?? ch04.2.htm
字號:
<HTML><HEAD> <META NAME="GENERATOR" CONTENT="Adobe PageMill 2.0 Mac"> <LINK REL="STYLESHEET" HREF="ch04.css"> <TITLE> 4.2 Operands </TITLE></HEAD><BODY BGCOLOR="#ffffff"><P><A NAME="pgfId=626"></A><HR ALIGN=LEFT></P><P><A HREF="ch04.htm">Chapter start</A> <A HREF="ch04.1.htm">Previous page</A> <A HREF="ch04.3.htm">Next page</A></P><H1>4.2 Operands</H1><P><P CLASS="Body"><A NAME="pgfId=635"></A>As stated before, there are severaltypes of operands that can be specified in expressions. The simplest typeis a reference to a net or register in its complete form--that is, justthe name of the net or register is given. In this case, all of the bitsmaking up the net or register value shall be used as the operand.</P><P><P CLASS="Body"><A NAME="pgfId=668"></A>If a single bit of a vector netor register is required, then a bit-select operand shall be used. A part-selectoperand shall be used to reference a group of adjacent bits in a vectornet or register.</P><P><P CLASS="Body"><A NAME="pgfId=669"></A>A memory element can be referencedas an operand. A concatenation of other operands, (including nested concatenations)can be specified as an operand. A function call is an operand.</P><P><P CLASS="SubSection"><A NAME="pgfId=493"></A>Net and register bit-selectand part-select addressing</P><P><P CLASS="Body"><A NAME="pgfId=674"></A><I>Bit-selects</I> extract aparticular bit from a vector net or register. The bit can be addressed usingan expression. If the bit select is out of the address bounds or the bit-selectis <CODE>x</CODE> or <CODE>z</CODE> , then the value returned by the referenceshall be <CODE>x</CODE> .</P><P><P CLASS="Body"><A NAME="pgfId=679"></A>Several contiguous bits in avector register or net can be addressed, and are known as <I>part-selects</I>. A part-select of a vector register or net is given with the followingsyntax:</P><PRE>vect[msb_expr:lsb_expr]</PRE><P><P CLASS="Body"><A NAME="pgfId=681"></A>Both expressions must be constantexpressions. The first expression must address a more significant bit thanthe second expression. If the part-select is out of the address bounds orthe part-select is <CODE>x</CODE> or <CODE>z</CODE> , then the value returnedby the reference shall be <CODE>x</CODE> .</P><P><P CLASS="Body"><A NAME="pgfId=805"></A>The bit-select or part-selectof a variable declared as real and realtime shall be considered illegal.</P><P><P CLASS="Body"><A NAME="pgfId=912"></A>1. The following example specifiesthe single bit of <CODE>acc</CODE> vector that is addressed by the operand<CODE>index</CODE> .</P><PRE>acc[index]</PRE><P><P CLASS="Body"><A NAME="pgfId=676"></A>The actual bit that is accessedby an address is, in part, determined by the declaration of <CODE>acc</CODE>. For instance, each of the declarations of <CODE>acc</CODE> shown in thenext example causes a particular value of <CODE>index</CODE> to access adifferent bit:</P><PRE><B>reg</B>[15:0] acc;<B>reg</B> [1:16] acc;</PRE><P><P CLASS="Body"><A NAME="pgfId=1274"></A>2. The next example and thebullet items that follow it illustrate the principles of bit addressing.The code declares an 8-bit register called <CODE>vect</CODE> and initializesit to a value of 4. The bullet items describe how the separate bits of thatvector can be addressed.</P><PRE><B>reg</B> [7:0] vect;vect = 4; // fills vect with the pattern 00000100 // msb is bit 7, lsb is bit 0</PRE><UL> <LI><A NAME="pgfId=683"></A>if the value of <CODE>addr</CODE> is 2, then <CODE>vect[addr]</CODE> returns <CODE>1</CODE> <LI><A NAME="pgfId=684"></A>if the value of <CODE>addr</CODE> is out of bounds, then <CODE>vect[addr]</CODE> returns <CODE>x</CODE> <LI><A NAME="pgfId=685"></A>if <CODE>addr</CODE> is 0, 1, or 3 through 7, <CODE>vect[addr]</CODE> returns <CODE>0</CODE> <LI><A NAME="pgfId=686"></A><CODE>vect[3:0]</CODE> returns the bits <CODE>0100</CODE> <LI><A NAME="pgfId=687"></A><CODE>vect[5:1]</CODE> returns the bits <CODE>00010</CODE> <LI><A NAME="pgfId=688"></A><CODE>vect[</CODE> <I>expression that returns x</I> <CODE>]</CODE> returns <CODE>x</CODE> <LI><A NAME="pgfId=689"></A><CODE>vect[</CODE> <I>expression that returns z</I> <CODE>]</CODE> returns <CODE>x</CODE> <LI><A NAME="pgfId=690"></A>if any bit of <CODE>addr</CODE> is <CODE>x/z</CODE> , then the value of <CODE>addr</CODE> is <CODE>x</CODE></UL><P><P CLASS="Note"><A NAME="pgfId=1273"></A>NOTES</P><OL> <P><P CLASS="NumberedNote1"><A NAME="pgfId=1277"></A>1) --Part-select Indices that evaluate to x or z may be flagged as a compile time error. <P><P CLASS="NumberedNote2"><A NAME="pgfId=1278"></A>2) --Bit-select or Part-select indices that are outside of the declared range may be flagged as a compile time error.</OL><P><P CLASS="SubSection"><A NAME="pgfId=691"></A>Memory addressing</P><P><P CLASS="Body"><A NAME="pgfId=693"></A>Declaration of memory is discussedin section 3.8. This section discusses memory addressing. </P><P><P CLASS="Body"><A NAME="pgfId=913"></A>The next example declares a memoryof 1024 8-bit words:</P><PRE><B>reg</B> [7:0] mem_name[0:1023];</PRE><P><P CLASS="Body"><A NAME="pgfId=695"></A>The syntax for a memory addressshall consist of the name of the memory and an expression for the address--specifiedwith the following format:</P><PRE>mem_name[addr_expr]</PRE><P><P CLASS="Body"><A NAME="pgfId=697"></A>The <CODE>addr_expr</CODE> canbe any expression; therefore, memory indirections can be specified in asingle expression. The next example illustrates memory indirection:</P><PRE>mem_name[mem_name[3]]</PRE><P><P CLASS="Body"><A NAME="pgfId=699"></A>In the above example, <CODE>mem_name[3]</CODE>addresses word three of the memory called <CODE>mem_name</CODE> . The valueat word three is the index into <CODE>mem_name</CODE> that is used by thememory address <CODE>mem_name[mem_name[3]]</CODE> . As with bit-selects,the address bounds given in the declaration of the memory determine theeffect of the address expression. If the index is out of the address boundsor if any bit in the address is <CODE>x</CODE> or <CODE>z</CODE> , thenthe value of the reference shall be <CODE>x</CODE> .</P><P><P CLASS="Note"><A NAME="pgfId=700"></A>NOTE--There is no mechanism toexpress bit-selects or part-selects of memory elements directly. If thisis required, then the memory element has to be first transferred to an appropriatelysized temporary register.</P><P><P CLASS="SubSection"><A NAME="pgfId=702"></A>Strings</P><P><P CLASS="Body"><A NAME="pgfId=706"></A>String operands shall be treatedas constant numbers consisting of a sequence of 8-bit ASCII codes, one percharacter. Any Verilog HDL operator can manipulate string operands. Theoperator shall behave as though the entire string were a single numericvalue.</P><P><P CLASS="Body"><A NAME="pgfId=1272"></A>When a variable is larger thanrequired to hold the value being assigned, the contents after the assignmentshall be padded on the left with zeros. This is consistent with the paddingthat occurs during assignment of non-string values.</P><P><P CLASS="Body"><A NAME="pgfId=704"></A>The following example declaresa string variable large enough to hold 14 characters and assigns a valueto it. The example then manipulates the string using the concatenation operator.</P><PRE><B>module</B> string_test;<B>reg</B> [8*14:1] stringvar;<B>initial </B> <B>begin</B> stringvar = "Hello world"; <B>$display</B> ("%s is stored as %h", stringvar, stringvar); stringvar = {stringvar,"!!!"}; <B>$display</B> ("%s is stored as %h", stringvar, stringvar);<B>end</B><B>endmodule</B></PRE><P><P CLASS="Body"><A NAME="pgfId=709"></A>The result of simulating theabove description is:</P><PRE>Hello world is stored as 00000048656c6c6f20776f726c64Hello world!!! is stored as 48656c6c6f20776f726c64212121</PRE><P><P CLASS="SubSubSect"><A NAME="pgfId=712"></A>String operations</P><P><P CLASS="Body"><A NAME="pgfId=713"></A>The common string operationscopy, concatenate, and compare are supported by Verilog HDL operators. Copyis provided by simple assignment. Concatenation is provided by the concatenationoperator. Comparison is provided by the equality operators.</P><P><P CLASS="Body"><A NAME="pgfId=503"></A>When manipulating string valuesin vector variables, at least <CODE>8*n</CODE> bits shall be required inthe vector, where <CODE>n</CODE> is the number of characters in the string.</P><P><P CLASS="SubSubSect"><A NAME="pgfId=716"></A>String value padding andpotential problems</P><P><P CLASS="Body"><A NAME="pgfId=717"></A>When strings are assigned tovariables, the values stored shall be padded on the left with zeros. Paddingcan affect the results of comparison and concatenation operations. The comparisonand concatenation operators shall not distinguish between zeros resultingfrom padding and the original string characters.</P><P><P CLASS="Body"><A NAME="pgfId=1275"></A>The following example illustratesthe potential problem.</P><PRE><B>reg</B> [8*10:1] s1, s2;<B>initial begin</B> s1 = "Hello"; s2 = " world!";<B> if</B> ({s1,s2} == "Hello world!") <B>$display</B> ("strings are equal");<B>end</B></PRE><P><P CLASS="Body"><A NAME="pgfId=721"></A>The comparison in the exampleabove fails because during the assignment the string variables get paddedas illustrated in the next example:</P><PRE>s1 = 000000000048656c6c6fs2 = 00000020776f726c6421</PRE><P><P CLASS="Body"><A NAME="pgfId=723"></A>The concatenation of <CODE>s1</CODE>and <CODE>s2</CODE> includes the zero padding, resulting in the followingvalue:</P><PRE>000000000048656c6c6f00000020776f726c6421</PRE><P><P CLASS="Body"><A NAME="pgfId=725"></A>Since the string "Helloworld!" contains no zero padding, the comparison fails, as shown below:</P><P><P CLASS="Body"><A NAME="pgfId=726"></A> </P><P><IMG SRC="ch04-7.gif" WIDTH="416" HEIGHT="94" NATURALSIZEFLAG="3" ALIGN="BOTTOM"> <P CLASS="Body"><A NAME="pgfId=727"></A>The above comparison yieldsa result of zero, which is equivalent to false.</P><P><P CLASS="SubSubSect"><A NAME="pgfId=729"></A>Null string handling</P><P><P CLASS="Body"><A NAME="pgfId=730"></A>The null string (<CODE> ""</CODE>) shall be considered equivalent to the ASCII value zero (0), which is differentfrom a string <CODE>"0"</CODE> .</P><P><HR ALIGN=LEFT></P><P><A HREF="ch04.htm">Chapter start</A> <A HREF="ch04.1.htm">Previous page</A> <A HREF="ch04.3.htm">Next page</A></BODY></HTML>?? 快捷鍵說明
復制代碼
Ctrl + C
搜索代碼
Ctrl + F
全屏模式
F11
切換主題
Ctrl + Shift + D
顯示快捷鍵
?
增大字號
Ctrl + =
減小字號
Ctrl + -