ch04.1.htm

Verilog DHL教程

the result of the expression can be determined as zero without evaluating
the sub-expression <CODE>regB | regC</CODE> .</P>

<P><P CLASS="SubSection"><A NAME="pgfId=711"></A>Arithmetic operators</P>

<P><P CLASS="Body"><A NAME="pgfId=568"></A>The binary arithmetic operators
are the following:</P>

<P><TABLE BORDER="1" CELLSPACING="2" CELLPADDING="2">
<CAPTION ALIGN="TOP"><P CLASS="TableTitle"><A NAME="pgfId=775"></A>Table&nbsp;4-5: Arithmetic
operators defined</CAPTION>
<TR>
<TD><P CLASS="CellBody"><A NAME="pgfId=841"></A><CODE>a + b</CODE></TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=842"></A>a plus b</TD></TR>
<TR>
<TD><P CLASS="CellBody"><A NAME="pgfId=972"></A>a<CODE> - </CODE>b</TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=973"></A>a minus b</TD></TR>
<TR>
<TD><P CLASS="CellBody"><A NAME="pgfId=974"></A>a<CODE> * b</CODE></TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=975"></A>a multiply by b</TD></TR>
<TR>
<TD><P CLASS="CellBody"><A NAME="pgfId=976"></A><CODE>a / b</CODE></TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=977"></A>a divide by b</TD></TR>
<TR>
<TD><P CLASS="CellBody"><A NAME="pgfId=978"></A><CODE>a % b</CODE></TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=979"></A>a modulo b</TD></TR>
</TABLE>
<P CLASS="Body"><A NAME="pgfId=572"></A>The integer division shall truncate
any fractional part. The modulus operator, for example <CODE>y % z</CODE>
, gives the remainder when the first operand is divided by the second, and
thus is zero when <CODE>z</CODE> divides <CODE>y</CODE> exactly. The result
of a modulus operation shall take the sign of the first operand.</P>

<P><P CLASS="Body"><A NAME="pgfId=576"></A>The unary arithmetic operators
shall take precedence over the binary operators. The unary operators are
the following:</P>

<P><TABLE BORDER="1" CELLSPACING="2" CELLPADDING="2">
<CAPTION ALIGN="TOP"><P CLASS="TableTitle"><A NAME="pgfId=571"></A>Table&nbsp;4-6: Unary operators
defined</CAPTION>
<TR>
<TD><P CLASS="CellBody"><A NAME="pgfId=840"></A>+m</TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=980"></A>unary plus m (same as m)</TD></TR>
<TR>
<TD><P CLASS="CellBody"><A NAME="pgfId=981"></A>-m</TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=982"></A>unary minus m</TD></TR>
</TABLE>
<P CLASS="Body"><A NAME="pgfId=575"></A>For the arithmetic operators, if
any operand bit value is the unknown value <CODE>x</CODE> , or high impedance
value <CODE>z</CODE> , then the entire result value shall be <CODE>x</CODE>
.</P>

<P><P CLASS="Body"><A NAME="pgfId=952"></A><A HREF="#pgfId=555">Table&nbsp;4-7</A>
gives examples of modulus operations.</P>

<P><TABLE BORDER="1" CELLSPACING="2" CELLPADDING="2">
<CAPTION ALIGN="TOP"><P CLASS="TableTitle"><A NAME="pgfId=555"></A>Table&nbsp;4-7: Examples of
modulus operations</CAPTION>
<TR>
<TH><P CLASS="CellHeading"><A NAME="pgfId=1046"></A>Modulus Expression</TH>
<TH><P CLASS="CellHeading"><A NAME="pgfId=1047"></A>Result</TH>
<TH><P CLASS="CellHeading"><A NAME="pgfId=1048"></A>Comments</TH></TR>
<TR>
<TD><P CLASS="CellBody"><A NAME="pgfId=1049"></A>10 % 3</TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=1050"></A>1</TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=1051"></A>10/3 yields a remainder of
1</TD></TR>
<TR>
<TD><P CLASS="CellBody"><A NAME="pgfId=1052"></A>11 % 3</TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=1053"></A>2</TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=1054"></A>11/3 yields a remainder of
2</TD></TR>
<TR>
<TD><P CLASS="CellBody"><A NAME="pgfId=1055"></A>12 % 3</TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=1056"></A>0</TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=1057"></A>12/3 yields no remainder</TD></TR>
<TR>
<TD><P CLASS="CellBody"><A NAME="pgfId=1058"></A>-10 % 3</TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=1059"></A>-1</TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=1060"></A>the result takes the sign of
the first operand</TD></TR>
<TR>
<TD><P CLASS="CellBody"><A NAME="pgfId=1061"></A>11 % -3</TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=1062"></A>2</TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=1067"></A>the result takes the sign of
the first operand</TD></TR>
<TR>
<TD><P CLASS="CellBody"><A NAME="pgfId=1064"></A>-4'd12 % 3</TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=1065"></A>1</TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=1066"></A>-4'd12 is seen as a large,
positive number that leaves a remainder of 1 when divided by 3</TD></TR>
</TABLE>
<P CLASS="SubSection"><A NAME="pgfId=520"></A>Arithmetic expressions with
registers and integers</P>

<P><P CLASS="Body"><A NAME="pgfId=942"></A>An arithmetic operation on a
reg type register shall be treated differently than an arithmetic operation
on an integer data type. A reg data type shall be treated as an <I>unsigned</I>
value and an integer data type shall be treated as a <I>signed</I> value.
Thus, if a sized constant with a negative value is stored in a reg type
register, a positive constant which is a two's complement of the sized constant
shall be the value stored in the reg type register. When this register is
used in an arithmetic expression, the positive constant shall be used as
the value of the register. In contrast, if a sized constant with a negative
value is stored in an integer type register and used in an arithmetic expression,
the expression shall evaluate using signed arithmetic.</P>

<P><P CLASS="Body"><A NAME="pgfId=933"></A><A HREF="#pgfId=939">Table&nbsp;4-8</A>
lists how arithmetic operators interpret each data type</P>

<P><TABLE BORDER="1" CELLSPACING="2" CELLPADDING="2">
<CAPTION ALIGN="TOP"><P CLASS="TableTitle"><A NAME="pgfId=939"></A>Table&nbsp;4-8: Data type
interpretation by arithmetic operators</CAPTION>
<TR>
<TH><P CLASS="CellHeading"><A NAME="pgfId=1290"></A>Data Type</TH>
<TH><P CLASS="CellHeading"><A NAME="pgfId=1291"></A>Interpretation</TH></TR>
<TR>
<TD><P CLASS="CellBody"><A NAME="pgfId=1292"></A>net</TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=1293"></A>unsigned</TD></TR>
<TR>
<TD><P CLASS="CellBody"><A NAME="pgfId=1294"></A>reg</TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=1302"></A>unsigned</TD></TR>
<TR>
<TD><P CLASS="CellBody"><A NAME="pgfId=1296"></A>integer</TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=1303"></A>signed, two's complement</TD></TR>
<TR>
<TD><P CLASS="CellBody"><A NAME="pgfId=1298"></A>time</TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=1304"></A>unsigned</TD></TR>
<TR>
<TD><P CLASS="CellBody"><A NAME="pgfId=1300"></A>real, realtime</TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=1305"></A>signed, floating point</TD></TR>
</TABLE>
</P>

<P><P CLASS="Body"><A NAME="pgfId=971"></A>The example below shows various
ways to divide<I> </I>&quot;minus twelve by three&quot;--using <B>integer</B>
and <B>reg</B> data types in expressions.</P>

<PRE><B>integer</B> intA;
<B>reg</B> [15:0] regA;

intA = -4'd12;
regA = intA / 3;   // expression result is -4,
                   // intA is an integer data type

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 reg data type

regA = -12 / 3;    // expression result is -4, -12 is effectively
                   // an integer data type. regA is 65532</PRE>

<P><P CLASS="SubSection"><A NAME="pgfId=584"></A>Relational operators</P>

<P><P CLASS="Body"><A NAME="pgfId=574"></A><A HREF="#pgfId=499">Table&nbsp;4-9</A>
lists and defines the relational operators.</P>

<P><TABLE BORDER="1" CELLSPACING="2" CELLPADDING="2">
<CAPTION ALIGN="TOP"><P CLASS="TableTitle"><A NAME="pgfId=499"></A>Table&nbsp;4-9: The relational
operators defined</CAPTION>
<TR>
<TD><P CLASS="CellBody"><A NAME="pgfId=501"></A>a&nbsp;&lt;&nbsp;b</TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=504"></A>a&nbsp;less&nbsp;than&nbsp;b</TD></TR>
<TR>
<TD><P CLASS="CellBody"><A NAME="pgfId=516"></A>a&nbsp;&gt;&nbsp;b</TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=517"></A>a&nbsp;greater&nbsp;than&nbsp;b</TD></TR>
<TR>
<TD><P CLASS="CellBody"><A NAME="pgfId=518"></A>a&nbsp;&lt;=&nbsp;b</TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=519"></A>a&nbsp;less&nbsp;than&nbsp;or&nbsp;equal&nbsp;to&nbsp;b</TD></TR>
<TR>
<TD><P CLASS="CellBody"><A NAME="pgfId=521"></A>a&nbsp;&gt;=&nbsp;b</TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=573"></A>a&nbsp;greater&nbsp;than&nbsp;or&nbsp;equal&nbsp;to&nbsp;b</TD></TR>
</TABLE>
<P CLASS="Body"><A NAME="pgfId=1068"></A>An expression using these <I>relational
operators</I> shall yield the scalar value <CODE>0</CODE> if the specified
relation is <I>false</I> , or the value <CODE>1</CODE> if it is <I>true</I>
. If, due to unknown or high impedance bits in the operands, the relation
is <I>ambiguous</I> , then the result shall be one bit unknown value <CODE>(x)</CODE>
.</P>

<P><P CLASS="Body"><A NAME="pgfId=914"></A>When two operands of unequal
bit-lengths are used, the smaller operand shall be zero filled on the most
significant bit side to extend to the size of the larger operand.</P>

<P><P CLASS="Body"><A NAME="pgfId=589"></A>All the relational operators
shall have the same precedence. Relational operators shall have lower precedence
than arithmetic operators.</P>

<P><P CLASS="Body"><A NAME="pgfId=764"></A>The following examples illustrate
the implications of this precedence rule:</P>

<PRE>a&nbsp;&lt;&nbsp;foo - 1   // this expression is the same as
a&nbsp;&lt;&nbsp;(foo - 1) // this expression, but . . .
foo&nbsp;-&nbsp;(1&nbsp;&lt;&nbsp;a) // this one is not the same as 
foo&nbsp;-&nbsp;1&nbsp;&lt;&nbsp;a   // this expression</PRE>

<P><P CLASS="Body"><A NAME="pgfId=591"></A>When <CODE>foo - (1 &lt; a)</CODE>
evaluates, the relational expression evaluates first and then either zero
or one is subtracted from <CODE>foo</CODE> . When <CODE>foo - 1 &lt; a</CODE>
evaluates, the value of <CODE>foo</CODE> operand is reduced by one and then
compared with <CODE>a</CODE> .</P>

<P><P CLASS="SubSection"><A NAME="pgfId=593"></A>Equality operators</P>

<P><P CLASS="Body"><A NAME="pgfId=1072"></A>The <I>equality operators</I>
shall rank lower in precedence than the relational operators. <A HREF="#pgfId=585">Table&nbsp;4-10</A>
lists and defines the equality operators.</P>

<P><TABLE BORDER="1" CELLSPACING="2" CELLPADDING="2">
<CAPTION ALIGN="TOP"><P CLASS="TableTitle"><A NAME="pgfId=585"></A>Table&nbsp;4-10: The equality
operators defined</CAPTION>
<TR>
<TD><P CLASS="CellBody"><A NAME="pgfId=587"></A>a&nbsp;=== b</TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=588"></A>a&nbsp;equal&nbsp;to&nbsp;b,&nbsp;including&nbsp;x&nbsp;and&nbsp;z</TD></TR>
<TR>
<TD><P CLASS="CellBody"><A NAME="pgfId=592"></A>a&nbsp;!== b</TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=868"></A>a&nbsp;not&nbsp;equal&nbsp;to&nbsp;b,&nbsp;including&nbsp;x&nbsp;and&nbsp;z</TD></TR>
<TR>
<TD><P CLASS="CellBody"><A NAME="pgfId=1063"></A>a&nbsp;&nbsp;== b</TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=1069"></A>a&nbsp;equal&nbsp;to&nbsp;b,&nbsp;result&nbsp;may
be&nbsp;unknown</TD></TR>
<TR>
<TD><P CLASS="CellBody"><A NAME="pgfId=1070"></A>a&nbsp;&nbsp;!= b</TD>
<TD><P CLASS="CellBody"><A NAME="pgfId=1071"></A>a&nbsp;not&nbsp;equal&nbsp;to&nbsp;b,&nbsp;result&nbsp;may
be&nbsp;unknown</TD></TR>
</TABLE>
<P CLASS="Body"><A NAME="pgfId=597"></A>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 <CODE>0</CODE> if comparison
fails, 1 if it succeeds.</P>

<P><P CLASS="Body"><A NAME="pgfId=598"></A>For the<I> logical equality</I>
and <I>logical inequality</I> operators (<CODE> ==</CODE> and <CODE>!=</CODE>
), if either operand contains an <CODE>x</CODE> or a <CODE>z</CODE> , then
the result shall be the unknown value (<CODE> x</CODE> ).</P>

<P><P CLASS="Body"><A NAME="pgfId=599"></A>For the <I>case equality</I>
and <I>case inequality</I> operators(<CODE> ===</CODE> and <CODE>!==</CODE>
), the comparison shall be done just as it is in the procedural case statement
(see 9.5). Bits which are <CODE>x</CODE> or <CODE>z</CODE> shall be included
in the comparison and must match for the result to be true. The result of
these operators shall always be a known value, either <CODE>1</CODE> or
<CODE>0</CODE> .</P>

<P><P CLASS="SubSection"><A NAME="pgfId=601"></A>Logical operators</P>

<P><P CLASS="Body"><A NAME="pgfId=602"></A>The operators <I>logical and</I>
(<CODE> &amp;&amp;</CODE> ) and <I>logical or</I> (||) are logical connectives.
The result of the evaluation of a logical comparison shall be <CODE>1</CODE>
(defined as true), <CODE>0</CODE> (defined as false), or, if the result
is ambiguous, then the result shall be the unknown value (<CODE> x</CODE>
). The precedence of <CODE>&amp;&amp;</CODE> is greater than that of <CODE>||</CODE>
, and both are lower than relational and equality operators.</P>

<P><P CLASS="Body"><A NAME="pgfId=984"></A>A third logical operator is the
unary <I>logical negation</I> operator <CODE>!</CODE> . The negation operator
converts a non-zero or true operand into <CODE>0</CODE> and a zero or false
operand into <CODE>1</CODE> . An ambiguous truth value remains as <CODE>x</CODE>
.</P>

<P><P CLASS="Body"><A NAME="pgfId=983"></A>1. If register <CODE>alpha</CODE>
holds the integer value 237 and <CODE>beta</CODE> holds the value zero,
then the following examples perform as described:</P>

<PRE>regA = alpha &amp;&amp; beta; // regA is set to 0
regB&nbsp;=&nbsp;alpha || beta; // regB is set to 1</PRE>

<P><P CLASS="Body"><A NAME="pgfId=604"></A>2. The following expression performs
a logical and of three sub-expressions without needing any parentheses:</P>

<PRE>a&nbsp;&lt;&nbsp;size-1&nbsp;&amp;&amp;&nbsp;b&nbsp;!=&nbsp;c&nbsp;&amp;&amp;&nbsp;index&nbsp;!=&nbsp;lastone</PRE>

<P><P CLASS="Body"><A NAME="pgfId=606"></A>However, it is recommended for