blob: cdc08a8f4ae22315e059543a9957f02ee8ff002d [file] [log] [blame]
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
<html>
<head>
<title>binop</title>
<link rel=stylesheet href="opcode.css">
</head>
<body>
<h1>binop</h1>
<h2>Purpose</h2>
<p>
Perform the identified binary operation on the two source registers, storing
the result in the first source register.
</p>
<h2>Details</h2>
<table class="instruc">
<thead>
<tr>
<th>Op &amp; Format</th>
<th>Mnemonic / Syntax</th>
<th>Arguments</th>
</tr>
</thead>
<tbody>
<tr>
<td>90..af 23x</td>
<td><i>binop</i> vAA, vBB, vCC<br/>
90: add-int<br/>
91: sub-int<br/>
92: mul-int<br/>
93: div-int<br/>
94: rem-int<br/>
95: and-int<br/>
96: or-int<br/>
97: xor-int<br/>
98: shl-int<br/>
99: shr-int<br/>
9a: ushr-int<br/>
9b: add-long<br/>
9c: sub-long<br/>
9d: mul-long<br/>
9e: div-long<br/>
9f: rem-long<br/>
a0: and-long<br/>
a1: or-long<br/>
a2: xor-long<br/>
a3: shl-long<br/>
a4: shr-long<br/>
a5: ushr-long<br/>
a6: add-float<br/>
a7: sub-float<br/>
a8: mul-float<br/>
a9: div-float<br/>
aa: rem-float<br/>
ab: add-double<br/>
ac: sub-double<br/>
ad: mul-double<br/>
ae: div-double<br/>
af: rem-double
</td>
<td><code>A:</code> destination register or pair (8 bits)<br/>
<code>B:</code> first source register or pair (8 bits)<br/>
<code>C:</code> second source register or pair (8 bits)</td>
</tr>
</tbody>
</table>
<h2>Constraints</h2>
<ul>
<li>
All A, B and C must be valid register indices in the current stackframe.
</li>
<li>
For the -long and -double variants, also A+1, B+1 and C+1 must be valid
register indices.
</li>
<li>
Registers vB and vC must be defined. They must both contain values that
match the variant of the instruction (it's probably better to split this up
into multiple pages again).
</li>
</ul>
<h2>Behavior</h2>
<ul>
<li>
The given operation &lt;binop&gt; is performed according to the semantics
specified in table XXX.
</li>
<li>
The result is stored in register vA, that is, vA'=&lt;biop&gt; vB.
</li>
<li>
For the -double and -long variants, (vA+1) is also affected.
</li>
<li>
As usual, neighboring registers might get undefined, if vA (and vA+1) were
part of a register pair originally.
</li>
</ul>
<h2>Exceptions</h2>
<ul>
<li>
ArithmeticException if an error occurs during the instruction.
</li>
</ul>
</body>
</html>