCount is set to 2. When set to “from_user”, the toZ delay shall not be calculated. The default is “from_user”. This parameter shall be ignored when accMinTypMaxDelays is set to “true”. Sets the number of delay arguments to be used by the ACC routines for module path, intermodule path and module input port delays. Shall be set to “1”, “2”, “3”, “6”, or “12”. The default shall be “6”. 22.8.3 Determining the number of arguments for ACC delay routines The ACC routines acc_fetch_delays(), acc_replace_delays(), and acc_append_delays() shall require a different number of arguments based on — The type of object handle — The setting of configuration parameters The following subsections discuss how these factors affect the number of arguments for delay ACC routines. 22.8.3.1 Single delay value mode When the configuration parameter accMinTypMaxDelays is “false” (the default), a single value shall be used for each delay transition. In this mode, the routines acc_fetch_delays(), acc_replace_delays(), and acc_append_delays() shall require each delay value as a separate argument. For acc_replace_delays() and acc_append_delays(), the arguments shall be a literal value of type double or variables of type double. For acc_fetch_delays(), the arguments shall be pointers to variables of type double. The number of arguments required is determined by the type of object handle passed to the delay ACC routine, as shown in Table 117. Table 117—Number of delay arguments in single delay mode Object handle type Timing check 2-state primitive 3-state primitive Configuration parameters accToHiZDelay set to “min”, “max”, or “average” accToHiZDelay set to “from_user” Number and order of delay arguments 1 argument: timing check limit 2 arguments: rise, fall transitions 2 arguments: rise, fall transitions (toZ delay is calculated; see 22.8.3.3) 3 arguments: rise, fall, toZ transitions Copyright © 2001 IEEE. All rights reserved. 391 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Table 117—Number of delay arguments in single delay mode (continued) Object handle type Module paths Intermodule paths Module ports Module port bits Configuration parameters accPathDelayCount set to “1” accPathDelayCount set to “2” accPathDelayCount set to “3” accPathDelayCount set to “6” accPathDelayCount set to “12” Number and order of delay arguments 1 argument: all transitions 2 arguments: rise, fall transitions 3 arguments: rise, fall, toZ transitions 6 arguments: 0->1, 1->0, 0->Z, Z->1, 1->Z, Z->0 12 arguments: 0->1, 1->0, 0->Z, Z->1, 1->Z, Z->0 0->X, X->1, 1->X, X->0, X->Z, Z->X 22.8.3.2 Min:typ:max delay value mode When the configuration parameter accMinTypMaxDelays is “true”, a three-value set shall be used for each delay transition. In this mode, the routines acc_fetch_delays(), acc_replace_delays(), and acc_append_delays() shall require the delay argument to be a pointer of an array of variables of type double. The number of elements placed into or read from the array shall be determined by the type of object handle passed to the delay ACC routine, as shown in Table 118. Table 118—Number of delay elements in min:typ:max delay mode Object handle type Timing check 2-state primitive 3-state primitive Configuration parameters Size and order of the delay array 3 elements: array[0] = min limit array[1] = typ limit array[2] = max limit 9 elements: array[0] = min rise delay array[1] = typ rise delay array[2] = max rise delay array[3] = min fall delay array[4] = typ fall delay array[5] = max fall delay array[6] = min toZ delay array[7] = typ toZ delay array[8] = max toZ delay (an array of at least 9 elements shall be declared, even if toZ delays are not used by the object) 392 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Table 118—Number of delay elements in min:typ:max delay mode (continued) Object handle type Module path Intermodule paths Module ports Module port bits Configuration parameters accPathDelayCount set to “1” accPathDelayCount set to “2” accPathDelayCount is set to “3” accPathDelayCount set to “6” Size and order of the delay array 3 elements: array[0] = min delay array[1] = typ delay array[2] = max delay 6 elements: array[0] = min rise delay array[1] = typ rise delay array[2] = max rise delay array[3] = min fall delay array[4] = typ fall delay array[5] = max fall delay 9 elements: array[0] = min rise delay array[1] = typ rise delay array[2] = max rise delay array[3] = min fall delay array[4] = typ fall delay array[5] = max fall delay array[6] = min toZ delay array[7] = typ toZ delay array[8] = max toZ delay 18 elements: array[0] = min 0->1 delay array[1] = typ 0->1 delay array[2] = max 0->1 delay array[3] = min 1->0 delay array[4] = typ 1->0 delay array[5] = max 1->0 delay array[6] = min 0->Z delay array[7] = typ 0->Z delay array[8] = max 0->Z delay array[9] = min Z->1 delay array[10] = typ Z->1 delay array[11] = max Z->1 delay array[12] = min 1->Z delay array[13] = typ 1->Z delay array[14] = max 1->Z delay array[15] = min Z->0 delay array[16] = typ Z->0 delay array[17] = max Z->0 delay Copyright © 2001 IEEE. All rights reserved. 393 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Table 118—Number of delay elements in min:typ:max delay mode (continued) Object handle type Module path (continued) Configuration parameters accPathDelayCount set to “12” Size and order of the delay array 36 elements: array[0] = min 0->1 delay array[1] = typ 0->1 delay array[2] = max 0->1 delay array[3] = min 1->0 delay array[4] = typ 1->0 delay array[5] = max 1->0 delay array[6] = min 0->Z delay array[7] = typ 0->Z delay array[8] = max 0->Z delay array[9] = min Z->1 delay array[10] = typ Z->1 delay array[11] = max Z->1 delay array[12] = min 1->Z delay array[13] = typ 1->Z delay array[14] = max 1->Z delay array[15] = min Z->0 delay array[16] = typ Z->0 delay array[17] = max Z->0 delay array[18] = min 0->X delay array[19] = typ 0->X delay array[20] = max 0->X delay array[21] = min X->1 delay array[22] = typ X->1 delay array[23] = max X->1 delay array[24] = min 1->X delay array[25] = typ 1->X delay array[26] = max 1->X delay array[27] = min X->0 delay array[28] = typ X->0 delay array[29] = max X->0 delay array[30] = min X->Z delay array[31] = typ X->Z delay array[32] = max X->Z delay array[33] = min Z->X delay array[34] = typ Z->X delay array[35] = max Z->X delay 22.8.3.3 Calculating turn-off delays from rise and fall delays In single delay mode (accMinTypMaxDelays set to “false”), the routines acc_replace_delays() and acc_append_delays() can be instructed to calculate automatically the turn-off delays from rise and fall delays. How the calculation shall be performed is controlled by the configuration parameter accToHiZDelay, as shown in Table 119. 394 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Table 119—Configuring accToHiZDelay to determine the toZ delay Configuration of accToHiZDelay “average” “min” “max” “from_user” (the default) Value of the toZ delay The toZ turn-off delay shall be the average of the rise and fall delays. The toZ turn-off delay shall be the smaller of the rise and fall delays. The toZ turn-off delay shall be the larger of the rise and fall delays. The toZ turn-off delay shall be set to the value passed as a user-supplied argument. 22.9 String handling 22.9.1 ACC routines share an internal string buffer ACC routines that return pointers to strings can share an internal buffer to store string values. These routines shall return a pointer to the location in the buffer that contains the first character of the string, as illustrated in Figure 54. In this example, mod_name points to the location in the buffer where top.m1 (the name of the module associated with module_handle) is stored. mod_name = acc_fetch_name(module_handle); THE INTERNAL STRING BUFFER ’d’ ’f’ ’f’ ’\0’ ’t’ ’o’ ’p’ ’.’ ’m’ ’1’ ’\0’ end of a previous string Figure 54—How ACC routines store strings in the internal buffer Copyright © 2001 IEEE. All rights reserved. 395 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 22.9.2 String buffer reset ACC routines shall place strings at the next available sequential location in the string buffer, which stores at least 4096 characters. If there is not enough room to store an entire string starting at the next location, a condition known as buffer reset shall occur. When buffer reset occurs, ACC routines shall place the next string starting at the beginning of the buffer, overwriting data already stored there. The result can be a loss of data, as illustrated in Figure 55. Action: mod_name = acc_fetch_fullname(module_handle); mod_name THE INTERNAL STRING BUFFER ’d’ ’f’ ’f’ ’\0’ ’t’ ’o’ ’p’ ’.’ ’m’ ’1’ ’\0’ Results: mod_name points to the string “top.m1”. The string happens to be stored near the end of the buffer. net_name = acc_fetch_fullname(net_handle); net_name mod_name THE INTERNAL STRING BUFFER ’t’ ’o’ ’p’ ’.’ ’m’ ’1’ ’.’ ’w’ ’4’ ’\0’ ’\0’ acc_fetch_fullname() cannot place the next string at the end of the buffer. Therefore, a buffer reset occurs. net_name points to the string “top.m1.w4” The data at the beginning of the buffer is overwritten; The old mod_name pointer now points to corrupted data, which in this example is “m1.w4”. Figure 55—Buffer reset causes data in the string buffer to be overwritten 22.9.2.1 The buffer reset warning ACC routines shall issue a warning whenever the internal string buffer resets. To view the warning message, the configuration parameter accDisplayWarnings shall be set to “true”, using the ACC routine acc_configure(). 396 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 22.9.3 Preserving string values Applications that use strings immediately—for example, to print names of objects—do not need to be concerned about overwrites after a string buffer reset. Applications that have to preserve string values while calling other ACC routines that write to the string buffer should preserve the string value before it is overwritten. To preserve a string value, the C routine strcpy can be used to copy the string to a local character array. 22.9.4 Example of preserving string values The following example code illustrates preserving string values. If the module in this example contains many cells, one of the calls to acc_fetch_name() could eventually overwrite the module name in the string buffer with a cell name. To preserve the module name, strcpy is used to store it locally in an array called mod_name. #include "acc_user.h" void display_cells_in_module(mod) handle mod; { handle cell; char *mod_name; PLI_BYTE8 *temp; /* save the module name in local buffer mod_name */ storage the size of the full temp = acc_fetch_fullname(mod); module name is allocated mod_name = (char*)malloc((strlen((char*)temp)+1) * sizeof(PLI_BYTE8)); strcpy(mod_name,(char *)temp); strcpy saves the full module name in mod_name cell = null; while (cell = acc_next_cell( mod, cell ) ) io_printf( "%s.%s\n", mod_name, acc_fetch_name( cell ) ); free(mod_name); } 22.10 Using VCL ACC routines The VCL routines add or delete value change monitors on a specified object. If a value change monitor is placed on an object, then whenever the object changes logic value or strength, a PLI consumer routine shall be called. The ACC routine acc_vcl_add() adds a value change monitor on an object. The arguments for acc_vcl_add() specify — A handle to an object in the Verilog HDL structure — The name of a consumer routine — A user_data value — A VCL reason_flag The following example illustrates the usage of acc_vcl_add(). acc_vcl_add(net, netmon_consumer, net_name, vcl_verilog_logic); Copyright © 2001 IEEE. All rights reserved. 397 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® The purpose of each of these arguments is described in the following paragraphs. Refer to 23.97 for the full syntax and usage of acc_vcl_add() and its arguments. The handle argument shall be a handle to any object type in the list in 22.10.1. The consumer routine argument shall be the name of a C application that shall be called for the reasons specified by the reason_flag, such as a logic value change. When a consumer routine is called, it shall be passed a pointer to a C record, called vc_record. This record shall contain information about the object, including the simulation time of the change and the new logic value of the object. The vc_record is defined in the file acc_user.h and is listed in Figure 56. The user_data argument shall be a PLI_BYTE8 pointer. The value of the user_data argument shall be passed to the consumer routine as part of the vc_record. The user_data argument can be used to pass a single value to the consumer routine, or it can be used to pass a pointer to information. For example, the name of the object could be stored in a global character string array, and a pointer to that array could be passed as the user_data argument. The consumer routine could then have access to the object name. Another example is to allocate memory for a user-defined structure with several values that need to be passed to the consumer routine. A pointer to the memory for the user-defined structure is then passed as the user_data argument. Note that the user_data argument is defined as a PLI_BYTE8 pointer; therefore, any other data type should be cast to a PLI_BYTE8 pointer. The VCL reason_flag argument is one of two predefined constants that sets up the VCL callback mechanism to call the consumer routine under specific circumstances. The constant vcl_verilog_logic sets up the VCL to call the consumer routine whenever the monitored object changes logic value. The constant vcl_verilog_strength sets up the VCL to call the consumer routine when the monitored object changes logic value or logic strength. An object can have any number of VCL monitors associated with it, as long as each monitor is unique in some way. VCL monitors can be deleted using the ACC routine acc_vcl_delete(). 22.10.1 VCL objects The VCL shall monitor value changes for the following objects: — Scalar variables and bit-selects of vector variables — Scalar nets, unexpanded vector nets, and bit-selects of expanded vector nets — Integer, real, and time variables — Module ports — Primitive output or inout terminals — Named events Note—Adding a value change link to a module port is equivalent to adding a value change link to the loconn of the port. The vc_reason returned shall be based on the loconn of the port. 22.10.2 The VCL record definition Each time a consumer routine is called, it shall be passed a pointer to a record structure called vc_record. This structure shall contain information about the most recent change that occurred on the monitored object. The vc_record structure is defined in acc_user.h and is listed in Figure 56. 398 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C typedef struct t_vc_record { PLI_INT32 vc_reason; PLI_INT32 vc_hightime; PLI_INT32 vc_lowtime; PLI_BYTE8 *user_data; union { PLI_UBYTE8 logic_value; double real_value; handle vector_handle; s_strengths strengths_s; } out_value; } s_vc_record, *p_vc_record; Figure 56—The VCL s_vc_record structure The vc_reason field of vc_record shall contain a predefined integer constant that shall describe what type of change occurred. The constants that can be passed in the vc_reason field are described in Table 120. Table 120—Predefined vc_reason constants Predefined vc_reason constant logic_value_change strength_value_change vector_value_change sregister_value_change vregister_value_change integer_value_change real_value_change time_value_change event_value_change Description A scalar net or bit-select of a vector net changed logic value. A scalar net or bit-select of a vector net changed logic value or strength. A vector net or part-select of a vector net changed logic value. A scalar reg changed logic value. A vector reg or part-select of a vector reg changed logic value. An integer variable changed value. A real variable changed value. A time variable changed value. A named event occured. The vc_hightime and vc_lowtime fields of vc_record shall be 32-bit integers that shall contain the simulation time in the simulator's time units during which the change occurred, as follows: msb vc_hightime lsb vc_lowtime 64 32 31 0 The user_data field of vc_record shall be a PLI_BYTE8 pointer, and it shall contain the value specified as the user_data argument in the acc_vcl_add() ACC routine. The out_value field of vc_record shall be a union of several data types. Only one data type shall be passed in the structure, based on the reason the callback occurred, as shown Table 121. Copyright © 2001 IEEE. All rights reserved. 399 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Table 121—Predefined out_value constants If vc_reason is logic_value_change strength_value_change vector_value_change sregister_value_change vregister_value_change integer_value_change real_value_change time_value_change event_value_change The out_value shall be a type of PLI_UBYTE8 s_strengths structure handle PLI_UBYTE8 handle handle double handle none Description A predefined constant, from the following: vcl0 vcl1 vclX vclx vclZ vclz A structure with logic and strength, as shown in Figure 57 A handle to a vector net or part-select of a vector net A predefined constant, from the following: vcl0 vcl1 vclX vclx vclZ vclz A handle to a vector reg or part-select of a vector reg A handle to an integer variable The value of a real variable A handle to a time variable Event types have no value When the vc_reason field of the vc_record is strength_value_change, the s_strengths structure fields of the out_value field of vc_record shall contain the value. This structure shall contain three fields, as shown in Figure 57. typedef struct t_strengths { PLI_UBYTE8 logic_value; PLI_UBYTE8 strength1; PLI_UBYTE8 strength2; } s_strengths, *p_strengths; Figure 57—The VCL s_strengths structure The values of the s_strengths structure fields are defined in Table 122. Table 122—Predefined out_value constants s_strengths field logic_value strength1 strength2 C data type PLI_UBYTE8 PLI_UBYTE8 Description A predefined constant, from the following: vcl0 vcl1 vclX vclx vclZ vclz A predefined constant, from the following: vclSupply vclWeak vclStrong vclMedium vclPull vclSmall vclLarge vclHighZ 400 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The strength1 and strength2 fields of the s_strengths structure can represent a) A known strength—when strength1 and strength2 contain the same value, the signal strength shall be that value. b) An ambiguous strength with a known logic_value—when strength1 and strength2 contain different values and the logic_value contains either vcl0 or vcl1, the signal strength shall be an ambiguous strength, where the strength1 value shall be the maximum possible strength and strength2 shall be the minimum possible strength. c) An ambiguous strength with an unknown logic_value—when strength1 and strength2 contain different values and the logic_value contains vclX, the signal strength shall be an ambiguous strength, where the strength1 value shall be the logic 1 component and strength2 shall be the logic 0 component. 22.10.3 Effects of acc_initialize() and acc_close() on VCL consumer routines The ACC routines acc_initialize() and acc_close() shall reset all configuration parameters set by the routine acc_configure() back to default values. Care should be taken to ensure that the VCL consumer routine does not depend on any configuration parameters, as these parameters might not have the same value when a VCL callback occurs. Refer to 23.6 on acc_configure() for a list of routines that are affected by configuration parameters. 22.10.4 An example of using VCL ACC routines The following example contains three PLI routines: a checktf application, a calltf application, and a consumer routine. The example is based on the checktf and calltf applications both being associated with two user-defined system tasks, using the PLI interface mechanism described in Clause 21. $net_monitor(,, ...); $net_monitor_off(,, ...); The checktf application, netmon_checktf, is shown below. This application performs syntax checking on instances of the user-defined system tasks to ensure there is at least one argument and that the arguments are valid net names. PLI_INT32 netmon_checktf() { int i; PLI_INT32 arg_cnt = tf_nump(); /* initialize the environment for access routines */ acc_initialize(); /* check number and type of task/function arguments */ if (arg_cnt == 0) tf_error("$net_monitor[_off] must have at least one argument"); else for (i = 1; i <= arg_cnt; i++) if (acc_fetch_type(acc_handle_tfarg(i)) != accNet) { tf_error("$net_monitor[_off] arg %d is not a net type",i); } acc_close(); return(0); } Copyright © 2001 IEEE. All rights reserved. 401 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® The calltf application, netmon_calltf, follows. This application gets a handle to each task function argument and either adds or deletes a VCL monitor on the net. The application checks the data C argument associated with each system task name to determine whether the application was called by $net_monitor or $net_monitor_off. PLI_INT32 netmon_calltf(data) PLI_INT32 data; { handle net; PLI_INT32 netmon_consumer(); PLI_INT32 tfnum; #define ADD 0 /* data value associated with $net_monitor */ #define DELETE 1 /* data value associated with $net_monitor_off */ /* initialize the environment for access routines */ acc_initialize(); switch (data) /* see which system task name called this application */ { case ADD: /* called by $net_monitor */ /* add a VCL flag to each net in the task/function argument list */ tfnum = 1; while ((net = acc_handle_tfarg(tfnum++)) != null) { /* add a VCL monitor; pass net pointer as user_data argument*/ acc_vcl_add(net, netmon_consumer, (PLI_BYTE8*)net, vcl_verilog_logic); } break; case DELETE: /* called by $net_monitor_off */ tfnum = 1; while ((net = acc_handle_tfarg(tfnum++)) != null) { /* delete the VCL monitor */ acc_vcl_delete(net, netmon_consumer, (PLI_BYTE8*)net, vcl_verilog); } break; } acc_close(); } 402 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The consumer routine, netmon_consumer, is shown in the following example. The consumer routine is called by the VCL callback mechanism. Since the checktf application only permits net data types to be used, the consumer routine only needs to check for scalar and vector net value changes when it is called. In this example, it is assumed that $net_monitor is associated with a data value of 0, and $net_monitor_off is associated with a data value of 1. Refer to 21.3.1 for a description of associating data values. PLI_INT32 netmon_consumer(vc_record) p_vc_record vc_record; /* record type passed to consumer routine */ { PLI_BYTE8 net_value; char value; handle vector_value; /* check reason VCL call-back occurred */ switch (vc_record->vc_reason) { case logic_value_change : /* scalar signal changed logic value */ { net_value = vc_record->out_value.logic_value; /* convert logic value constant to a character for printing */ switch (net_value) { case vcl0 : value = '0'; break; case vcl1 : value = '1'; break; case vclX : value = 'X'; break; case vclZ : value = 'Z'; break; } io_printf("%d : %s = %c\n", vc_record->vc_lowtime, acc_fetch_name((handle) vc_record->user_data), value); break; } case vector_value_change :/* vector signal changed logic value */ { vector_value = vc_record->out_value.vector_handle; io_printf("%d : %s = %s\n", vc_record->vc_lowtime, acc_fetch_name((handle) vc_record->user_data), acc_fetch_value(vector_value, "%b",NULL) ); break; } } } Copyright © 2001 IEEE. All rights reserved. 403 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® 23. ACC routine definitions This clause describes the PLI access (ACC) routines, explaining their function, syntax, and usage. The routines are listed in alphabetical order. The following conventions are used in the definitions of the PLI routines described in Clause 23, Clause 25, and Clause 27. Synopsis: A brief description of the PLI routine functionality, intended to be used as a quick reference when searching for PLI routines to perform specific tasks. Syntax: The exact name of the PLI routine and the order of the arguments passed to the routine. Returns: The definition of the value returned when the PLI routine is called, along with a brief description of what the value represents. The return definition contains the fields — Type: The data type of the C value that is returned. The data type is either a standard ANSI C type or a special type defined within the PLI. — Description: A brief description of what the value represents. Arguments: The definition of the arguments passed with a call to the PLI routine. The argument definition contains the fields — Type: The data type of the C values that are passed as arguments. The data type is either a standard ANSI C type, or a special type defined within the PLI. — Name: The name of the argument used in the Syntax definition. — Description: A brief description of what the value represents. All arguments shall be considered mandatory unless specifically noted in the definition of the PLI routine. Two tags are used to indicate arguments that may not be required: — Conditional: Arguments tagged as conditional shall be required only if a previous argument is set to a specific value, or if a call to another PLI routine has configured the PLI to require the arguments. The PLI routine definition explains when conditional arguments are required. — Optional: Arguments tagged as optional may have default values within the PLI, but they may be required if a previous argument is set to a specific value, or if a call to another PLI routine has configured the PLI to require the arguments. The PLI routine definition explains the default values and when optional arguments are required. Related routines: A list of PLI routines that are typically used with, or provide similar functionality to, the PLI routine being defined. This list is provided as a convenience to facilitate finding information in this standard. It is not intended to be all-inclusive, and it does not imply that the related routines have to be used. 404 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE 23.1 acc_append_delays() IEEE Std 1364-2001 Version C acc_append_delays() for single delay values (accMinTypMaxDelays set to “false”) Synopsis: Syntax: Add delays to existing delay on primitives, module paths, intermodule paths, timing checks, and module input ports. Primitives Module paths Intermodule paths Ports or port bits Timing checks Returns: Arguments: Conditional acc_append_delays(object_handle, rise_delay, fall_delay, z_delay) acc_append_delays(object_handle, d1,d2,d3,d4,d5,d6,d7,d8,d9,d10,d11,d12) acc_append_delays(object_handle, limit) Type Description PLI_INT32 1 if successful; 0 if an error occurred Type Name Description handle object_handle Handle of a primitive, module path, intermodule path, timing check, module input port or bit of a module input port double rise_delay fall_delay Rise and fall transition delay for 2-state primitives, 3-state primitives double z_delay If accToHiZDelay is set to “from_user”: turn-off (to Z) transition delay for 3-state primitives double d1 For module/intermodule paths and input ports/port bits: If accPathDelayCount is set to “1”: delay for all transitions If accPathDelayCount is set to “2” or “3”: rise transition delay If accPathDelayCount is set to “6” or “12”: 0->1 transition delay Conditional Conditional Conditional Conditional double double double double double d2 For module/intermodule paths and input ports/port bits: If accPathDelayCount is set to “2” or “3”: fall transition delay If accPathDelayCount is set to “6” or “12”: 1->0 transition delay d3 For module/intermodule paths and input ports/port bits: If accPathDelayCount is set to “3”: turn-off transition delay If accPathDelayCount is set to “6” or “12”: 0->Z transition delay d4 For module/intermodule paths and input ports/port bits: d5 If accPathDelayCount is set to “6” or “12”: d6 d4 is Z->1 transition delay d5 is 1->Z transition delay d6 is Z->0 transition delay d7 For module/intermodule paths and input ports/port bits: d8 If accPathDelayCount is set to “12”: d9 d7 is 0->X transition delay d10 d8 is X->1 transition delay d11 d9 is 1->X transition delay d12 d10 is X->0 transition delay d11 is X->Z transition delay d12 is Z->X transition delay limit Limit of timing check Copyright © 2001 IEEE. All rights reserved. 405 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® acc_append_delays() for min:typ:max delays (accMinTypMaxDelays set to “true”) Synopsis: Syntax: Add min:typ:max delay values to existing delay values for primitives, module paths, intermodule paths, timing checks or module input ports; the delay values are contained in an array. acc_append_delays(object_handle, array_ptr) Type Description Returns: PLI_INT32 Type 1 if successful; 0 if an error is encountered Name Description Arguments: handle object_handle Handle of a primitive, module path, intermodule path, timing check, module input port or bit of a module input port Related routines: double address array_ptr Pointer to array of min:typ:max delay values; the size of the array depends on the type of object and the setting of accPathDelayCount (see 22.8) Use acc_fetch_delays() to retrieve an object’s delay values Use acc_replace_delays() to replace an object’s delay values Use acc_configure() to set accPathDelayCount and accMinTypMaxDelays The ACC routine acc_append_delays() shall work differently depending on how the configuration parameter accMinTypMaxDelays is set. When this parameter is set to false, a single delay per transition shall be assumed, and delays shall be passed as individual arguments. For this single delay mode, the first syntax table in this section shall apply. When accMinTypMaxDelays is set to true, acc_append_delays() shall pass one or more sets of minimum:typical:maximum delays contained in an array, rather than single delays passed as individual arguments. For this min:typ:max delay mode, the second syntax table in this section shall apply. The number of delay values appended by acc_append_delays() shall be determined by the type of object and the setting of configuration parameters. Refer to 22.8 for a description of how the number of delay values are determined. The acc_append_delays() routine shall write delays in the timescale of the module that contains the object_handle. When altering the delay via acc_append_delays() the value of the reject/error region will not be affected unless they exceed the value of the delay. If the reject/error limits exceed the delay they will be truncated down to the new delay limit. The example shown in Figure 58 is an example of backannotation. It reads new delay values from a file called primdelay.dat and uses acc_append_delays() to add them to the current delays on a gate. The format of the file is shown in Figure 58. 406 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C . . top.m1.buf4 10.5 name of gate . . rise delay 15.0 20.7 fall delay turn-off delay #include #include "acc_user.h" PLI_INT32 write_gate_delays() { FILE *infile; PLI_BYTE8 full_gate_name[NAME_SIZE]; double rise,fall,toz; handle gate_handle; /*initialize the environment for ACC routines*/ acc_initialize(); /*read delays from file - "r" means read only*/ infile = fopen("primdelay.dat","r"); while(fscanf(infile, “%s %lf %lf %lf”, full_gate_name,rise,fall,toz) != EOF) { /*get handle for the gate*/ gate_handle = acc_handle_object(full_gate_name); /*add new delays to current values for the gate*/ acc_append_delays(gate_handle, rise, fall, toz); } acc_close(); } Figure 58—Using acc_append_delays() in single delay value mode The example shown in Figure 59 shows how to append min:typ:max delays for a 2-state primitive (no highimpedance state). The C application follows these steps: a) Declares an array of nine double-precision floating-point values to hold three sets of min:typ:max values, one set each for rising transitions, falling transitions, and transitions to Z. b) Sets the configuration parameter accMinTypMaxDelays to true to instruct acc_append_delays() to write delays in min:typ:max format. c) Calls acc_append_delays() with a valid primitive handle and the array pointer. Since the primitive to be used in this example does not have a high-impedance state, acc_append_delays() automatically appends just the rise and fall delay value sets. The last three array elements for the toZ delay values are not used. However, even though the last three array elements are not used with a 2-state primitive, the syntax for using min:typ:max delays requires that the array contain all nine elements. Copyright © 2001 IEEE. All rights reserved. 407 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® For this example, the C application, append_mintypmax_delays, is associated through the ACC interface mechanism with a user-defined system task called $appendprimdelays. A primitive with no Z state and new delay values are passed as task/function arguments to $appendprimdelays as follows: a 2-state primitive typical fall delay maximum fall delay $appendprimdelays( g1, 3.0, 5.0, 6.7, 2.4, 8.1, 9.1 ); minimum rise delay maximum typical rise delay rise delay minimum fall delay #include "acc_user.h" PLI_INT32 append_mintypmax_delays() { handle prim; double delay_array[9]; int i; delay_array has to be large enough to hold nine values to handle both 2-state primitives and 3-state primitives acc_configure(accMinTypMaxDelays, "true"); /* get handle for primitive */ prim = acc_handle_tfarg(1); /* store new delay values in array */ for (i = 0; i < 9; i++) delay_array[i] = acc_fetch_tfarg(i+2); /* append min:typ:max delays */ acc_append_delays(prim, delay_array); } Figure 59—Using acc_append_delays() in min:typ:max mode 408 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE 23.2 acc_append_pulsere() IEEE Std 1364-2001 Version C Synopsis: Syntax: Returns: Arguments: Related routines: acc_append_pulsere() Add delays to existing pulse handling reject_limit and e_limit for a module path, intermodule path or module input port. acc_append_pulsere(object,r1,e1, r2,e2, r3,e3, r4,e4, r5,e5, r6,e6, r7,e7, r8,e8, r9,e9, r10,e10, r11,e11, r12,e12) Type Description PLI_INT32 1 if successful; 0 if an error is encountered Type Name Description handle double double object r1...r12 e1...e12 Handle of module path, intermodule path or module input port reject_limit values; the number of arguments is determined by accPathDelayCount e_limit values; the number of arguments is determined by accPathDelayCount Use acc_fetch_pulsere() to get current pulse handling values Use acc_replace_pulsere() to replace existing pulse handling values Use acc_set_pulsere() to set pulse handling values as a percentage of the path delay Use acc_configure() to set accPathDelayCount The ACC routine acc_append_pulsere() shall add to an existing pulse handling reject_limit value and e_limit value for a module path, intermodule path and module input port. The reject_limit and e_limit values are used to control how pulses are propagated through paths. A pulse is defined as two transitions that occur in a shorter period of time than the delay. Pulse control values determine whether a pulse should be rejected, propagated through to the output, or considered an error. The pulse control values consist of a reject_limit and an e_limit pair of values, where — The reject_limit shall set a threshold for determining when to reject a pulse—any pulse less than the reject_limit shall not propagate. — The e_limit shall set a threshold for determining when a pulse is considered to be an error—any pulse less than the e_limit and greater than or equal to the reject_limit shall propagate a logic x. — A pulse that is greater than or equal to the e_limit shall propagate. Table 123 illustrates the relationship between the reject_limit and the e_limit. Table 123—Pulse control example When reject_limit = 10.5 e_limit = 22.6 The pulse shall be Rejected if < 10.5 An error if >= 10.5 and < 22.6 Passed if >= 22.6 Copyright © 2001 IEEE. All rights reserved. 409 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® The following rules shall apply when specifying pulse handling values: a) The value of reject_limit shall be less than or equal to the value of e_limit. b) The reject_limit and e_limit shall not be greater than the delay. If any of the limits do not meet the above rules, they shall be truncated. The number of pulse control values that acc_append_pulsere() sets shall be controlled using the ACC routine acc_configure() to set the delay count configuration parameter accPathDelayCount, as shown in Table 124. Table 124—How the value of accPathDelayCount affects acc_append_pulsere() When accPathDelayCount is “1” “2” “3” “6” (the default) “12” acc_append_pulsere() shall write One pair of reject_limit and e_limit values: one pair for all transitions, r1 and e1 Two pairs of reject_limit and e_limit values: one pair for rise transitions, r1 and e1 one pair for fall transitions, r2 and e2 Three pairs of reject_limit and e_limit values: one pair for rise transitions, r1 and e1 one pair for fall transitions, r2 and e2 one pair for turn-off transitions, r3 and e3 Six pairs of reject_limit and e_limit values—a different pair for each possible transition among 0, 1, and Z: one pair for 0->1 transitions, r1 and e1 one pair for 1->0 transitions, r2 and e2 one pair for 0->Z transitions, r3 and e3 one pair for Z->1 transitions, r4 and e4 one pair for 1->Z transitions, r5 and e5 one pair for Z->0 transitions, r6 and e6 Twelve pairs of reject_limit and e_limit values—a different pair for each possible transition among 0, 1, X, and Z: one pair for 0->1 transitions, r1 and e1 one pair for 1->0 transitions, r2 and e2 one pair for 0->Z transitions, r3 and e3 one pair for Z->1 transitions, r4 and e4 one pair for 1->Z transitions, r5 and e5 one pair for Z->0 transitions, r6 and e6 one pair for 0->X transitions, r7 and e7 one pair for X->1 transitions, r8 and e8 one pair for 1->X transitions, r9 and e9 one pair for X->0 transitions, r10 and e10 one pair for X->Z transitions, r11 and e11 one pair for Z->X transitions, r12 and e12 The minimum number of pairs of reject_limit and e_limit arguments to pass to acc_append_pulsere() has to equal the value of accPathDelayCount. Any unused reject_limit and e_limit argument pairs shall be ignored by acc_append_pulsere() and can be dropped from the argument list. If accPathDelayCount is not set explicitly, it shall default to six; therefore, six pairs of pulse reject_limit and e_limit arguments have to be passed when acc_append_pulsere() is called. Note that the value assigned to accPathDelayCount also affects acc_append_delays(), acc_fetch_delays(), acc_replace_delays(), acc_fetch_pulsere(), and acc_replace_pulsere(). Pulse control values shall be appended using the timescale of the module that contains the object handle. 410 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE 23.3 acc_close() IEEE Std 1364-2001 Version C Synopsis: Syntax: Returns: Arguments: Related routines: acc_close() Free internal memory used by ACC routines; reset all configuration parameters to default values. acc_close() Type Description void No return Type Name Description None Use acc_initialize() to initialize the ACC routine environment The ACC routine acc_close() shall free internal memory used by ACC routines and reset all configuration parameters to default values. No other ACC routines should be called after calling acc_close(); in particular, ACC routines that are affected by acc_configure() should not be called. Potentially, multiple PLI applications running in the same simulation session can interfere with each other because they share the same set of configuration parameters. To guard against application interference, both acc_initialize() and acc_close() reset all configuration parameters to their default values. The example shown in Figure 60 presents a C language routine that calls acc_close() before exiting. #include "acc_user.h" void show_versions() { /*initialize environment for ACC routines*/ acc_initialize(); /*show version of ACC routines and simulator */ io_printf("Running %s with %s\n",acc_version(),acc_product_version()); acc_close(); } Figure 60—Using acc_close() Copyright © 2001 IEEE. All rights reserved. 411 IEEE Std 1364-2001 Version C 23.4 acc_collect() IEEE STANDARD VERILOG® Synopsis: Syntax: Returns: Arguments: Related routines: acc_collect() Obtain an array of handles for all objects related to a particular reference object; get the number of objects collected. acc_collect(acc_next_routine_name, object_handle, number_of_objects) Type Description handle array An address pointer to an array of handles of the objects collected address Type Name Description pointer to acc_next_ routine acc_next_routine_name Actual name (unquoted) of the acc_next_ routine that finds the objects to be collected handle object_handle Handle of the reference object for acc_next_ routine PLI_INT32 * number_of_objects Integer pointer where the count of objects collected shall be written All acc_next_ routines except acc_next_topmod() Use acc_free() to free memory allocated by acc_collect() The ACC routine acc_collect() shall scan through a reference object, such as a module, and collect handles to all occurrences of a specific target object. The collection of handles shall be stored in an array, which can then be used by other ACC routines. The object associated with object_handle shall be a valid type of handle for the reference object required by the acc_next routine to be called. The routine acc_collect() should be used in the following situations: — To retrieve data that can be used more than once — Instead of using nested or concurrent calls to acc_next_loconn(), acc_next_hiconn(), acc_next_load(), and acc_next_cell_load() routines Otherwise, it can be more efficient to use the acc_next_ routine directly. The routine acc_collect() shall allocate memory for the array of handles it returns. When the handles are no longer needed, the memory can be freed by calling the routine acc_free(). The ACC routine acc_next_topmod() does not work with acc_collect(). However, top-level modules can be collected by passing acc_next_child() with a null reference object argument. For example: acc_collect(acc_next_child, null, &count); The example shown in Figure 61 presents a C language routine that uses acc_collect() to collect and display all nets in a module. 412 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C #include "acc_user.h" PLI_INT32 display_nets() { handle *list_of_nets, module_handle; PLI_INT32 net_count, i; /*initialize environment for ACC routines*/ acc_initialize(); /*get handle for the module*/ module_handle = acc_handle_tfarg(1); /*collect all nets in the module*/ list_of_nets = acc_collect(acc_next_net, module_handle, &net_count); /*display names of net instances*/ for(i=0; i < net_count; i++) io_printf("Net name is: %s\n", acc_fetch_name(list_of_nets[i])); /*free memory used by array list_of_nets*/ acc_free(list_of_nets); acc_close(); } Figure 61—Using acc_collect() Copyright © 2001 IEEE. All rights reserved. 413 IEEE Std 1364-2001 Version C 23.5 acc_compare_handles() IEEE STANDARD VERILOG® Synopsis: Syntax: Returns: Arguments: acc_compare_handles() Determine if two handles refer to the same object. acc_compare_handles(handle1, handle2) Type Description PLI_INT32 true if handles refer to the same object; false if different objects Type Name Description handle handle1 Handle to any object handle handle2 Handle to any object The ACC routine acc_compare_handles() shall determine if two handles refer to the same object. In some cases, two different handles might reference the same object if each handle is retrieved in a different way— for example, if an acc_next routine returns one handle and acc_handle_object() returns the other. The C == operator cannot be used to determine if two handles reference the same object. if (handle1 == handle2) /* this does not work */ The example shown in Figure 62 uses acc_compare_handles() to determine if a primitive drives the specified output of a scalar port of a module. #include "acc_user.h" PLI_INT32 prim_drives_scalar_port(prim, mod, port_num) handle prim, mod; PLI_INT32 port_num; { /* retrieve net connected to scalar port */ handle port = acc_handle_port(mod, port_num); handle port_conn = acc_next_loconn(port, null); /* retrieve net connected to primitive output */ handle out_term = acc_handle_terminal(prim, 0); handle prim_conn = acc_handle_conn(out_term); /* compare handles */ if (acc_compare_handles(port_conn, prim_conn) ) return(true); else return(false); } If port_conn and prim_conn refer to the same connection, then the prim drives port Figure 62—Using acc_compare_handles() 414 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE 23.6 acc_configure() IEEE Std 1364-2001 Version C Synopsis: Syntax: Returns: Arguments: Related routines: acc_configure() Set parameters that control the operation of various ACC routines. acc_configure(configuration_parameter, configuration_value) Type Description PLI_INT32 1 if successful; 0 if an error occurred Type Name Description integer constant quoted string For accDefaultAttr0 acc_fetch_attribute() acc_fetch_attribute_int() acc_fetch_attribute_str() For accDisplayErrors all ACC routines For accDisplayWarnings all ACC routines For accEnableArgs acc_handle_modpath() acc_handle_tchk() acc_set_scope() configuration_parameter configuration_value For accMapToMipd acc_append_delays() acc_replace_delays() For accMinTypMaxDelays acc_append_delays() acc_fetch_delays() acc_replace_delays() For accPathDelayCount acc_append_delays() acc_fetch_delays() acc_replace_delays() acc_append_pulsere() acc_fetch_pulsere() acc_replace_pulsere() One of the following predefined constants: accDefaultAttr0 accDevelopmentVersion accDisplayErrors accDisplayWarnings accEnableArgs accMapToMipd accMinTypMaxDelays accPathDelayCount accPathDelimStr accToHiZDelay One of a fixed set of string values for the configuration_parameter For accPathDelimStr acc_fetch_attribute() acc_fetch_attribute_int() acc_fetch_attribute_str() acc_fetch_fullname() acc_fetch_name() For accToHiZDelay acc_append_delays() acc_replace_delays() The ACC routine acc_configure() shall set parameters that control the operation of various ACC routines. Tables 125 through 134 describe each parameter and its set of values. Note that a call to either acc_initialize() or acc_close() shall set each configuration parameter back to its default value. Copyright © 2001 IEEE. All rights reserved. 415 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Table 125—accDefaultAttr0 configuration parameter accDefaultAttr0 Set of values “true” “false” Effect acc_fetch_attribute() shall return zero when it does not find the attribute requested and shall ignore the default_value argument acc_fetch_attribute() shall return the value passed as the default_value argument when it does not find the attribute requested Default “false” Table 126—accDevelopmentVersion configuration parameter accDevelopmentVersion Set of values Quoted string of letters, numbers, and the period character that form a valid PLI version, such as: “IEEE 1364 PLI” Effect None (can be used to document which version of ACC routines was used to develop a PLI application) Default Current version of ACC routines Software vendors can define version strings specific to their products Table 127—accDisplayErrors configuration parameter accDisplayErrors Set of values “true” “false” Effect ACC routines shall display error messages ACC routines shall not display error messages Default “true” Table 128—accDisplayWarnings configuration parameter accDisplayWarnings Set of values “true” “false” Effect ACC routines shall display warning messages ACC routines shall not display warning messages Default “false” 416 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C accEnableArgs Table 129—accEnableArgs configuration parameter Set of values “acc_handle_modpath” “no_acc_handle_modpath” “acc_handle_tchk” “no_acc_handle_tchk” “acc_set_scope” “no_acc_set_scope” Effect acc_handle_modpath() shall recognize its optional arguments acc_handle_modpath() shall ignore its optional arguments acc_handle_tchk() shall recognize its optional arguments acc_handle_tchk() shall ignore its optional arguments acc_set_scope() shall recognize its optional arguments acc_set_scope() shall ignore its optional arguments Default “no_acc_handle_modpath” “no_acc_handle_tchk” “no_acc_set_scope” Table 130—accMapToMipd configuration parameter accMapToMipd Set of values “max” “min” “latest” Effect acc_replace_delays() and acc_append_delays() shall map the longest intermodule path delay to the MIPD acc_replace_delays() and acc_append_delays() shall map the shortest intermodule path delay to the MIPD acc_replace_delays() and acc_append_delays() shall map the last intermodule path delay to the MIPD Default “max” Table 131—accMinTypMaxDelays configuration parameter accMinTypMaxDelays Set of values “true” “false” Effect acc_append_delays(), acc_fetch_delays(), acc_replace_delays(), acc_append_pulsere(), acc_fetch_pulsere(), and acc_replace_pulsere() shall use min:typ:max delay sets acc_append_delays(), acc_fetch_delays(), acc_replace_delays(), acc_append_pulsere(), acc_fetch_pulsere(), and acc_replace_pulsere() shall use a single delay value Default “false” Copyright © 2001 IEEE. All rights reserved. 417 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Table 132—accPathDelayCount configuration parameter accPathDelayCount Set of values “1” “2” “3” “6” “12” Effect acc_append_delays(), acc_fetch_delays(), acc_replace_delays(), acc_append_pulsere(), acc_fetch_pulsere(), and acc_replace_pulsere() shall use 1 delay value or value set acc_append_delays(), acc_fetch_delays(), acc_replace_delays(), acc_append_pulsere(), acc_fetch_pulsere(), and acc_replace_pulsere() shall use 2 delay values or value sets acc_append_delays(), acc_fetch_delays(), acc_replace_delays(), acc_append_pulsere(), acc_fetch_pulsere(), and acc_replace_pulsere() shall use 3 delay values or value sets acc_append_delays(), acc_fetch_delays(), acc_replace_delays(), acc_append_pulsere(), acc_fetch_pulsere(), and acc_replace_pulsere() shall use 6 delay values or value sets acc_append_delays(), acc_fetch_delays(), acc_replace_delays(), acc_append_pulsere(), acc_fetch_pulsere(), and acc_replace_pulsere() shall use 12 delay values or value sets Default “6” Table 133—accPathDelimStr configuration parameter accPathDelimStr Set of values Quoted string of letters, numbers, $ or _ Effect acc_fetch_name(), acc_fetch_fullname(), and acc_fetch_attribute() shall use the string literal as the delimiter separating the source and destination in module path names Default “$” 418 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Table 134—accToHiZDelay configuration parameter accToHiZDelay Set of values “average” “max” “min” “from_user” Effect acc_append_delays() and acc_replace_delays() shall derive turn-off delays from the average of the rise and fall delays acc_append_delays() and acc_replace_delays() shall derive turn-off delays from the larger of the rise and fall delays acc_append_delays() and acc_replace_delays() shall derive turn-off delays from the smaller of the rise and fall delays acc_append_delays() and acc_replace_delays() shall derive turn-off delays from user-supplied argument(s) Default “from_user” Copyright © 2001 IEEE. All rights reserved. 419 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® The example shown in Figure 63 presents a C language application that obtains the load capacitance of all scalar nets connected to the ports in a module. This application uses acc_configure() to direct acc_fetch_attribute() to return zero if a load capacitance is not found for a net; as a result, the third argument, default_value, can be dropped from the call to acc_fetch_attribute(). #include "acc_user.h" PLI_INT32 display_load_capacitance() { handle module_handle, port_handle, net_handle; double cap_val; /*initialize environment for ACC routines*/ acc_initialize(); /*configure acc_fetch_attribute to return 0 when it does not find*/ /* the attribute*/ acc_configure(accDefaultAttr0, "true"); /*get handle for module*/ module_handle = acc_handle_tfarg(1); /*scan all ports in module; display load capacitance*/ port_handle = null; while(port_handle = acc_next_port(module_handle, port_handle) ) { /*ports are scalar, so pass "null" to get single net connection*/ net_handle = acc_next_loconn(port_handle, null); /*since accDefaultAttr0 is "true", drop default_value argument*/ cap_val = acc_fetch_attribute(net_handle,"LoadCap_" ); if (!acc_error_flag) default_value io_printf("Load capacitance of net #%d = %1f\n", argument is dropped acc_fetch_index(port_handle), cap_val); } acc_close(); } Figure 63—Using acc_configure() to set accDefaultAttr0 420 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The example shown in Figure 64 presents a C language application that displays the name of a module path. It uses acc_configure() to set accEnableArgs and, therefore, forces acc_handle_modpath() to ignore its null name arguments and recognize its optional handle arguments, src_handle and dst_handle. #include "acc_user.h" PLI_INT32 get_path() { handle path_handle,mod_handle,src_handle,dst_handle; /*initialize the environment for ACC routines*/ acc_initialize(); /*set accEnableArgs for acc_handle_modpath*/ acc_configure(accEnableArgs, "acc_handle_modpath"); /*get handles to the three system task arguments:*/ /* arg 1 is module name */ /* arg 2 is module path source */ /* arg 3 is module path destination*/ mod_handle = acc_handle_tfarg(1); acc_handle_modpath() uses optional handle arguments src_handle and dst_handle because: src_handle = acc_handle_tfarg(2); dst_handle = acc_handle_tfarg(3); accEnableArgs is set and the name arguments are null /*display name of module path*/ path_handle = acc_handle_modpath( mod_handle, null, null, src_handle, dst_handle); io_printf("Path is %s \n", acc_fetch_fullname(path_handle) ); acc_close(); } Figure 64—Using acc_configure() to set accEnableArgs Copyright © 2001 IEEE. All rights reserved. 421 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® The example shown in Figure 65 fetches the rise and fall delays of each path in a module and backannotates the maximum delay value as the delay for all transitions. The value of accPathDelayCount specifies the minimum number of arguments that have to be passed to routines that read or write delay values. By setting accPathDelayCount to the minimum number of arguments needed for acc_fetch_delays() and again for acc_replace_delays(), all unused arguments can be eliminated from each call. #include "acc_user.h" PLI_INT32 set_path_delays() { handle mod_handle; handle path_handle; double rise_delay,fall_delay,max_delay; /*initialize environment for ACC routines*/ acc_initialize(); /*get handle to module*/ mod_handle = acc_handle_tfarg(1); /*fetch rise delays for all paths in module "top.m1"*/ path_handle = null; while(path_handle = acc_next_modpath(mod_handle, path_handle) ) { /*configure accPathDelayCount for rise and fall delays only*/ acc_configure(accPathDelayCount, "2"); only 2 delay acc_fetch_delays(path_handle, &rise_delay, &fall_delay); arguments are needed /*find the maximum of the rise and fall delays*/ max_delay = (rise_delay > fall_delay) ? rise_delay : fall_delay; /*configure accPathDelayCount to apply one delay for all transitions*/ acc_configure(accPathDelayCount, "1"); acc_replace_delays(path_handle, max_delay); } acc_close(); only 1 delay argument is needed } Figure 65—Using acc_configure() to set accPathDelayCount 422 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The example shown in Figure 66 shows how accToHiZDelay is used to direct acc_append_delays() to derive the turn-off delay for a Z-state primitive automatically as the smaller of its rise and fall delays. #include "acc_user.h" PLI_INT32 set_buf_delays() { handle primitive_handle; handle path_handle; double added_rise, added_fall; /*initialize environment for ACC routines*/ acc_initialize(); /*configure accToHiZDelay so acc_append_delays derives turn-off */ /* delay from the smaller of the rise and fall delays*/ acc_configure(accToHiZDelay, "min"); /*get handle to Z-state primitive*/ primitive_handle = acc_handle_tfarg(1); /*get delay values*/ added_rise = tf_getrealp(2); added_fall = tf_getrealp(3); acc_append_delays(primitive_handle, added_rise, added_fall); acc_close(); } Figure 66—Using acc_configure() to set accToHiZDelay Copyright © 2001 IEEE. All rights reserved. 423 IEEE Std 1364-2001 Version C 23.7 acc_count() IEEE STANDARD VERILOG® Synopsis: Syntax: Returns: Arguments: Related routines: acc_count() Count the number of objects related to a particular reference object. acc_count(acc_next_routine_name, object_handle) Type Description PLI_INT32 Number of objects Type Name Description pointer to an acc_next_ routine acc_next_routine_name Actual name (unquoted) of the acc_next_ routine that finds the objects to be counted handle object_handle Handle of the reference object for the acc_next_ routine All acc_next_ routines except acc_next_topmod() The ACC routine acc_count() shall find the number of objects that exist for a specific acc_next_ routine with a given reference object. The object associated with object_handle shall be a valid reference object for the type acc_next_ routine to be called. Note that the ACC routine acc_next_topmod() does not work with acc_count(). However, top-level modules can be counted using acc_next_child() with a null reference object argument. For example: acc_count(acc_next_child, null); The example shown in Figure 67 uses acc_count() to count the number of nets in a module. , #include "acc_user.h" PLI_INT32 count_nets() { handle module_handle; PLI_INT32 number_of_nets; /*initialize environment for ACC routines*/ acc_initialize(); /*get handle for module*/ module_handle = acc_handle_tfarg(1); /*count and display number of nets in the module*/ number_of_nets = acc_count(acc_next_net, module_handle); io_printf("number of nets = %d\n", number_of_nets); acc_close(); } Figure 67—Using acc_count() 424 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE 23.8 acc_fetch_argc() IEEE Std 1364-2001 Version C Synopsis: Syntax: Returns: Arguments: Related routines: acc_fetch_argc() Get the number of command-line arguments supplied with a Verilog software tool invocation. acc_fetch_argc() Type Description PLI_INT32 Number of command-line arguments Type Name Description None Use acc_fetch_argv() to get a character string array of the invocation options The ACC routine acc_fetch_argc() shall obtain the number of command-line arguments given on a Verilog software product invocation command line. The example shown in Figure 68 uses acc_fetch_argc() to determine the number of invocation arguments used. #include "acc_user.h" #include /* string.h is implementation dependent */ PLI_BYTE8* my_scan_plusargs(str) PLI_BYTE8 *str; { PLI_INT32 i; int length = strlen(str); PLI_BYTE8 *curStr; PLI_BYTE8 **argv = acc_fetch_argv(); for (i = acc_fetch_argc()-1; i>0; i--) { curStr = argv[i]; if ((curStr[0] == ’+’) && (!strncmp(curStr+1,str,length))) { PLI_BYTE8 *retVal; length = strlen(&(curStr[length]) + 1); retVal = (PLI_BYTE8 *)malloc(sizeof(PLI_BYTE8) * length); strcpy(retVal, &(curStr[length])); return(retVal); } } return(null); } Figure 68—Using acc_fetch_argc() Copyright © 2001 IEEE. All rights reserved. 425 IEEE Std 1364-2001 Version C 23.9 acc_fetch_argv() IEEE STANDARD VERILOG® Synopsis: Syntax: Returns: Arguments: Related routines: acc_fetch_argv() Get an array of character pointers that make up the command-line arguments for a Verilog software product invocation. acc_fetch_argv() Type Description PLI_BYTE8 ** An array of character pointers that make up the command-line arguments Type Name Description None Use acc_fetch_argc() to get a count of the number of invocation arguments The ACC routine acc_fetch_argv() shall obtain an array of character pointers that make up the commandline arguments. The format of the argv array is that each pointer in the array shall point to a NULL terminated character array which contains the string located on the tool's invocation command line. There shall be ‘argc’ entries in the argv array. The value in entry zero shall be the tool’s name. The argument following a -f argument shall contain a pointer to a NULL terminated array of pointers to characters. This new array shall contain the parsed contents of the file. The value in entry zero shall contain the name of the file. The remaining entries shall contain pointers to NULL terminated character arrays containing the different options in the file. The last entry in this array shall be a NULL. If one of the options is a -f then the next pointer shall behave the same as described above. The example shown in Figure 69 uses acc_fetch_argv() to retrieve the invocation arguments used. 426 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C #include "acc_user.h" #include /* string.h is implementation dependent */ PLI_BYTE8* my_scan_plusargs(str) PLI_BYTE8 *str; { PLI_INT32 i; int length = strlen(str); PLI_BYTE8 *curStr; PLI_BYTE8 **argv = acc_fetch_argv(); for (i = acc_fetch_argc()-1; i>0; i--) { curStr = argv[i]; if ((curStr[0] == ’+’) && (!strncmp(curStr+1,str,length))) { PLI_BYTE8 *retVal; length = strlen(&(curStr[length]) + 1); retVal = (PLI_BYTE8 *)malloc(sizeof(PLI_BYTE8) * length); strcpy(retVal, &(curStr[length])); return(retVal); } } return(null); } Figure 69—Using acc_fetch_argv() Copyright © 2001 IEEE. All rights reserved. 427 IEEE Std 1364-2001 Version C 23.10 acc_fetch_attribute() IEEE STANDARD VERILOG® Synopsis: Syntax: Returns: Arguments: Optional Related routines: acc_fetch_attribute() Get the value of a parameter or specparam named as an attribute in the Verilog source description. acc_fetch_attribute(object_handle, attribute_string, default_value) Type Description double Value of the parameter or specparam Type Name Description handle object_handle Handle of a named object quoted string or PLI_BYTE8 * attribute_string Literal string or character string pointer with the attribute portion of the parameter or specparam declaration double default_value Double-precision value to be returned if the attribute is not found (depends on accDefaultAttr0) Use acc_fetch_attribute_int() to get an attribute value as an integer Use acc_fetch_attribute_str() to get an attribute value as a string Use acc_configure(accDefaultAttr0...) to set default value returned when attribute is not found Use acc_fetch_paramtype() to get the data type of the parameter value Use acc_fetch_paramval() to get parameters or specparam values not declared in attribute/object format The ACC routine acc_fetch_attribute() shall obtain the value of a parameter or specparam that is declared as an attribute in the Verilog HDL source description. The value shall be returned as a double. Any parameter or specparam can be an attribute by naming it in one of the following ways: — As a general attribute associated with more than one object in the module where the parameter or specparam attribute is declared — As a specific attribute associated with a particular object in the module where the parameter or specparam attribute is declared Each of these methods uses its own naming convention, as described in Table 135. For either convention, attribute_string shall name the attribute and shall be passed as the second argument to acc_fetch_attribute(). The object_name shall be the actual name of a design object in a Verilog HDL source description. For A general attribute A specific attribute associated with a particular object Table 135—Naming conventions for attributes Naming convention Example attribute_string specparam DriveStrength$ = 2.8; A mnemonic name that describes the attribute attribute_string is DriveStrength$ attribute_string—object_name specparam DriveStrength$g1 = 2.8; Concatenate a mnemonic name that describes the attribute with the name of the object attribute_string is DriveStrength$ object_name is g1 428 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The ACC routine acc_fetch_attribute() shall identify module paths in terms of their sources and destinations in the following format: source path_delimiter destination The acc_fetch_attribute() routine shall look for module path names in this format, and acc_fetch_name() and acc_fetch_fullname() shall return names of module paths in this format. Therefore, the same naming convention should be used when associating an attribute with a module path. Note that names of module paths with multiple sources or destinations shall be derived from the first source or destination only. By default, the path_delimiter used in path names is the “$” character. This default can be changed by using the ACC routine acc_configure() to set the delimiter parameter accPathDelimStr to another character string. The examples in Table 136 show how to name module paths using different delimiter strings. Table 136—Example module path names using delimiter strings For module path (a => q) = 10; (b *> q1,q2) = 8; (d,e,f *> r,s)= 8; If accPathDelimStr is “$” “_$_” “_” Then the module path name is a$q b_$_q1 d_r The following example shows an attribute name for a particular module path object: Given the module path: An attribute name is: (a => q) = 10; specparam RiseStrength$a$q = 20; In this example, the attribute_string is RiseStrength$, the object_name is a$q, and the path_delimiter is $ (the default path delimiter). Copyright © 2001 IEEE. All rights reserved. 429 IEEE Std 1364-2001 Version C The following flowchart illustrates how acc_fetch_attribute() shall work: IEEE STANDARD VERILOG® search for attribute 1) associated with specified object found? yes return attribute’s value as a double-precision floating-point number no search for attribute 2) without an associated object yes found? no 3) return default value This flowchart shows that when acc_fetch_attribute() finds the attribute requested, it returns the value of the attribute as a double-precision floating-point number. 1) The routine shall first look for the attribute name that concatenates attribute_string with the name associated with object_handle. For example, to find an attribute InputLoad$ for a net n1, acc_fetch_attribute() would search for InputLoad$n1. 2) If acc_fetch_attribute() does not find the attribute associated with the object specified with object_handle, the routine shall then search for a name that matches attribute_string. Assume that, in the previous example, acc_fetch_attribute() does not find InputLoad$n1. It would then look for InputLoad$. Other variants of that name, such as InputLoad$n3 or InputLoad$n, shall not be considered matches. 3) Failing both search attempts, the routine acc_fetch_attribute() shall return a default value. The default value is controlled by using the ACC routine acc_configure() to set or reset the configuration parameter accDefaultAttr0 as shown in Table 137. Table 137—Controlling the default value returned by acc_fetch_attribute() When accDefaultAttr0 is “true” “false” acc_fetch_attribute() shall return Zero when the attribute is not found; the default_value argument can be dropped The value passed as the default_value argument when the attribute is not found The example shown in Figure 70 presents a C language application that uses acc_fetch_attribute() to obtain the load capacitance of all scalar nets connected to the ports in a module. Note that acc_fetch_attribute() does not require its third argument, default_value, because acc_configure() is used to set accDefaultAttr0 to true. 430 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C #include "acc_user.h" PLI_INT32 display_load_capacitance() { handle module_handle, port_handle, net_handle; double cap_val; /*initialize environment for ACC routines*/ acc_initialize(); /*configure acc_fetch_attribute to return 0 when it does not find*/ /*the attribute*/ acc_configure(accDefaultAttr0, "true"); /*get handle for module*/ module_handle = acc_handle_tfarg(1); /*scan all ports in module; display load capacitance*/ port_handle = null; while(port_handle = acc_next_port(module_handle, port_handle) ) { /*ports are scalar, so pass "null" to get single net connection*/ net_handle = acc_next_loconn(port_handle, null); /*since accDefaultAttr0 is "true", drop default_value argument*/ cap_val = acc_fetch_attribute(net_handle,"LoadCap_"); if (!acc_error_flag) io_printf("Load capacitance of net #%d = %1f\n", acc_fetch_index(port_handle), cap_val); } acc_close(); } Figure 70—Using acc_fetch_attribute() Copyright © 2001 IEEE. All rights reserved. 431 IEEE Std 1364-2001 Version C 23.11 acc_fetch_attribute_int() IEEE STANDARD VERILOG® Synopsis: Syntax: Returns: Arguments: Optional Related routines: acc_fetch_attribute_int() Get the integer value of a parameter or specparam named as an attribute in the Verilog source description. acc_fetch_attribute_int(object_handle, attribute_string, default_value) Type Description PLI_INT32 Value of the parameter or specparam Type Name Description handle object_handle Handle of a named object quoted string or PLI_BYTE8 * attribute_string Literal string or character string pointer with the attribute portion of the parameter or specparam declaration PLI_INT32 default_value Integer value to be returned if the attribute is not found (depends on accDefaultAttr0) Use acc_fetch_attribute() to get an attribute value as a double Use acc_fetch_attribute_str() to get an attribute value as a string Use acc_configure(accDefaultAttr0...) to set default value returned when attribute is not found Use acc_fetch_paramtype() to get the data type of the parameter value Use acc_fetch_paramval() to get parameters or specparam values not declared in attribute/object format The ACC routine acc_fetch_attribute_int() shall obtain the value of a parameter or specparam that is declared as an attribute in the Verilog HDL source description. The value shall be returned as an integer. Any parameter or specparam can be an attribute. Refer to 23.10 for a description of attribute naming and how attribute values are fetched. 432 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE 23.12 acc_fetch_attribute_str() IEEE Std 1364-2001 Version C Synopsis: Syntax: Returns: Arguments: Optional Related routines: acc_fetch_attribute_str() Get the value of a parameter or specparam named as an attribute in the Verilog source description. acc_fetch_attribute_str(object_handle, attribute_string, default_value) Type Description PLI_BYTE8 * Value of the parameter or specparam Type Name Description handle object_handle Handle of a named object quoted string or PLI_BYTE8 * attribute_string Literal string or character string pointer with the attribute portion of the parameter or specparam declaration quoted string or PLI_BYTE8 * default_value Character string value to be returned if the attribute is not found (depends on accDefaultAttr0) Use acc_fetch_attribute() to get an attribute value as a double Use acc_fetch_attribute_int() to get an attribute value as an integer Use acc_configure(accDefaultAttr0...) to set default value returned when attribute is not found Use acc_fetch_paramtype() to get the data type of the parameter value Use acc_fetch_paramval() to get parameters or specparam values not declared in attribute/object format The ACC routine acc_fetch_attribute_str() shall obtain the value of a parameter or specparam that is declared as an attribute in the Verilog HDL source description. The value shall be returned as a pointer to a character string. The return value for this routine is placed in the ACC internal string buffer. See 22.9 for explanation of strings in ACC routines. Any parameter or specparam can be an attribute. Refer to 23.10 for a description of attribute naming and how attribute values are fetched. Copyright © 2001 IEEE. All rights reserved. 433 IEEE Std 1364-2001 Version C 23.13 acc_fetch_defname() IEEE STANDARD VERILOG® Synopsis: Syntax: Returns: Arguments: Related routines acc_fetch_defname() Get the definition name of a module instance or primitive instance. acc_fetch_defname(object_handle) Type Description PLI_BYTE8 * Pointer to a character string containing the definition name Type Name Description handle object_handle Handle of the module instance or primitive instance Use acc_fetch_name() to display the instance name of an object The ACC routine acc_fetch_defname() shall obtain the definition name of a module instance or primitive instance. The definition name is the declared name of the object as opposed to the instance name of the object. In the illustration shown below, the definition name is “dff”, and the instance name is “i15”. definition name dff i15 (q, clk, d); //instance of a module or primitive instance name The return value for this routine is placed in the ACC internal string buffer. See 22.9 for explanation of strings in ACC routines. The example shown in Figure 71 presents a C language application that uses acc_fetch_defname() to display the definition names of all primitives in a module. #include "acc_user.h" void get_primitive_definitions(module_handle) handle module_handle; { handle prim_handle; /*get and display defining names of all primitives in the module*/ prim_handle = null; while(prim_handle = acc_next_primitive(module_handle,prim_handle)) io_printf("primitive definition is %s\n", acc_fetch_defname(prim_handle) ); } Figure 71—Using acc_fetch_defname() 434 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE 23.14 acc_fetch_delay_mode() IEEE Std 1364-2001 Version C Synopsis: Syntax: Returns: Arguments: acc_fetch_delay_mode() Get the delay mode of a module instance. acc_fetch_delay_mode(module_handle) Type Description PLI_INT32 A predefined integer constant representing the delay mode of the module instance: accDelayModeNone accDelayModeZero accDelayModeUnit accDelayModePath accDelayModeDistrib accDelayModeMTM Type Name Description handle module_handle Handle to a module instance The ACC routine acc_fetch_delay_mode() shall return the delay mode of a module or cell instance. The delay mode determines how delays are stored for primitives and paths within the module or cell. The routine shall return one of the predefined constants given in Table 138. Table 138—Predefined constants used by acc_fetch_delay_mode() Predefined constant accDelayModeNone accDelayModeZero accDelayModeUnit accDelayModeDistrib accDelayModePath accDelayModeMTM Description No delay mode specified. All primitive delays are zero; all path delays are ignored. All primitive delays are one; all path delays are ignored. If a logical path has both primitive delays and path delays specified, the primitive delays shall be used. If a logical path has both primitive delays and path delays specified, the path delays shall be used. If this property is true, Minimum:Typical:Maximum delay sets for each transition are being stored; if this property is false, a single delay for each transition is being stored. Copyright © 2001 IEEE. All rights reserved. 435 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Figure 72 uses acc_fetch_delay_mode() to retrieve the delay mode of all children of a specified module. #include "acc_user.h" PLI_INT32 display_delay_mode() { handle mod, child; /*reset environment for ACC routines*/ acc_initialize(); /*get module passed to user-defined system task*/ mod = acc_handle_tfarg(1); /*find and display delay mode for each module instance*/ child = null; while(child = acc_next_child(mod, child)) { io_printf("Module %s set to: ",acc_fetch_fullname(child)); switch(acc_fetch_delay_mode(child) ) { case accDelayModePath: io_printf(" path delay mode\n"); break; case accDelayModeDistrib: io_printf(" distributed delay mode\n"); break; ... } } } Figure 72—Using acc_fetch_delay_mode() 436 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE 23.15 acc_fetch_delays() IEEE Std 1364-2001 Version C acc_fetch_delays() for single delay values (accMinTypMaxDelays set to “false”) Synopsis: Syntax: Get existing delays for primitives, module paths, timing checks, module input ports, and intermodule paths. Primitives acc_fetch_delays(object_handle, rise_delay, fall_delay, z_delay) Module paths Intermodule paths Ports or port bits acc_fetch_delays(object_handle, d1,d2,d3,d4,d5,d6,d7,d8,d9,d10,d11,d12) Timing checks Returns: acc_fetch_delays(object_check_handle, limit) Type Description PLI_INT32 1 if successful; 0 if an error occurred Type Name Description Arguments: handle object_handle Handle of a primitive, module path, timing check, module input port, bit of a module input port, or intermodule path double * rise_delay fall_delay Rise and fall delay for 2-state primitive or 3-state primitive Conditional double * z_delay Turn-off (to Z) transition delay for 3-state primitives double * d1 For module/intermodule paths and input ports/port bits: If accPathDelayCount is set to “1”: delay for all transitions If accPathDelayCount is set to “2” or “3”: rise transition delay If accPathDelayCount is set to “6” or “12”: 0->1 transition delay Conditional double * d2 For module/intermodule paths and input ports/port bits: If accPathDelayCount is set to “2” or “3”: fall transition delay If accPathDelayCount is set to “6” or “12”: 1->0 transition delay Conditional double * d3 For module/intermodule paths and input ports/port bits: If accPathDelayCount is set to “3”: turn-off transition delay If accPathDelayCount is set to “6” or “12”: 0->Z transition delay Conditional double * d4 For module/intermodule paths and input ports/port bits: d5 If accPathDelayCount is set to “6” or “12”: d6 d4 is Z->1 transition delay d5 is 1->Z transition delay d6 is Z->0 transition delay Conditional double * d7 For module/intermodule paths and input ports/port bits: d8 If accPathDelayCount is set to “12”: d9 d7 is 0->X transition delay d10 d8 is X->1 transition delay d11 d9 is 1->X transition delay d12 d10 is X->0 transition delay d11 is X->Z transition delay d12 is Z->X transition delay double * limit Limit of timing check Copyright © 2001 IEEE. All rights reserved. 437 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Synopsis: Syntax: Returns: Arguments: acc_fetch_delays() for min:typ:max delays (accMinTypMaxDelays set to “true”) Get existing delay values for primitives, module paths, timing checks, module input ports, or intermodule paths; the delay values are contained in an array. acc_fetch_delays(object_handle, array_ptr), Type Description PLI_INT32 Type 1 if successful; 0 if an error is encountered Name Description handle double address object_handle array_ptr Handle of a primitive, module path, timing check, module input port, bit of a module input port, or intermodule path Pointer to array of min:typ:max delay values; the size of the array depends on the type of object and the setting of accPathDelayCount (see 22.8) The ACC routine acc_fetch_delays() shall work differently depending on how the configuration parameter accMinTypMaxDelays is set. When this parameter is set to “false”, a single delay per transition shall be assumed, and each delay shall be fetched into variables pointed to as individual arguments. For this single delay mode, the first syntax table in this section shall apply. When accMinTypMaxDelays is set to “true”, acc_fetch_delays() shall fetch one or more sets of minimum:typical:maximum delays into an array, rather than single delays fetched as individual arguments. For this min:typ:max delay mode, the second syntax table in this section shall apply. The number of delay values that shall be fetched by acc_fetch_delays() is determined by the type of object and the setting of configuration parameters. Refer to 22.8 for a description of how the number of delay values is determined. The ACC routine acc_fetch_delays() shall retrieve delays in the timescale of the module that contains the object_handle. The example shown in Figure 73 presents a C language application that uses acc_fetch_delays() to retrieve the rise, fall, and turn-off delays of all paths through a module. 438 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C #include "acc_user.h" void display_path_delays() { handle mod_handle; handle path_handle; double rise_delay,fall_delay,toz_delay; /*initialize environment for ACC routines*/ acc_initialize(); /*set accPathDelayCount to return rise, fall and turn-off delays */ acc_configure(accPathDelayCount, "3"); /*get handle to module*/ mod_handle = acc_handle_tfarg(1); /*fetch rise delays for all paths in module "top.m1"*/ path_handle = null; while(path_handle = acc_next_modpath(mod_handle, path_handle) ) { acc_fetch_delays(path_handle, &rise_delay,&fall_delay,&toz_delay); /*display rise, fall and turn-off delays for each path*/ io_printf("For module path %s,delays are:\n", acc_fetch_fullname(path_handle) ); io_printf("rise = %lf, fall = %lf, turn-off = %lf\n", rise_delay,fall_delay,toz_delay); } acc_close(); } Figure 73—Using acc_fetch_delays() in single delay mode The example shown in Figure 74 is a C language code fragment of an application that shows how to fetch min:typ:max delays for the intermodule paths. The example follows these steps: a) Declares an array of nine double-precision floating-point values as a buffer for storing three sets of min:typ:max values, one set each for rise, fall, and turn-off delays. b) Sets the configuration parameter accMinTypMaxDelays to “true” to instruct acc_fetch_delays() to retrieve delays in min:typ:max format. c) Calls acc_fetch_delays() with a valid intermodule path handle and the array pointer. Copyright © 2001 IEEE. All rights reserved. 439 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® #include "acc_user.h" void fetch_mintypmax_delays(port_output, port_input) handle port_output, port_input; { . . . handle intermod_path; double delay_array[9]; . . . acc_configure(accMinTypMaxDelays, "true"); acc_handle_path returns a handle to a wire path that represents the connection from an output (or inout) port to an input (or inout) port . . . intermod_path = acc_handle_path(port_output, port_input); acc_fetch_delays(intermod_path, delay_array); . . . } acc_fetch_delays places the following values in delay_array: delay_array[0] = delay_array[1] = delay_array[2] = min:typ:max rise delay delay_array[3] = delay_array[4] = delay_array[5] = min:typ:max fall delay delay_array[6] = delay_array[7] = delay_array[8] = min:typ:max turn-off delay Figure 74—Using acc_fetch_delays() in min:typ:max delay mode 440 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE 23.16 acc_fetch_direction() IEEE Std 1364-2001 Version C Synopsis: Syntax: Returns: Arguments: acc_fetch_direction() Get the direction of a port or terminal. acc_fetch_direction(object_handle) Type Description PLI_INT32 A predefined integer constant representing the direction of a port or terminal accInput accOutput accInout accMixedIo Type Name Description handle object_handle Handle of a port or terminal The ACC routine acc_fetch_direction() shall return a predefined integer constant indicating the direction of a module port or primitive terminal. The values returned are given in Table 139. Table 139—The operation of acc_fetch_direction() When direction is Input only Output only Bidirectional (input and output) A concatenation of input ports and output ports acc_fetch_direction() shall return accInput accOutput accInout accMixedIo The example shown in Figure 75 presents a C language application that uses acc_fetch_direction() to determine whether or not a port is an input. #include "acc_user.h" int is_port_input(port_handle) handle port_handle; { PLI_INT32 direction; direction = acc_fetch_direction(port_handle); if (direction == accInput || direction == accInout) return(true); else return(false); } Figure 75—Using acc_fetch_direction() Copyright © 2001 IEEE. All rights reserved. 441 IEEE Std 1364-2001 Version C 23.17 acc_fetch_edge() IEEE STANDARD VERILOG® Synopsis: Syntax: Returns: Arguments: acc_fetch_edge() Get the edge specifier of a module path or timing check terminal. acc_fetch_edge(pathio_handle) Type Description PLI_INT32 A predefined integer constant representing the edge specifier of a path input or output ter- minal: accNoedge accEdge01 accEdgex1 accPosedge accEdge10 accEdge1x accNegedge accEdge0x accEdgex0 Type Name Description handle pathio_handle Handle to a module path input or output, or handle to a timing check terminal The ACC routine acc_fetch_edge() shall return a value that is a masked integer representing the edge specifier for a module path or timing check terminal. Table 140 lists the predefined edge specifiers as they are specified in acc_user.h. Table 140—Edge specifiers constants Edge type None Positive edge (0→1,0→x,x→1) Negative edge (1→0,1→x,x→0) 0→1 edge 1→0 edge 0→x edge x→1 edge 1→x edge x→0 edge Defined constant accNoedge accPosedge accNegedge accEdge01 accEdge10 accEdge0x accEdgex1 accEdge1x accEdgex0 Binary value 0 00001101 00110010 00000001 00000010 00000100 00001000 00010000 00100000 The integer mask returned by acc_fetch_edge() is usually either accPosedge or accNegedge. Occasionally, however, the mask is a hybrid mix of specifiers that is equal to neither. The example shown in Figure 76 illustrates how to check for these hybrid edge specifiers. The value accNoEdge is returned if no edge is found. The example takes a path input or output and returns the string corresponding to its edge specifier. It provides analogous functionality to that of acc_fetch_type_str() in that it returns a string corresponding to an integer value that represents a type. 442 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C This example first checks to see whether the returned mask is equal to accPosedge or accNegedge, which are the most likely cases. If it is not, the application does a bitwise AND with the returned mask and each of the other edge specifiers to find out which types of edges it contains. If an edge type is encoded in the returned mask, the corresponding edge type string suffix is appended to the string “accEdge”. PLI_BYTE8 *acc_fetch_edge_str(pathio) handle pathio; { PLI_INT32 edge = acc_fetch_edge(pathio); static PLI_BYTE8 edge_str[32]; if (! acc_error_flag) { if (edge == accNoEdge) strcpy(edge_str, "accNoEdge"); /* accPosedge == (accEdge01 & accEdge0x & accEdgex1) */ else if (edge == accPosEdge) strcpy(edge_str, "accPosEdge"); /* accNegedge == (accEdge10 & accEdge 1x & accEdgex0) */ else if (edge == accNegEdge) strcpy(edge_str, "accNegEdge"); /* edge is neither posedge nor negedge, but some combination of other edges */ else { strcpy(edge_str, "accEdge"); if (edge & accEdge01) strcat(edge_str, "_01"); if (edge & accEdge10) strcat(edge_str, "_10"); if (edge & accEdge0x) strcat(edge_str, "_0x"); if (edge & accEdgex1) strcat(edge_str, "_x1"); if (edge & accEdge1x) strcat(edge_str, "_1x"); if (edge & accEdgex0) strcat(edge_str, "_x0"); } return(edge_str); } else return(null); } Figure 76—Using acc_fetch_edge() Copyright © 2001 IEEE. All rights reserved. 443 IEEE Std 1364-2001 Version C 23.18 acc_fetch_fullname() IEEE STANDARD VERILOG® Synopsis: Syntax: Returns: Arguments: Related routines: acc_fetch_fullname() Get the full hierarchical name of any named object or module path. acc_fetch_fullname(object_handle) Type Description PLI_BYTE8 * Character pointer to a string containing the full hierarchical name of the object Type Name Description handle object_handle Handle of the object Use acc_fetch_name() to find the lowest-level name of the object Use acc_configure(accPathDelimStr...) to set the delimiter string for module path names The ACC routine acc_fetch_fullname() shall obtain the full hierarchical name of an object. The full hierarchical name is the name that uniquely identifies an object. In Figure 77, the top-level module, top1, contains module instance mod3, which contains net w4. In this example, the full hierarchical name of the net is top1.mod3.w4. top1 mod3 w4 Figure 77—A design hierarchy; the fullname of net w4 is “top1.mod3.w4” Table 141 lists the objects in a Verilog HDL description for which acc_fetch_fullname() shall return a name. Table 141—Named objects supported by acc_fetch_fullname() Modules Module ports Module paths Data paths Primitives Nets Regs or Variables Integer, time and real variables Named events Parameters Specparams Named blocks Verilog HDL tasks Verilog HDL functions 444 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Module path names shall be derived from their sources and destinations in the following format: source path_delimiter destination By default, the path_delimiter shall be the character $. However, the delimiter can be changed by using the ACC routine acc_configure() to set the delimiter parameter accPathDelimStr to another character string. The following examples show names of paths within a top-level module m3, as returned by acc_fetch_fullname() when the path_delimiter is $. Note that names of module paths with multiple sources or destinations shall be derived from the first source and destination only. Table 142—Module path names returned by acc_fetch_fullname() For paths in module m3 (a => q) = 10; (b *> q1,q2) = 8; (d,e,f *> r,s)= 8; acc_fetch_fullname() returns a pointer to m3.a$q m3.b$q1 m3.d$r If a Verilog software product creates default names for unnamed instances, acc_fetch_fullname() shall return the full hierarchical default name. Otherwise, the routine shall return null for unnamed instances. Using acc_fetch_fullname() with a module port handle shall return the full hierarchical implicit name of the port. The routine acc_fetch_fullname() shall store the returned string in a temporary buffer. To preserve the string for later use in an application, it should be copied to another variable (refer to 22.9). In the example shown in Figure 78, the routine uses acc_fetch_fullname() to display the full hierarchical name of an object if the object is a net. #include "acc_user.h" PLI_INT32 display_if_net(object_handle) handle object_handle; { /*get and display full name if object is a net*/ if (acc_fetch_type(object_handle) == accNet) io_printf("Object is a net: %s\n", acc_fetch_fullname(object_handle) ); else io_printf("Object is not a net: %s\n", acc_fetch_fullname(object_handle) ); } Figure 78—Using acc_fetch_fullname() Copyright © 2001 IEEE. All rights reserved. 445 IEEE Std 1364-2001 Version C 23.19 acc_fetch_fulltype() IEEE STANDARD VERILOG® Synopsis: Syntax: Returns: Arguments: Related routines: acc_fetch_fulltype() Get the fulltype of an object. acc_fetch_fulltype(object_handle) Type Description PLI_INT32 A predefined integer constant from the list shown in 22.6 Type Name Description handle object_handle Handle of the object Use acc_fetch_type() to get the general type classification of an object Use acc_fetch_type_str() to get the fulltype as a character string The ACC routine acc_fetch_fulltype() shall return the fulltype of an object. The fulltype is a specific classification of a Verilog HDL object, represented as a predefined constant (defined in acc_user.h). Table 113 lists all of the fulltype constants that can be returned by acc_fetch_fulltype(). Many Verilog HDL objects have both a type and a fulltype. The type of an object is its general Verilog HDL type classification. The fulltype is the specific type of the object. The examples in Table 143 illustrate the difference between the type of an object and the fulltype of the same object for selected objects. Table 143—The difference between the type and the fulltype of an object For a handle to A setup timing check An and gate primitive A sequential UDP acc_fetch_type() shall return accTchk accPrimitive accPrimitive acc_fetch_fulltype() shall return accSetup accAndGate accSeqPrim 446 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The example shown in Figure 79 uses acc_fetch_fulltype() to find and display the fulltypes of timing checks. This application is called by a higher-level application, display_object_type, presented as the usage example for acc_fetch_type(). #include "acc_user.h" PLI_INT32 display_timing_check_type(tchk_handle) handle tchk_handle; { /*display timing check type*/ io_printf("Timing check is"); switch(acc_fetch_fulltype(tchk_handle) ) { case accHold: io_printf(" hold\n"); break; case accNochange: io_printf(" nochange\n"); break; case accPeriod: io_printf(" period\n"); break; case accRecovery: io_printf(" recovery\n"); break; case accSetup: io_printf(" setup\n"); break; case accSkew: io_printf(" skew\n"); break; case accWidth: io_printf(" width\n"); } } Figure 79—Using acc_fetch_fulltype() to display the fulltypes of timing checks Copyright © 2001 IEEE. All rights reserved. 447 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® The example shown in Figure 80 uses acc_fetch_fulltype() to find and display the fulltypes of primitive objects passed as input arguments. This application is called by a higher-level application, display_object_type, presented as the usage example for acc_fetch_type(). #include "acc_user.h" PLI_INT32 display_primitive_type(primitive_handle) handle primitive_handle; { /*display primitive type*/ io_printf("Primitive is"); switch(acc_fetch_fulltype(primitive_handle) ) { case accAndGate: io_printf(" and gate\n"); break; case accBufGate: io_printf(" buf gate\n"); break; case accBufif0Gate:case accBufif1Gate: io_printf(" bufif gate\n"); break; case accCmosGate:case accNmosGate:case accPmosGate: case accRcmosGate:case accRnmosGate:case accRpmosGate: io_printf(" MOS or Cmos gate\n"); break; case accCombPrim: io_printf(" combinational UDP\n"); break; case accSeqPrim: io_printf(" sequential UDP\n"); break; case accNotif0Gate:case accNotif1Gate: io_printf(" notif gate\n"); break; case accRtranGate: io_printf(" rtran gate\n"); break; case accRtranif0Gate:case accRtranif1Gate: io_printf(" rtranif gate\n"); break; case accNandGate: io_printf(" nand gate\n"); break; case accNorGate: io_printf(" nor gate\n"); break; case accNotGate: io_printf(" not gate\n"); break; case accOrGate: io_printf(" or gate\n"); break; case accPulldownGate: io_printf(" pulldown gate\n"); break; case accPullupGate: io_printf(" pullup gate\n"); break; case accXnorGate: io_printf(" xnor gate\n"); break; case accXorGate: io_printf(" xor gate\n"); } } Figure 80—Using acc_fetch_fulltype() to display the fulltypes of primitives 448 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE 23.20 acc_fetch_index() IEEE Std 1364-2001 Version C Synopsis: Syntax: Returns: Arguments: acc_fetch_index() Get the index number for a port or terminal. acc_fetch_index(object_handle) Type Description PLI_INT32 Integer index for a port or terminal, starting with zero Type Name Description handle object_handle Handle of the port or terminal The ACC routine acc_fetch_index() shall return the index number for a module port or primitive terminal. Indices are integers that shall start at zero and increase from left to right. — The index of a port shall be its position in a module definition in the Verilog HDL source description. — The index of a terminal shall be its position in a gate, switch, or UDP instance. Table 144 shows how indices shall be derived. Table 144—Deriving indices For Terminals: nand g1(out, in1, in2); Implicit ports: module A(q, a, b); Explicit ports: module top; reg ra,rb; wire wq; explicit_port_mod epm1(.b(rb), .a(ra), .q(wq)); endmodule module explicit_port_mod(q, a, b); input a, b; output q; nand (q, a, b); endmodule Indices are 0 for terminal out 1 for terminal in1 2 for terminal in2 0 for port q 1 for port a 2 for port b 0 for explicit port epm1.q 1 for explicit port epm1.a 2 for explicit port epm1.b The example shown in Figure 81 presents a C language application that uses acc_fetch_index() to find and display the input ports of a module. Copyright © 2001 IEEE. All rights reserved. 449 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® #include "acc_user.h" PLI_INT32 display_inputs(module_handle) handle module_handle; { handle port_handle; PLI_INT32 direction; /*get handle for the module and each of its ports*/ port_handle = null; while (port_handle = acc_next_port(module_handle, port_handle) ) { /*determine if port is an input*/ direction = acc_fetch_direction(port_handle); /*give the index of each input port*/ if (direction == accInput) io_printf("Port #%d of %s is an input\n", acc_fetch_index(port_handle), acc_fetch_fullname(module_handle) ); } } Figure 81—Using acc_fetch_index() 450 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE 23.21 acc_fetch_location() IEEE Std 1364-2001 Version C Synopsis: Syntax: Returns: Arguments: acc_fetch_location() Get the location of an object in a Verilog-HDL source file. acc_fetch_location(loc_p, object_handle) Type Description PLI_INT32 1 if successful; 0 if an error is encountered Type Name Description p_location loc_p Pointer to a predefined location structure handle object_handle Handle to an object The ACC routine acc_fetch_location() shall return the file name and line number in the file for the specified object. The file name and line number shall be returned in an s_location data structure. This data structure is defined in acc_user.h, and listed in Figure 82. typedef struct t_location { PLI_INT32 line_no; PLI_BYTE8 *filename; } s_location, *p_location; Figure 82—s_location data structure filename field is a character pointer. line_no field is a nonzero positive integer. Table 145 lists the objects that shall be supported by acc_fetch_location(). Table 145—Objects supported by acc_fetch_location() Object type Modules Module ports Module paths Data paths Primitives Explicit nets Implicit nets Reg variables Integer, time and real variables Location returned Module instantiation line Module definition Module path line Module path line Instantiation line Definition line Line where first used Definition line Definition line Copyright © 2001 IEEE. All rights reserved. 451 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Table 145—Objects supported by acc_fetch_location() (continued) Object type Named events Parameters Specparams Named blocks Verilog HDL tasks Verilog HDL functions Location returned Definition line Definition line Definition line Definition line Definition line Definition line The return value for filename is placed in the ACC internal string buffer. See 22.9 for an explanation of strings in ACC routines. The example shown in Figure 83 uses acc_fetch_location() to print the file name and line number for an object. PLI_INT32 find_object_location (object) handle object; { s_location s_loc; p_location loc_p = &s_loc; acc_fetch_location(loc_p, object); /*get the filename and line_no*/ if (! acc_error_flag) /* On success */ io_printf (“Object located in file %s on line %d \n”, loc_p->filename, loc_p->line_no); } Figure 83—Using acc_fetch_location() 452 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE 23.22 acc_fetch_name() IEEE Std 1364-2001 Version C Synopsis: Syntax: Returns: Arguments: Related routines: acc_fetch_name() Get the instance name of any named object or module path. acc_fetch_name(object_handle) Type Description PLI_BYTE8 * Character pointer to a string containing the instance name of the object Type Name Description handle object_handle Handle of the named object Use acc_fetch_fullname() to get the full hierarchical name of the object Use acc_fetch_defname() to get the definition name of the object Use acc_configure(accPathDelimStr...) to set the naming convention for module paths The ACC routine acc_fetch_name() shall obtain the name of an object. The name of an object is its lowestlevel name. In the following example, the top-level module, top1, contains module instance mod3, which contains net w4, as shown in Figure 84. In this example, the name of the net is w4. top1 mod3 w4 Figure 84—A design hierarchy; the name of net w4 is “w4” The return value for this routine is placed in the ACC internal string buffer. See 22.9 for an explanation of strings in ACC routines. Table 146 lists the objects in a Verilog HDL description for which acc_fetch_name() shall return a name. Table 146—Named objects supported by acc_fetch_name() Modules Module ports Module paths Data paths Primitives Nets Regs or Variables Integer, time and real variables Named events Parameters Specparams Named blocks Verilog HDL tasks Verilog HDL functions Copyright © 2001 IEEE. All rights reserved. 453 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Module path names shall be derived from their sources and destinations in the following format: source path_delimiter destination By default, the path_delimiter is the character $. However, the delimiter can be changed by using the ACC routine acc_configure() to set the delimiter parameter accPathDelimStr to another character string. Table 147 shows names of paths within a top-level module m3, as returned by acc_fetch_name() when the path_delimiter is $. Note that names of module paths with multiple sources or destinations shall be derived from the first source and destination only. Table 147—Module path names returned by acc_fetch_name() For paths in module m3 (a => q) = 10; (b *> q1,q2) = 8; (d,e,f *> r,s)= 8; acc_fetch_name() returns a pointer to a$q b$q1 d$r If a Verilog software implementation creates default names for unnamed instances, acc_fetch_name() shall return the default name. Otherwise, the routine shall return null for unnamed instances. Using acc_fetch_name() with a module port handle shall return the implicit name of the port. The following example uses acc_fetch_name() to display the names of top-level modules. #include "acc_user.h" PLI_INT32 show_top_mods() { handle module_handle; /*initialize environment for ACC routines*/ acc_initialize(); /*scan all top-level modules*/ io_printf("The top-level modules are:\n"); module_handle = null; while (module_handle = acc_next_topmod(module_handle) ) io_printf(" %s\n",acc_fetch_name(module_handle)); acc_close(); } Figure 85—Using acc_fetch_name() 454 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE 23.23 acc_fetch_paramtype() IEEE Std 1364-2001 Version C Synopsis: Syntax: Returns: Arguments: Related routines: acc_fetch_paramtype() Get the data type of a parameter or specparam. acc_fetch_paramtype(parameter_handle) Type Description PLI_INT32 A predefined integer constant representing the data type of a parameter: accIntParam accIntegerParam accRealParam accStringParam Type Name Description handle parameter_handle Handle to a parameter or specparam Use acc_next_parameter() to get all parameters within a module Use acc_next_specparam() to get all specparams within a module The ACC routine acc_fetch_paramtype() shall return an integer constant that represents the data type of a value that has been assigned to a parameter or specparam. Figure 86 uses acc_fetch_paramtype() to display the values of all parameters within a module. #include "acc_user.h" PLI_INT32 print_parameter_values() { handle module_handle, param_handle; /*initialize environment for ACC routines*/ acc_initialize(); module_handle = acc_handle_tfarg(1); param_handle = null; while(param_handle = acc_next_parameter(module_handle,param_handle)) { io_printf("Parameter %s has value: ", acc_fetch_fullname(param_handle)); switch(acc_fetch_paramtype(param_handle) ) { case accRealParam: io_printf("%lf\n", acc_fetch_paramval(param_handle) ); break; case accIntegerParam: io_printf("%d\n", (int)acc_fetch_paramval(param_handle) ); break; case accStringParam: io_printf("%s\n", (char*)(int)acc_fetch_paramval(param_handle) ); break; } } acc_close(); } Figure 86—Using acc_fetch_paramtype() Copyright © 2001 IEEE. All rights reserved. 455 IEEE Std 1364-2001 Version C 23.24 acc_fetch_paramval() IEEE STANDARD VERILOG® Synopsis: Syntax: Returns: Arguments: Related routines: acc_fetch_paramval() Get the value of a parameter or specparam. acc_fetch_paramval(parameter_handle) Type Description double The value of a parameter or specparam Type Name Description handle parameter_handle Handle to a parameter or specparam Use acc_fetch_paramtype() to retrieve the data type of a parameter Use acc_next_parameter() to scan all parameters within a module Use acc_next_specparam() to scan all specparams within a module The ACC routine acc_fetch_paramval() shall return the value stored in a parameter or specparam. The value shall be returned as a double-precision floating-point number. A parameter value can be stored as one of three data types: — A double-precision floating-point number — An integer value — A string Therefore, it can be necessary to call acc_fetch_paramtype() to determine the data type of the parameter value, as shown in the example in Figure 87. The routine acc_fetch_paramval() returns values as type double. The values can be converted back to integers or character pointers using the C language cast mechanism, as shown in Table 148. Note that some C language compilers do not allow casting a double-precision value directly to a character pointer; it is therefore necessary to use a two-step cast to first convert the double value to an integer and then convert the integer to a character pointer. If a character string is returned, it is placed in the ACC internal string buffer. See 22.9 for explanation of strings in ACC routines. Table 148—Casting acc_fetch_paramval() return values To convert to Integer String Follow these steps Cast the return value to the integer data type using the C language cast operator (int): int_val= (int) acc_fetch_paramval(...); Cast the return value to a character pointer using the C language cast operators (char*)(int): str_ptr= (char*)(int) acc_fetch_paramval(...); 456 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The example shown in Figure 87 presents a C language application, print_parameter_values, that uses acc_fetch_paramval() to display the values of all parameters within a module. #include "acc_user.h" PLI_INT32 print_parameter_values() { handle module_handle; handle param_handle; /*initialize environment for ACC routines*/ acc_initialize(); /*get handle for module*/ module_handle = acc_handle_tfarg(1); /*scan all parameters in the module and display their values*/ /* according to type*/ param_handle = null; while(param_handle = acc_next_parameter(module_handle,param_handle)) { io_printf("Parameter %s has value:", acc_fetch_fullname(param_handle)); switch(acc_fetch_paramtype(param_handle) ) { case accRealParam: io_printf("%lf\n", acc_fetch_paramval(param_handle) ); break; case accIntegerParam: io_printf("%d\n", (int)acc_fetch_paramval(param_handle) ); break; case accStringParam: io_printf("%s\n", (char*)(int)acc_fetch_paramval(param_handle) ); break; } } acc_close(); } two-step cast Figure 87—Using acc_fetch_paramval() Copyright © 2001 IEEE. All rights reserved. 457 IEEE Std 1364-2001 Version C 23.25 acc_fetch_polarity() IEEE STANDARD VERILOG® Synopsis: Syntax: Returns: Arguments: acc_fetch_polarity() Get the polarity of a path. acc_fetch_polarity(path_handle) Type Description PLI_INT32 A predefined integer constant representing the polarity of a path: accPositive accNegative accUnknown Type Name Description handle path_handle Handle to a module path or data path The ACC routine acc_fetch_polarity() shall return an integer constant that represents the polarity of the specified path. The polarity of a path describes how a signal transition at its source propagates to its destination in the absence of logic simulation events. The return value shall be one of the predefined integer constant polarity types listed in Table 149. Table 149—Polarity types returned by acc_fetch_polarity() Integer constant accPositive accNegative accUnknown Description A rise at the source causes a rise at the destination. A fall at the source causes a fall at the destination. A rise at the source causes a fall at the destination. A fall at the source causes a rise at the destination. Unpredictable; a rise or fall at the source causes either a rise or fall at the destination. The example shown in Figure 88 takes a path argument and returns the string corresponding to its polarity. PLI_BYTE8 *fetch_polarity_str(path) { switch (acc_fetch_polarity(path)) { case accPositive: return(“accPositive”); case accNegative: return(“accNegative”); case accUnknown: return(“accUnknown”); default: return(null); } } Figure 88—Using acc_fetch_polarity() 458 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE 23.26 acc_fetch_precision() IEEE Std 1364-2001 Version C Synopsis: Syntax: Returns: Arguments: Related routines: acc_fetch_precision() Get the smallest time precision argument specified in all `timescale compiler directives in a given design. acc_fetch_precision() Type Description PLI_INT32 Type An integer value that represents a time precision Name Description None Use acc_fetch_timescale_info() to get the timescale and precision of a specific object The ACC routine acc_fetch_precision() shall return the smallest time precision argument specified in all `timescale compiler directives for a given design. The value returned shall be the order of magnitude of one second, as shown in Table 150. Table 150—Value returned by acc_fetch_precision() Integer value returned 2 1 0 -1 -2 -3 -4 -5 -6 -7 -8 -9 -10 -11 -12 -13 -14 -15 Simulation time precision represented 100 s 10 s 1s 100 ms 10 ms 1 ms 100 us 10 us 1 us 100 ns 10 ns 1 ns 100 ps 10 ps 1 ps 100 fs 10 fs 1 fs If there are no `timescale compiler directives specified for a design, acc_fetch_precision() shall return a value of 0 (1 s). Copyright © 2001 IEEE. All rights reserved. 459 IEEE Std 1364-2001 Version C 23.27 acc_fetch_pulsere() IEEE STANDARD VERILOG® Synopsis: Syntax: Returns: Arguments: Related routines: acc_fetch_pulsere() Get current pulse handling reject_limit and e_limit for a module path, intermodule path or module input port. acc_fetch_pulsere(object,r1,e1, r2,e2, r3,e3, r4,e4, r5,e5, r6,e6, r7,e7, r8,e8, r9,e9, r10,e10, r11,e11, r12,e12) Type Description PLI_INT32 Type 1 if successful; 0 if an error is encountered Name Description handle double * double * object r1...r12 e1...e12 Handle of module path, intermodule path or module input port reject_limit values; the number of arguments is determined by accPathDelayCount e_limit values; the number of arguments is determined by accPathDelayCount Use acc_append_pulsere() to add to the existing pulse handling values Use acc_replace_pulsere() to replace existing pulse handling values Use acc_set_pulsere() to set pulse handling values as a percentage of the path delay Use acc_configure() to set accPathDelayCount The ACC routine acc_fetch_pulsere() shall obtain the current values controlling how pulses are propagated through a module path, intermodule path or module input port. A pulse is defined as two transitions that occur in a shorter period of time than the delay. Pulse control values determine whether a pulse should be rejected, propagated through to the output, or considered an error. The pulse control values consist of a reject_limit and an e_limit pair of values, where — The reject_limit shall set a threshold for determining when to reject a pulse—any pulse less than the reject_limit shall not propagate — The e_limit shall set a threshold for determining when a pulse is an error—any pulse less than the e_limit and greater than or equal to the reject_limit shall propagate a logic x — A pulse that is greater than or equal to the e_limit shall propagate — Table 151 illustrates the relationship between the reject_limit and the e_limit. Table 151—Pulse control example When reject_limit = 10.5 e_limit = 22.6 The pulse shall be Rejected if < 10.5 An error if >= 10.5 and < 22.6 Passed if >= 22.6 The number of pulse control values that acc_fetch_pulsere() shall retrieve is controlled using the ACC routine acc_configure() to set the delay count configuration parameter accPathDelayCount, as shown in Table 152. 460 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C Table 152—How the accPathDelayCount affects acc_fetch_pulsere() When accPathDelayCount is “1” “2” “3” “6” (the default) “12” acc_fetch_pulsere() shall retrieve One pair of reject_limit and e_limit values: one pair for all transitions, r1 and e1 Two pairs of reject_limit and e_limit values: one pair for rise transitions, r1 and e1 one pair for fall transitions, r2 and e2 Three pairs of reject_limit and e_limit values: one pair for rise transitions, r1 and e1 one pair for fall transitions, r2 and e2 one pair for turn-off transitions, r3 and e3 Six pairs of reject_limit and e_limit values—a different pair for each possible transition among 0, 1, and Z: one pair for 0->1 transitions, r1 and e1 one pair for 1->0 transitions, r2 and e2 one pair for 0->Z transitions, r3 and e3 one pair for Z->1 transitions, r4 and e4 one pair for 1->Z transitions, r5 and e5 one pair for Z->0 transitions, r6 and e6 Twelve pairs of reject_limit and e_limit values—a different pair for each possible transition among 0, 1, X, and Z: one pair for 0->1 transitions, r1 and e1 one pair for 1->0 transitions, r2 and e2 one pair for 0->Z transitions, r3 and e3 one pair for Z->1 transitions, r4 and e4 one pair for 1->Z transitions, r5 and e5 one pair for Z->0 transitions, r6 and e6 one pair for 0->X transitions, r7 and e7 one pair for X->1 transitions, r8 and e8 one pair for 1->X transitions, r9 and e9 one pair for X->0 transitions, r10 and e10 one pair for X->Z transitions, r11 and e11 one pair for Z->X transitions, r12 and e12 The minimum number of pairs of reject_limit and e_limit arguments to pass to acc_fetch_pulsere() shall equal the value of accPathDelayCount. Any unused reject_limit and e_limit argument pairs shall be ignored by acc_fetch_pulsere() and can be dropped from the argument list. If accPathDelayCount is not set explicitly, it shall default to 6, and therefore six pairs of pulse reject_limit and e_limit arguments have to be used when acc_fetch_pulsere() is called. Note that the value assigned to accPathDelayCount also affects acc_append_delays(), acc_fetch_delays(), acc_replace_delays(), acc_append_pulsere(), and acc_replace_pulsere(). Pulse control values shall be retrieved using the timescale of the module that contains the object handle. The example shown in Figure 89 shows how an application, get_pulsevals, uses acc_fetch_pulsere() to retrieve rise and fall pulse handling values of paths listed in a file called path.dat. The format of the file is shown in the following diagram. Copyright © 2001 IEEE. All rights reserved. 461 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® • • path source top.m1 in out • name of module • path destination #include #include "acc_user.h" #define NAME_SIZE 256 PLI_INT32 get_pulsevals() { FILE *infile; PLI_BYTE8 mod_name[NAME_SIZE]; PLI_BYTE8 pathin_name[NAME_SIZE], pathout_name[NAME_SIZE]; handle mod, path; double rise_reject_limit,rise_e_limit,fall_reject_limit,fall_e_limit; /*initialize environment for ACC routines*/ acc_initialize(); /* set accPathDelayCount to return two pairs of pulse handling */ /* values, one each for rise and fall transitions */ acc_configure(accPathDelayCount, "2"); /*read all module path specifications from file "path.dat"*/ infile = fopen("path.dat", "r"); while(fscanf(infile, "%s %s %s", mod_name,pathin_name,pathout_name)!=EOF) { mod=acc_handle_object(mod_name); path=acc_handle_modpath(mod,pathin_name,pathout_name); if(acc_fetch_pulsere(path, &rise_reject_limit,&rise_e_limit, &fall_reject_limit, &fall_e_limit)) { io_printf("rise reject limit = %lf, rise e limit = %lf\n", rise_reject_limit, rise_e_limit); io_printf("fall reject limit = %lf, fall e limit = %lf\n", fall_reject_limit, fall_e_limit); } } acc_close(); } Figure 89—Using acc_fetch_pulsere() 462 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE 23.28 acc_fetch_range() IEEE Std 1364-2001 Version C Synopsis: Syntax: Returns: Arguments: Related routines acc_fetch_range() Get the most significant bit and least significant bit range values for a vector. acc_fetch_range(vector_handle, msb, lsb) Type Description PLI_INT32 Zero if successful; nonzero upon error Type Name Description handle vector_handle Handle to a vector net or reg PLI_INT32 * msb Pointer to an integer variable to hold the most significant bit of vector_handle PLI_INT32 * lsb Pointer to an integer variable to hold the least significant bit of vector_handle Use acc_fetch_size() to get the number of bits in a vector The ACC routine acc_fetch_range() shall obtain the most significant bit (msb) and least significant bit (lsb) numbers of a vector. The msb shall be the left range element, while the lsb shall be the right range element in the Verilog HDL source code. The example shown in Figure 90 takes a handle to a module instance as its input. It then uses acc_fetch_range() to display the name and range of each vector net found in the module as: [:]. PLI_INT32 display_vector_nets() { handle mod = acc_handle_tfarg(1); handle net; PLI_INT32 msb, lsb; io_printf (“Vector nets in module %s:\n”, acc_fetch_fullname (mod)); net = null; while (net = acc_next_net(mod, net)) if (acc_object_of_type(net, accVector)) { acc_fetch_range(net, &msb, &lsb); io_printf(“ %s[%d:%d]\n”, acc_fetch_name(net), msb, lsb); } } Figure 90—Using acc_fetch_range() Copyright © 2001 IEEE. All rights reserved. 463 IEEE Std 1364-2001 Version C 23.29 acc_fetch_size() IEEE STANDARD VERILOG® Synopsis: Syntax: Returns: Arguments: acc_fetch_size() Get the bit size of a net, reg, integer, time, real or port. acc_fetch_size(object_handle) Type Description PLI_INT32 Number of bits in the net, reg, integer, time, real or port Type Name Description handle object_handle Handle to a net, reg, integer, time, real or port, or a bitselect or part select thereof The ACC routine acc_fetch_size() shall return the number of bits of a net, reg, integer, time, real or port. The example shown in Figure 91 uses acc_fetch_size() to display the size of a vector net. #include "acc_user.h" PLI_INT32 display_vector_size() { handle net_handle; PLI_INT32 size_in_bits; /* reset environment for ACC routines */ acc_initialize(); /*get first argument passed to user-defined system task*/ /* associated with this routine*/ net_handle = acc_handle_tfarg(1); /*if net is a vector, find and display its size in bits*/ if (acc_object_of_type(net_handle, accVector) ) { size_in_bits = acc_fetch_size(net_handle); io_printf("Net %s is a vector of size %d\n", acc_fetch_fullname(net_handle),size_in_bits); } else io_printf("Net %s is not a vector net\n", acc_fetch_fullname(net_handle) ); } Figure 91—Using acc_fetch_size() 464 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE 23.30 acc_fetch_tfarg(), acc_fetch_itfarg() IEEE Std 1364-2001 Version C Synopsis: Syntax: Returns: Arguments: Related routines: acc_fetch_tfarg(), acc_fetch_itfarg() Get the value of the specified argument of the system task or function associated with the PLI application; the value is returned as a double-precision number. acc_fetch_tfarg(argument_number) acc_fetch_itfarg(argument_number, tfinst) Type Description double The value of the task/function argument, returned as a double-precision number Type Name Description PLI_INT32 argument_number Integer number that references the system task or function argument by its position in the argument list handle tfinst Handle to a specific instance of a user-defined system task or function Use acc_fetch_tfarg_int() or acc_fetch_itfarg_int() to get the task/function argument value as an integer Use acc_fetch_tfarg_str() or acc_fetch_itfarg_str() to get the task/function argument value as a string Use acc_handle_tfinst() to get a handle to a specific instance of a user-defined system task or function The ACC routine acc_fetch_tfarg() shall return the value of arguments passed to the current instance of a user-defined system task or function. The ACC routine acc_fetch_itfarg() shall return the value of arguments passed to a specific instance of a user-defined system task or function, using a handle to the task or function. The value is returned as a double-precision floating-point number. Argument numbers shall start at one and increase from left to right in the order that they appear in the system task or function call. If an argument number is passed in that is out of range for the number of arguments in the user-defined system task/function call, acc_fetch_tfarg() and acc_fetch_itfarg() shall return a value of 0.0, and generate a warning message if warnings are enabled. Note that the acc_error_flag is not set for an out-of-range index number. If a user-defined system task/function argument that does not represent a valued object is referenced, acc_fetch_tfarg() and acc_fetch_itfarg() shall return a value of 0.0 and generate a warning message if warnings are enabled. Literal numbers, nets, regs, integer variables, and real variables all have values. Objects such as module instance names do not have a value. Note that the acc_error_flag is not set when a nonvalued argument is referenced. The routine acc_fetch_tfarg() returns values as type double. The routines acc_fetch_tfarg_int() and acc_fetch_tfarg_str() return values as integers or string pointers, respectively. The value returned by acc_fetch_tfarg() can also be converted to integers or character pointers using the C language cast mechanism, as shown in Table 153. Note that some C language compilers do not allow casting a double-precision value directly to a character pointer; it is therefore necessary to use a two-step cast to first convert the double value to an integer and then convert the integer to a character pointer. If a character string is returned, it is placed in the ACC internal string buffer. See 22.9 for explanation of strings in ACC routines. Copyright © 2001 IEEE. All rights reserved. 465 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® Table 153—Casting acc_fetch_tfarg() return values To convert to Integer String Follow these steps Cast the return value to the integer data type using the C language cast operator (PLI_INT32): int_val= (PLI_INT32) acc_fetch_tfarg(...); Cast the return value to a character pointer using the C language cast operators (char*)(int): str_ptr= (char*)(int) acc_fetch_tfarg(...); The example shown in Figure 92 uses acc_fetch_tfarg(), acc_fetch_tfarg_int(), and acc_fetch_tfarg_str() to return the value of the first argument of a user-defined system task or function. #include "acc_user.h" #include "veriuser.h" PLI_INT32 display_arg_value() { PLI_INT32 arg_type; /*initialize environment for ACC routines*/ acc_initialize(); /*check type of argument*/ io_printf("Argument value is "); switch(tf_typep(1) ) { case tf_readonlyreal: case tf_readwritereal: io_printf("%1f\n", acc_fetch_tfarg(1) ); break; case tf_readonly: returns value as a double-precision floating-point number case tf_readwrite: io_printf("%d\n", acc_fetch_tfarg_int(1) ); break; case tf_string: returns value as an integer number io_printf("%s\n", acc_fetch_tfarg_str(1) ); break; default: returns value as a pointer to a character string io_printf("Error in argument specification\n"); break; } acc_close(); } Figure 92—Using acc_fetch_tfarg(), acc_fetch_tfarg_int(), and acc_fetch_tfarg_str() 466 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE 23.31 acc_fetch_tfarg_int(), acc_fetch_itfarg_int() IEEE Std 1364-2001 Version C Synopsis: Syntax: Returns: Arguments: Related routines: acc_fetch_tfarg_int(), acc_fetch_itfarg_int() Get the value of the specified argument of the system task or function associated with the PLI application; the value is returned as an integer number. acc_fetch_tfarg_int(argument_number) acc_fetch_itfarg_int(argument_number, tfinst) Type Description PLI_INT32 The value of the task/function argument, returned as an integer number Type Name Description PLI_INT32 argument_number Integer number that references the system task or function argument by its position in the argument list handle tfinst Handle to a specific instance of a user-defined system task or function Use acc_fetch_tfarg() or acc_fetch_itfarg() to get the task/function argument value as a double Use acc_fetch_tfarg_str() or acc_fetch_itfarg_str() to get the task/function argument value as a string Use acc_handle_tfinst() to get a handle to a specific instance of a user-defined system task or function The ACC routine acc_fetch_tfarg_int() shall return the value of arguments passed to the current userdefined system task or function. The ACC routine acc_fetch_itfarg_int() shall return the value of arguments passed to a specific instance of a user-defined system task and function, using a handle to the task or function. The value is returned as an integer number. Argument numbers shall start at one and increase from left to right in the order that they appear in the system task or function call. If an argument number is passed in that is out of range for the number of arguments in the user-defined system task/function call, acc_fetch_tfarg_int() and acc_fetch_itfarg_int() shall return a value of 0 and generate a warning message if warnings are enabled. Note that the acc_error_flag is not set for an out-ofrange index number. If a user-defined system task/function argument that does not represent a valued object is referenced, acc_fetch_tfarg_int() and acc_fetch_itfarg_int() shall return a value of 0 and generate a warning message if warnings are enabled. Literal numbers, nets, regs, integer variables, and real variables all have values. Objects such as module instance names do not have a value. Note that the acc_error_flag is not set when a nonvalued argument is referenced. If a user-defined task/function argument is a real value, the value is cast to a PLI_INT32 and returned as an integer. If the task/function argument is a string value, the string is copied into the ACC string buffer and the pointer to the string is cast to the type PLI_INT32 and returned as an integer. Refer to Figure 92 for an example of using acc_fetch_tfarg_int(). Copyright © 2001 IEEE. All rights reserved. 467 IEEE Std 1364-2001 Version C 23.32 acc_fetch_tfarg_str(), acc_fetch_itfarg_str() IEEE STANDARD VERILOG® Synopsis: Syntax: Returns: Arguments: Related routines: acc_fetch_tfarg_str(), acc_fetch_itfarg_str() Get the value of the specified argument of the system task or function associated with the PLI application; the value is returned as a pointer to a character string. acc_fetch_tfarg_str(argument_number) acc_fetch_itfarg_str(argument_number, tfinst) Type Description PLI_BYTE8 * The value of the task/function argument, returned as a pointer to a character string Type Name Description PLI_INT32 argument_number Integer number that references the system task or function argument by its position in the argument list handle tfinst Handle to a specific instance of a user-defined system task or function Use acc_fetch_tfarg() or acc_fetch_itfarg() to get the task/function argument value as a double Use acc_fetch_tfarg_int() or acc_fetch_itfarg_int() to get the task/function argument value as an integer Use acc_handle_tfinst() to get a handle to a specific instance of a user-defined system task or function The ACC routine acc_fetch_tfarg_str() shall return the value of arguments passed to the current instance of a user-defined system task or function. The ACC routine acc_fetch_itfarg_str() shall return the value of arguments passed to a specific instance of a user-defined system task or function, using a handle to the task or function. The value shall be returned as a pointer to a character string. The return value for this routine is placed in the ACC internal string buffer. See 22.9 for explanation of strings in ACC routines. Argument numbers shall start at one and increase from left to right in the order that they appear in the system task or function call. If an argument number is passed in that is out of range for the number of arguments in the user-defined system task/function call, acc_fetch_tfarg_str() and acc_fetch_itfarg_str() shall return a value of null and generate a warning message if warnings are enabled. Note that the acc_error_flag is not set for an outof-range index number. If a user-defined system task/function argument that does not represent a valued object is referenced, acc_fetch_tfarg_str() and acc_fetch_itfarg_str() shall return a value of null and generate a warning message if warnings are enabled. Literal numbers, nets, regs, integer variables, and real variables all have values. Objects such as module instance names do not have a value. Note that the acc_error_flag is not set when a nonvalued argument is referenced. If a user-defined task/function argument is a value, each 8 bits of the value are converted into its equivalent ASCII character. Refer to Figure 92 for an example of using acc_fetch_tfarg_str(). 468 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE 23.33 acc_fetch_timescale_info() IEEE Std 1364-2001 Version C Synopsis: Syntax: Returns: Arguments: Related routines: acc_fetch_timescale_info() Get timescale information for an object or for an active $timeformat system task invocation. acc_fetch_timescale_info(object_handle, timescale_p) Type Description void Type Name Description handle object_handle Handle of a module instance, module definition, PLI userdefined system task/function call, or null p_timescale_info timescale_p Pointer to a variable defined as a s_timescale_info structure Use acc_fetch_precision() to fetch the smallest timescale precision in a design The ACC routine acc_fetch_timescale_info() shall obtain the timescale information for an object or for an active $timeformat built-in system task invocation. The timescale returned shall be based on the type of object handle, as defined in Table 154. Table 154—Return values from acc_fetch_timescale_info() If the object_handle is A handle to a module instance or module definition A handle to a user-defined system task or function null acc_fetch_timescale_info() shall return The timescale for the corresponding module definition The timescale for the corresponding module definition that represents the parent module instance of the object The timescale for an active $timeformat system task invocation The routine acc_fetch_timescale_info() shall return a value to an s_timescale_info structure pointed to by the timescale_p argument. This structure is declared in the file acc_user.h, as shown in Figure 93. typedef struct t_timescale_info { PLI_INT16 unit; PLI_INT16 precision; } s_timescale_info, *p_timescale_info; Figure 93—s_timescale_info data structure — The term unit is a short integer that shall represent the timescale unit in all cases of object — The term precision is a short integer that shall represent the timescale precision. In the case of a null object handle, precision shall be the number of decimal points specified in the active $timeformat system task invocation. Copyright © 2001 IEEE. All rights reserved. 469 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® The value returned for unit and precision shall be the order of magnitude of 1 s, as shown in Table 155. Table 155—Value returned by acc_fetch_timescale_info() Integer value returned 2 1 0 -1 -2 -3 -4 -5 -6 -7 -8 -9 -10 -11 -12 -13 -14 -15 Time unit r 100 s 10 s 1s 100 ms 10 ms 1 ms 100 us 10 us 1 us 100 ns 10 ns 1 ns 100 ps 10 ps 1 ps 100 fs 10 fs 1 fs For example, a call to acc_fetch_timescale_info(obj, timescale_p) Where obj is defined in a module that has `timescale 1us/1ns specified for its definition, shall return timescale_p->unit: -6 timescale_p->precision: -9 470 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE 23.34 acc_fetch_type() IEEE Std 1364-2001 Version C Synopsis: Syntax: Returns: Arguments: Related routines: acc_fetch_type() Get the type of an object. acc_fetch_type(object_handle) Type Description PLI_INT32 A predefined integer constant from the list shown in 22.6 Type Name Description handle object_handle Handle of the object Use acc_fetch_fulltype() to get the full type classification of an object Use acc_fetch_type_str() to get the type as a character string The ACC routine acc_fetch_type() shall return the type of an object. The type is a general classification of a Verilog HDL object, represented as a predefined constant (defined in acc_user.h). Refer to Table 113 for a list of all of the type constants that can be returned by acc_fetch_type(). Many Verilog HDL objects can have a type and a fulltype. The type of an object is its general Verilog HDL type classification. The fulltype is the specific type of the object. Table 156 illustrates the difference between the type of an object and the fulltype of the same object. Table 156—The difference between the type and the fulltype of an object For a handle to A setup timing check An and gate primitive A sequential UDP acc_fetch_type() shall return accTchk accPrimitive accPrimitive acc_fetch_fulltype() shall return accSetup accAndGate accSeqPrim The example shown in Figure 94 uses acc_fetch_type() to identify the type of an object (the functions display_primitive_type and display_timing_check_type used in this example are presented in the usage examples in 23.19). Copyright © 2001 IEEE. All rights reserved. 471 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® #include "acc_user.h" PLI_INT32 display_object_type() { handle object_handle; /*initialize environment for ACC routines*/ acc_initialize(); object_handle = acc_handle_tfarg(1); /*display object type*/ switch(acc_fetch_type(object_handle) ) { case accModule: io_printf("Object is a module\n"); break; case accNet: io_printf("Object is a net\n"); break; case accPath: io_printf("Object is a module path\n"); break; case accPort: io_printf("Object is a module port\n"); break; case accPrimitive: display_primitive_type(object_handle); break; case accTchk: display_timing_check_type(object_handle); break; case accTerminal: io_printf("Object is a primitive terminal\n"); break; } acc_close(); } Figure 94—Using acc_fetch_type() 472 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE 23.35 acc_fetch_type_str() IEEE Std 1364-2001 Version C Synopsis: Syntax: Returns: Arguments: Related routines: acc_fetch_type_str() Get a string that indicates the type of its argument. acc_fetch_type_str(type) Type Description PLI_BYTE8 * Pointer to a character string Type Name Description PLI_INT32 type A predefined integer constant that stands for an object type or fulltype Use acc_fetch_type() to get the type of an object as an integer constant Use acc_fetch_fulltype() to get the fulltype of an object as an integer constant The ACC routine acc_fetch_type_str() shall return the character string that specifies the type of its argument. The argument passed to acc_fetch_type_str() should be an integer value returned from either acc_fetch_type() or acc_fetch_fulltype(). The return value for this routine is placed in the ACC internal string buffer. See 22.9 for explanation of strings in ACC routines. In the example shown in Figure 95, a handle to an argument is passed to a C application. The application displays the name of the object and the type of the object. #include "acc_user.h" PLI_INT32 display_object_type(object) handle object; { PLI_INT32 type = acc_fetch_type(object); io_printf("Object %s is of type %s \n", acc_fetch_fullname(object), acc_fetch_type_str(type)); } Figure 95—Using acc_fetch_type_str() In this example, if the application is passed a handle to an object named top.param1, the application shall produce the following output: Object top.param1 is of type accParameter The output string, accParameter, is the name of the integer constant that represents the parameter type. Copyright © 2001 IEEE. All rights reserved. 473 IEEE Std 1364-2001 Version C 23.36 acc_fetch_value() IEEE STANDARD VERILOG® Synopsis: Syntax: Returns: Arguments: Optional Related routines: acc_fetch_value() Get the logic or strength value of a net, reg, or variable. acc_fetch_value(object_handle, format_string, value) Type Description PLI_BYTE8 * Pointer to a character string Type Name Description handle object_handle Handle of the object quoted string or PLI_BYTE8 * format_string A literal string or character string pointer with one of the following specifiers for formatting the return value: “%b” “%d” “%h” “%o” “%v” “%%” s_acc_value * value Pointer to a structure in which the value of the object is returned when the format string is “%%” (should be set to null when not used) Use acc_fetch_size() to determine how many bits wide the object is Use acc_set_value() to put a logic value on the object The ACC routine acc_fetch_value() shall return logic simulation values for scalar or vector nets, reg, and integer, time and real variables; acc_fetch_value() shall return strength values for scalar nets and scalar regs only. The routine acc_fetch_value() shall return the logic and strength values in one of two ways: — The value can be returned as a string — The value can be returned as an aval/bval pair in a predefined structure. The return method used by acc_fetch_value() shall be controlled by the format_string argument, as shown in Table 157. Table 157—How acc_fetch_value() returns values format_specifier “%b” “%d” “%h” “%o” “%v” “%%” Return format binary decimal hexadecimal octal strength s_acc_value structure Description Value shall be retrieved as a string, and a character pointer to the string shall be returned Value shall be retrieved and placed in a structure variable pointed to by the optional value argument The string value returned shall have the same form as output from the formatted built-in system task $display, in terms of value lengths and value characters used. The length shall be of arbitrary size, and unknown and high-impedance values shall be obtained. Note that strings are placed in a temporary buffer, and they should be preserved if not used immediately. Refer to 22.9 for details on preserving strings. 474 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The “%v” format shall return a three character string containing the strength code of a scalar net. Refer to 17.1.1.5 for the strength representations. When a format_string of “%%” is specified, acc_fetch_value() shall retrieve the logic value and strength to a predefined structure, s_acc_value, which is defined in acc_user.h and is shown below [note that this structure definition is also used with the acc_set_value() routine]. typedef struct t_setval_value { PLI_INT32 format; union { PLI_BYTE8 *str; PLI_INT32 scalar; PLI_INT32 integer; double real; p_acc_vecval vector; } value; } s_setval_value, *p_setval_value, s_acc_value, *p_acc_value; Figure 96—s_acc_value structure To use the “%%” format_string to retrieve values to a structure requires the following steps: a) A structure variable shall first be declared of type s_acc_value. b) The format field of the structure has to be set to a predefined constant. The format controls which fields in the s_acc_value structure shall be used when acc_fetch_value() returns the value. The predefined constants for the format shall be one of the constants shown in Table 158. c) The structure variable has to be passed as the third argument to acc_fetch_value(). d) The function return value from acc_fetch_value() should be ignored. Table 158—Format constants for the s_acc_value structure Format constant accBinStrVal accOctStrVal accDecStrVal accHexStrVal accStringVal accScalarVal accIntVal accRealVal accVectorVal acc_fetch_value() shall return the value to the s_acc_value union field str str str str str scalar integer real vector Description value is retrieved in the same format as “%b” value is retrieved in the same format as “%o” value is retrieved in the same format as “%d” value is retrieved in the same format as “%h” value is converted to a string, see 2.6 for a description of Verilog strings value is retrieved as one of the constants: acc0, acc1, accZ or accX value is retrieved as a C integer value is retrieved as a C double value is represented as aval/bval pairs stored in an array of s_acc_vecval structures Copyright © 2001 IEEE. All rights reserved. 475 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® For example, calling acc_fetch_value() with the following setup would return a string in the value.str field. (This is essentially the same as using acc_fetch_value() with a %b format string.) s_acc_value value; value.format = accBinStrVal; (void)acc_fetch_value(Net, “%%”, &value); If the format field for acc_fetch_value() is set to accVectorVal, then the value shall be placed in the record(s) pointed to by the value field. The value field shall be a pointer to an array of one or more s_acc_vecval structures. The s_acc_vecval structure is defined in the acc_user.h file and is listed in Figure 97. The structure shall contain two integers: aval and bval. Each s_acc_vecval record shall represent 32 bits of a vector. The encoding for each bit value is shown in Table 159. typedef struct t_acc_vecval { PLI_INT32 aval; PLI_INT32 bval; } s_acc_vecval, *p_acc_vecval; Figure 97—s_acc_vecval structure Table 159—Encoding of bits in the s_acc_vecval structure aval bval Value 0 0 0 1 0 1 0 1 Z 1 1 X The array of s_acc_vecval structures shall contain a record for every 32 bits of the vector, plus a record for any remaining bits. If a vector has N bits, then there shall be ((N-1)/32)+1 s_acc_vecval records. The routine acc_fetch_size() can be used to determine the value of N. The lsb of the vector shall be represented by the lsb of the first record of s_acc_vecval array. The 33rd bit of the vector shall be represented by the lsb of the second record of the array, and so on. See Figure 99 for an example of acc_fetch_value() used in this way. Note that when using aval/bval pairs, the s_acc_value record and the appropriately sized s_acc_vecval array shall first be declared. Setting the second parameter to acc_fetch_value() to “%%” and the third parameter to null shall be an error. 476 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The example application shown in Figure 98 uses acc_fetch_value() to retrieve the logic values of all nets in a module as strings. #include "acc_user.h" PLI_INT32 display_net_values() { handle mod, net; /*initialize environment for ACC routines*/ acc_initialize(); /*get handle for module*/ mod = acc_handle_tfarg(1); /*get all nets in the module and display their values*/ /* in binary format*/ net = null; while(net = acc_next_net(mod, net)) io_printf("Net value: %s\n", acc_fetch_value(net,"%b", null)); acc_close(); } Figure 98—Using acc_fetch_value() to retrieve the logic values as strings The example in Figure 99 uses acc_fetch_value() to retrieve a value into a structure, and then prints the value. The example assumes the application, my_fetch_value, is called from the following user-defined system task: $my_fetch_value(R); Copyright © 2001 IEEE. All rights reserved. 477 IEEE Std 1364-2001 Version C IEEE STANDARD VERILOG® #include "acc_user.h" PLI_INT32 my_fetch_value() { handle reg = acc_handle_tfarg(1); PLI_INT32 size = ((acc_fetch_size(reg) - 1) / 32) + 1; s_acc_value value; int index1, min_size; static PLI_BYTE8 table[4] = {’0’,’1’,’z’,’x’}; static PLI_BYTE8 outString[33]; io_printf("The value of %s is ",acc_fetch_name(reg)); value.format = accVectorVal; value.value.vector = (p_acc_vecval)malloc(size*sizeof(s_acc_vecval)); (void)acc_fetch_value(reg, "%%",&value); for (index1 = size - 1; index1 >= 0; index1--) { int index2; PLI_INT32 abits = value.value.vector[index1].aval; PLI_INT32 bbits = value.value.vector[index1].bval; if (index1 == size - 1) { min_size = (acc_fetch_size(reg) % 32); if (!min_size) min_size = 32; } else min_size = 32; outString[min_size] = ’\0’; min_size--; outString[min_size] = table[((bbits & 1) << 1) | (abits & 1)]; abits >>= 1; for (index2 = min_size - 1; index2 >= 0; index2--) { outString[index2] = table[(bbits & 2) | (abits & 1)]; abits >>= 1; bbits >>= 1; } io_printf("%s", outString); } io_printf("\n"); return(0); } Figure 99—Using acc_fetch_value() to retrieve values into a data structure 478 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE 23.37 acc_free() IEEE Std 1364-2001 Version C Synopsis: Syntax: Returns: Arguments: Related routines: acc_free() Frees memory allocated by acc_collect(). acc_free(handle_array_pointer) Type Description void Type No return Name Description handle * handle_array_pointer Pointer to the array of handles allocated by acc_collect() Use acc_collect() to collect handles returned by acc_next_ routines The ACC routine acc_free() shall deallocate memory that was allocated by the routine acc_collect(). The example shown in Figure 100 uses acc_free() to deallocate memory allocated by acc_collect() to collect handles to all nets in a module. #include "acc_user.h" PLI_INT32 display_nets() { handle *list_of_nets, module_handle; PLI_INT32 net_count, i; /*initialize environment for ACC routines*/ acc_initialize(); /*get handle for module*/ module_handle = acc_handle_tfarg(1); /*collect and display all nets in the module*/ list_of_nets = acc_collect(acc_next_net, module_handle, &net_count); for(i=0; i < net_count; i++) io_printf("Net name is: %s\n", acc_fetch_name(list_of_nets[i])); /*free memory used by array list_of_nets*/ acc_free(list_of_nets); acc_close(); } Figure 100—Using acc_free() Copyright © 2001 IEEE. All rights reserved. 479 IEEE Std 1364-2001 Version C 23.38 acc_handle_by_name() IEEE STANDARD VERILOG® Synopsis: Syntax: Returns: Arguments: Related Routines acc_handle_by_name() Get the handle to any named object based on its name and scope. acc_handle_by_name(object_name, scope_handle) Type Description handle A handle to the specified object Type Name Description quoted string or PLI_BYTE8 * object_name Literal name of an object or a character string pointer to the object name handle scope_handle Handle to scope, or null Use acc_handle_object() to get a handle based on the local instance name of an object The ACC routine acc_handle_by_name() shall return the handle to any named object based on its specified name and scope. The routine can be used in two ways, as shown in Table 160. Table 160—How acc_handle_by_name() works When the scope_handle is A valid scope handle null acc_handle_by_name() shall Search for the object_name in the scope specified Search for the object_name in the module containing the current system task or function The routine acc_handle_by_name() combines the functionality of acc_set_scope() and acc_handle_object(), making it possible to obtain handles for objects that are not in the local scope without having to first change scopes. Object searching shall conform to rules in 12.4 on hierarchical name referencing. Table 161 lists the objects in a Verilog HDL description for which acc_handle_by_name() shall return a handle. Table 161—Named objects supported by acc_handle_by_name() Modules Primitives Nets Regs Integer, time and real variables Named events Parameters Specparams Named blocks Verilog HDL tasks Verilog HDL functions 480 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C The routine acc_handle_by_name() does not return handles for module paths, intermodule paths, data paths, or ports. Use an appropriate acc_next_ or other ACC routines for these objects. The example shown in Figure 101 uses acc_handle_by_name() to set the scope and get the handle to an object if the object is in the module. #include "acc_user.h" PLI_INT32 is_net_in_module(module_handle, net_name) handle module_handle; PLI_BYTE8 *net_name; { handle net_handle; /*set scope to module and get handle for net */ net_handle = acc_handle_by_name(net_name, module_handle); if (net_handle) io_printf("Net %s found in module %s\n", net_name, acc_fetch_fullname(module_handle) ); else io_printf("Net %s not found in module %s\n", net_name, acc_fetch_fullname(module_handle) ); } Figure 101—Using acc_handle_by_name() Note that in this example net_handle = acc_handle_by_name(net_name, module_handle); could also have been written as follows: acc_set_scope(module_handle); net_handle = acc_handle_object(net_name); Copyright © 2001 IEEE. All rights reserved. 481 IEEE Std 1364-2001 Version C 23.39 acc_handle_calling_mod_m IEEE STANDARD VERILOG® Synopsis: Syntax: Returns: Arguments: acc_handle_calling_mod_m Get a handle to the module containing the instance of the user-defined system task or function that called the PLI application. acc_handle_calling_mod_m Type Description handle Handle to a module Type Name Description None The ACC routine acc_handle_calling_mod_m shall return a handle to the module that contains the instance of the user-defined system task or function that called the PLI application. 482 Copyright © 2001 IEEE. All rights reserved. HARDWARE DESCRIPTION LANGUAGE IEEE Std 1364-2001 Version C 23.40 acc_handle_condition() Synopsis: Syntax: Returns: Arguments: acc_handle_condition() Get a handle to the conditional expression of a module path, data path, or timing check terminal. acc_handle_condition(path_handle) Type Description handle Handle to a conditional expression Type Name Description handle path_handle Handle to a module path, data path, or timing check terminal The ACC routine acc_handle_condition() shall return a handle to a conditional expression for the specified module path, data path, or timing check terminal. The routine shall return null when — The module path, data path, or timing check terminal has no condition specified — The module path has an ifnone condition specified To determine if a module path has an ifnone condition specified, use the ACC routine acc_object_of_type() to check for the property type of accModPathHasIfnone. The example shown in Figure 102 provides functionality to see if a path is conditional, and, if it is, whether it is level-sensitive or edge-sensitive. The application assumes that the input is a valid handle to a module path. int is_path_conditional(path) { if (acc_handle_condition(path) ) return(TRUE); else return(FALSE); } int is_level_sensitive(path) { int flag; handle path_in = acc_next_input(path, null); if (is_path_conditional(path) && acc_fetch_edge(path_in)) flag = FALSE; /* path is edge-sensitive */ else flag = TRUE; /* path is level_sensitive */ acc_release_object(path_in); return (flag); } Figure 102—Using acc_handle_condition() Copyright © 2001 IEEE. All rights reserved. 483 IEEE Std 1364-2001 Version C 23.41 acc_handle_conn() IEEE STANDARD VERILOG® Synopsis: Syntax: Returns: Arguments: Related r