diff options
| author | Dan Bornstein <danfuzz@android.com> | 2011-03-23 17:53:04 -0700 |
|---|---|---|
| committer | Android Git Automerger <android-git-automerger@android.com> | 2011-03-23 17:53:04 -0700 |
| commit | b1101e847312da8ee98a26e25058f93ba381f1ac (patch) | |
| tree | 2f12fbaacbfeb599ffdc4853d227537230c6fee3 /docs | |
| parent | bf0cce6d471ffcd65161eca6e027e1b66f02f054 (diff) | |
| parent | 486ac3c0a90b05be5af4b82801ef325e83241057 (diff) | |
| download | android_dalvik-b1101e847312da8ee98a26e25058f93ba381f1ac.tar.gz android_dalvik-b1101e847312da8ee98a26e25058f93ba381f1ac.tar.bz2 android_dalvik-b1101e847312da8ee98a26e25058f93ba381f1ac.zip | |
am 486ac3c0: Removing inaccurate and unmaintained docs.
* commit '486ac3c0a90b05be5af4b82801ef325e83241057':
Removing inaccurate and unmaintained docs.
Diffstat (limited to 'docs')
47 files changed, 0 insertions, 4701 deletions
diff --git a/docs/opcodes/opcode-00-nop.html b/docs/opcodes/opcode-00-nop.html deleted file mode 100644 index 726f560ae..000000000 --- a/docs/opcodes/opcode-00-nop.html +++ /dev/null @@ -1,59 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>nop</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>nop</h1> - -<h2>Purpose</h2> - -<p> -Waste cycles. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>00 10x</td> - <td>nop</td> - <td> </td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<p> -None. -</p> - -<h2>Behavior</h2> - -<p> -No externally observable effects, that is, all registers and object state(s) -stay the same. The program counter silently advances to the next instruction. -</p> - -<h2>Exceptions</h2> - -<p> -None. -</p> - -</body> -</html> diff --git a/docs/opcodes/opcode-01-move.html b/docs/opcodes/opcode-01-move.html deleted file mode 100644 index 13c11502e..000000000 --- a/docs/opcodes/opcode-01-move.html +++ /dev/null @@ -1,92 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>move</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>move</h1> - -<h2>Purpose</h2> - -<p> -Move the contents of one non-object register to another. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>01 12x</td> - <td>move vA, vB</td> - <td><code>A:</code> destination register (4 bits)<br/> - <code>B:</code> source register (4 bits)</td> -</tr> -<tr> - <td>02 22x</td> - <td>move/from16 vAA, vBBBB</td> - <td><code>A:</code> destination register (8 bits)<br/> - <code>B:</code> source register (16 bits)</td> -</tr> -<tr> - <td>03 32x</td> - <td>move/16 vAAAA, vBBBB</td> - <td><code>A:</code> destination register (16 bits)<br/> - <code>B:</code> source register (16 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - Both A and B must be valid register indices in the current stack frame. - </li> - <li> - Register vB must be defined. - </li> - <li> - Register vB must not contain a reference value. - </li> - <li> - Register vB must not be part of a register pair. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - The value of register vB is moved to register vA, that is, vA' = vB. - </li> - <li> - If register v(A-1) is the first half of a register pair, register v(A-1)' - becomes undefined. - </li> - <li> - If register v(A+1) is the second half of a register pair, register v(A+1)' - becomes undefined. - </li> -</ul> - -<h2>Exceptions</h2> - -<p> -None. -</p> - -</body> -</html> diff --git a/docs/opcodes/opcode-04-move-wide.html b/docs/opcodes/opcode-04-move-wide.html deleted file mode 100644 index 8a3bd0078..000000000 --- a/docs/opcodes/opcode-04-move-wide.html +++ /dev/null @@ -1,106 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>move-wide</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>move-wide</h1> - -<h2>Purpose</h2> - -<p> -Move the contents of one register-pair to another. -</p> -<p> -Note: It is legal to move from vN to either vN-1 or vN+1, so implementations -must arrange for both halves of a register pair to be read before anything is -written. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>04 12x</td> - <td>move-wide vA, vB</td> - <td><code>A:</code> destination register pair (4 bits)<br/> - <code>B:</code> source register pair (4 bits)</td> -</tr> -<tr> - <td>05 22x</td> - <td>move-wide/from16 vAA, vBBBB</td> - <td><code>A:</code> destination register pair (8 bits)<br/> - <code>B:</code> source register pair (16 bits)</td> -</tr> -<tr> - <td>06 32x</td> - <td>move-wide/16 vAAAA, vBBBB</td> - <td><code>A:</code> destination register pair (16 bits)<br/> - <code>B:</code> source register pair (16 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - Both A+1 and B+1 must be valid register indices in the current stackframe - (which includes A and B being valid). - </li> - <li> - Register vB must be the lower half of a register pair (which excludes the - case of it containing a reference). - </li> - <li> - Both register vB and v(B+1) must be defined. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - The value of register vB is moved to register vA, that is, vA' = vB. - </li> - <li> - The value of register v(B+1) is moved to register v(A+1), that is, v(A+1)' - = v(B+1). - </li> - <li> - If register v(A-1) is the lower half of a register pair, then v(A-1)' - becomes undefined. - </li> - <li> - If register v(A+2) is the upper half of a register pair, then v(A+2)' - becomes undefined. - </li> - <li> - If A = B-1, then v(B+1)' becomes undefined. - </li> - <li> - If A = B+1, then v(B)' becomes undefined. - </li> -</ul> - -<h2>Exceptions</h2> - -<p> -None. -</p> - -</body> -</html> diff --git a/docs/opcodes/opcode-07-move-object.html b/docs/opcodes/opcode-07-move-object.html deleted file mode 100644 index f29027706..000000000 --- a/docs/opcodes/opcode-07-move-object.html +++ /dev/null @@ -1,90 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>move-object</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>move-object</h1> - -<h2>Purpose</h2> - -<p> -Move the contents of one object-bearing register to another. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>07 12x</td> - <td>move-object vA, vB</td> - <td><code>A:</code> destination register (4 bits)<br/> - <code>B:</code> source register (4 bits)</td> -</tr> -<tr> - <td>08 22x</td> - <td>move-object/from16 vAA, vBBBB</td> - <td><code>A:</code> destination register (8 bits)<br/> - <code>B:</code> source register (16 bits)</td> -</tr> -<tr> - <td>09 32x</td> - <td>move-object/16 vAAAA, vBBBB</td> - <td><code>A:</code> destination register (16 bits)<br/> - <code>B:</code> source register (16 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - Both A and B must be legal register indices in the current stack frame. - </li> - <li> - Register vB must be defined. - </li> - <li> - Register vB must contain a reference value (which excludes the case of it - being part of a register pair). - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - The value of register vB is moved to register vA, that is, vA' = vB. - </li> - <li> - If register v(A-1) is the lower half of a register pair, register v(A-1)' - becomes undefined. - </li> - <li> - If register v(A+1) is the upper half of a register pair, register v(A+1)' - becomes undefined. - </li> -</ul> - -<h2>Exceptions</h2> - -<p> -None. -</p> - -</body> -</html> diff --git a/docs/opcodes/opcode-0a-move-result.html b/docs/opcodes/opcode-0a-move-result.html deleted file mode 100644 index 616087f73..000000000 --- a/docs/opcodes/opcode-0a-move-result.html +++ /dev/null @@ -1,96 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>move-result</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>move-result</h1> - -<h2>Purpose</h2> - -<p> -Move the single-word non-object result of the most recent invoke-kind into the -indicated register. This must be done as the instruction immediately after an -invoke-kind whose (single-word, non-object) result is not to be ignored; -anywhere else is invalid. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>0a 11x</td> - <td>move-result vAA</td> - <td><code>A:</code> destination register (8 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - A must be a valid register index in the current stack frame. - </li> - <li> - The instruction must be immediately preceded (in the code array) by an - invoke-kind instruction. - </li> - <li> - The instruction must be immediately reached (in the actual control flow) - through returning from this invoke-kind instruction (it must not be jumped - to). - </li> - <li> - The result delivered by the invoke-kind instruction must not be a reference - value or require a register pair. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - The result delivered by the invoke-kind instruction is moved to register - vA, that is, vA' = result. - </li> - <li> - If register v(A-1) is the lower half of a register pair, register v(A-1)' - becomes undefined. - </li> - <li> - If register v(A+1) is the upper half of a register pair, register v(A+1)' - becomes undefined. - </li> -</ul> - -<h2>Exceptions</h2> - -<p> -None. -</p> - -<h2>Notes</h2> - -<p> -This instruction can also be thought of as reading the contents of a special -"result" register that is made valid and defined by executing a non-void return -instruction or a filled-new-array instruction. The execution of any other -instruction (including this one) renders this special register invalid. -</p> - -</body> -</html> diff --git a/docs/opcodes/opcode-0b-move-result-wide.html b/docs/opcodes/opcode-0b-move-result-wide.html deleted file mode 100644 index c53517aa5..000000000 --- a/docs/opcodes/opcode-0b-move-result-wide.html +++ /dev/null @@ -1,101 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>move-result-wide</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>move-result-wide</h1> - -<h2>Purpose</h2> - -<p> -Move the double-word result of the most recent invoke-kind into the indicated -register pair. This must be done as the instruction immediately after an -invoke-kind whose (double-word) result is not to be ignored; anywhere else is -invalid. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>0b 11x</td> - <td>move-result-wide vAA</td> - <td><code>A:</code> destination register pair (8 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - A+1 must be a valid register index in the current stack frame (which - includes A itself being valid). - </li> - <li> - The instruction must be immediately preceded (in the code array) by an - invoke-kind instruction. - </li> - <li> - The instruction must be immediately reached (in the actual control flow) - through returning from this invoke-kind instruction (it must not be jumped - to). - </li> - <li> - The result delivered by the invoke-kind instruction must be either a long - or a double value. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - The upper 32 bits of the result delivered by the invoke-kind instruction are - moved to register vA, that is, vA' = result >> 0x20. - </li> - <li> - The lower 32 bits of the result delivered by the invoke-kind instruction are - moved to register v(A+1), that is, v(A+1)' = result & 0xffffffff. - </li> - <li> - If register v(A-1) is the lower half of a register pair, register v(A-1)' - becomes undefined. - </li> - <li> - If register v(A+2) is the upper half of a register pair, register v(A+2)' - becomes undefined. - </li> -</ul> - -<h2>Exceptions</h2> - -<p> -None. -</p> - -<h2>Notes</h2> - -<p> -This instruction can also be thought of as reading the contents of a special -"result" register that is made valid and defined by executing a non-void return -instruction or a filled-new-array instruction. The execution of any other -instruction (including this one) renders this special register invalid. -</p> - -</body> -</html> diff --git a/docs/opcodes/opcode-0c-move-result-object.html b/docs/opcodes/opcode-0c-move-result-object.html deleted file mode 100644 index 153873598..000000000 --- a/docs/opcodes/opcode-0c-move-result-object.html +++ /dev/null @@ -1,97 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>move-result-object</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>move-result-object</h1> - -<h2>Purpose</h2> - -<p> -Move the object result of the most recent invoke-kind into the indicated -register. This must be done as the instruction immediately after an invoke-kind -or filled-new-array whose (object) result is not to be ignored; anywhere else -is invalid. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>0c 11x</td> - <td>move-result-object vAA</td> - <td><code>A:</code> destination register (8 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - A must be a valid register index in the current stack frame. - </li> - <li> - The instruction must be immediately preceded (in the code array) by an - invoke-kind, filled-new-array, or filled-new-array/range instruction. - </li> - <li> - The instruction must be immediately reached (in the actual control flow) - through returning from this invoke-kind instruction or by passing a - filled-new-array or filled-new-array/range instruction (it must not be - jumped to). - </li> - <li> - The result delivered by the invoke-kind instruction must be a reference - value (which excludes the case of a long and double values). - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - The result delivered by the invoke-kind instruction is moved to register - vA, that, is vA' = result. - </li> - <li> - If register v(A-1) is the lower half of a register pair, register v(A-1)' - becomes undefined. - </li> - <li> - If register v(A+1) is the upper half of a register pair, register v(A+1)' - becomes undefined. - </li> -</ul> - -<h2>Exceptions</h2> - -<p> -None. -</p> - -<h2>Notes</h2> - -<p> -This instruction can also be thought of as reading the contents of a special -"result" register that is made valid and defined by executing a non-void return -instruction or a filled-new-array instruction. The execution of any other -instruction (including this one) renders this special register invalid. -</p> - -</body> -</html> diff --git a/docs/opcodes/opcode-0d-move-exception.html b/docs/opcodes/opcode-0d-move-exception.html deleted file mode 100644 index 0f756d02e..000000000 --- a/docs/opcodes/opcode-0d-move-exception.html +++ /dev/null @@ -1,79 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>move-exception</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>move-exception</h1> - -<h2>Purpose</h2> - -<p> -Save a just-caught exception into the given register. This should be the first -instruction of any exception handler whose caught exception is not to be -ignored, and this instruction may only ever occur as the first instruction of an -exception handler; anywhere else is invalid. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>0d 11x</td> - <td>move-exception vAA</td> - <td><code>A:</code> destination register (8 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - A must be a valid register index in the current stack frame. - </li> - <li> - The instruction must be the first instruction (in the code array) of an - instruction handler, that is, its offset in the code array must match one of - the handlers defined for the method in the Dex file. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - The active exception of the current thread is moved to register vA, that is, - vA' = exception. - </li> - <li> - If register v(A-1) is the lower half of a register pair, register v(A-1)' - becomes undefined. - </li> - <li> - If register v(A+1) is the upper half of a register pair, register v(A+1)' - becomes undefined. - </li> -</ul> - -<h2>Exceptions</h2> - -<p> -None. -</p> - -</body> -</html> diff --git a/docs/opcodes/opcode-0e-return-void.html b/docs/opcodes/opcode-0e-return-void.html deleted file mode 100644 index 0498f81e1..000000000 --- a/docs/opcodes/opcode-0e-return-void.html +++ /dev/null @@ -1,84 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>return-void</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>return-void</h1> - -<h2>Purpose</h2> - -<p> -Return from a void method. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>0e 10x</td> - <td>return-void</td> - <td> </td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - The return type of the current method must be void. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - If the method is synchronized, the object's monitor is released in a way - similar to the monitor-exit instruction. - </li> - <li> - The stack frame of the current method invocation is removed from the stack. - This includes all its registers becoming invalid. - </li> - <li> - If the stack is now empty, the current thread terminates. - </li> - <li> - Otherwise, the following happens: - <ul> - <li> - The stack frame that caused this method invocation becomes valid. This - includes all its registers and their old values. - </li> - <li> - Execution continues at the bytecode instruction immediately following - the invoke-kind or invoke-kind/range instruction that caused this method - invocation. - </li> - </ul> - </li> -</ul> - -<h2>Exceptions</h2> - -<p> -None. -</p> - -</body> -</html> diff --git a/docs/opcodes/opcode-0f-return.html b/docs/opcodes/opcode-0f-return.html deleted file mode 100644 index 4de55ea9c..000000000 --- a/docs/opcodes/opcode-0f-return.html +++ /dev/null @@ -1,100 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>return</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>return</h1> - -<h2>Purpose</h2> - -<p> -Return from a single-width (32-bit) non-object value-returning method. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>0f 11x</td> - <td>return vAA</td> - <td><code>A:</code> return value register (8 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - The return type of the current method must not be double, long, or a - reference. - </li> - <li> - A must be a valid register index in the current stack frame. - </li> - <li> - Register vA must not be part of a register pair. - </li> - <li> - The type of vA must match the return type of the method. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - If the method is synchronized, the object's monitor is released in a way - similar to the monitor-exit instruction. - </li> - <li> - The stack frame of the current method invocation is removed from the stack. - This includes all its registers becoming invalid. - </li> - <li> - If the stack is now empty, the current thread terminates. - </li> - <li> - Otherwise, the following happens: - <ul> - <li> - The stack frame that caused this method invocation becomes valid. This - includes all its registers and their old values. - </li> - <li> - Execution continues at the bytecode instruction immediately following - the invoke-kind or invoke-kind/range instruction that caused this - method invocation. - </li> - <li> - The return value can be consumed by (exactly) the first instruction - following the invoke-kind or invoke-kind/range instruction that caused - this method invocation, and this instructions needs to be a move-result - instruction. - </li> - </ul> - </li> -</ul> - -<h2>Exceptions</h2> - -<p> -None. -</p> - -</body> -</html> diff --git a/docs/opcodes/opcode-10-return-wide.html b/docs/opcodes/opcode-10-return-wide.html deleted file mode 100644 index 4ccfce43d..000000000 --- a/docs/opcodes/opcode-10-return-wide.html +++ /dev/null @@ -1,99 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>return-wide</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>return-wide</h1> - -<h2>Purpose</h2> - -<p> -Return from a double-width (64-bit) value-returning method. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>10 11x</td> - <td>return-wide vAA</td> - <td><code>A:</code> return value register-pair (8 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - The return type of the current method must be double or long. - </li> - <li> - A+1 must be a valid register index in the current stack frame (which - includes A being valid). - </li> - <li> - Register vA must be the lower half of a register pair. - </li> - <li> - The type of vA must match the return type of the method. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - If the method is synchronized, the object's monitor is released in a way - similar to the monitor-exit instruction. - </li> - <li> - The stack frame of the current method invocation is removed from the stack. - This includes all its registers becoming invalid. - </li> - <li> - If the stack is now empty, the current thread terminates. - </li> - <li> - Otherwise, the following happens: - <ul> - <li> - The stack frame that caused this method invocation becomes valid. This - includes all its registers and their old values. - </li> - <li> - Execution continues at the bytecode instruction immediately following - the invoke instruction that caused this method invocation. - </li> - <li> - The return value can be consumed by (exactly) the first instruction - following the invoke-kind or invoke-kind/range instruction that caused - this method invocation, and this instructions needs to be a - move-result-wide instruction. - </li> - </ul> - </li> -</ul> - -<h2>Exceptions</h2> - -<p> -None. -</p> - -</body> -</html> diff --git a/docs/opcodes/opcode-11-return-object.html b/docs/opcodes/opcode-11-return-object.html deleted file mode 100644 index b4866ed70..000000000 --- a/docs/opcodes/opcode-11-return-object.html +++ /dev/null @@ -1,96 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>return-object</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>return-object</h1> - -<h2>Purpose</h2> - -<p> -Return from an object-returning method. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>11 11x</td> - <td>return-object vAA</td> - <td><code>A:</code> return value register (8 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - The return type of the current method must be a reference. - </li> - <li> - A must be a valid register index in the current stack frame. - </li> - <li> - Register vA must be known to be reference-bearing. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - If the method is synchronized, the object's monitor is released in a way - similar to the monitor-exit instruction. - </li> - <li> - The stack frame of the current method invocation is removed from the stack. - This includes all its registers becoming invalid. - </li> - <li> - If the stack is now empty, the current thread terminates. - </li> - <li> - Otherwise, the following happens: - </li> - <ul> - <li> - The stack frame that caused this method invocation becomes valid. This - includes all its registers and their old values. - </li> - <li> - Execution continues at the bytecode instruction immediately following - the invoke instruction that caused this method invocation. - </li> - <li> - The return value can be consumed by (exactly) the first instruction - following the invoke-kind or invoke-kind/range instruction that caused - this method invocation, and this instructions needs to be a - move-result-object instruction. - </li> - </ul> - </li> -</ul> - -<h2>Exceptions</h2> - -<p> -None. -</p> - -</body> -</html> diff --git a/docs/opcodes/opcode-12-const.html b/docs/opcodes/opcode-12-const.html deleted file mode 100644 index d2b6ef9eb..000000000 --- a/docs/opcodes/opcode-12-const.html +++ /dev/null @@ -1,102 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>const</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>const</h1> - -<h2>Purpose</h2> - -<p> -Move the given literal value (sign-extended to 32 bits, if necessary) into the -specified register. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>12 11n</td> - <td>const/4 vA, #+B</td> - <td><code>A:</code> destination register (4 bits)<br/> - <code>B:</code> signed int (4 bits)</td> -</tr> -<tr> - <td>13 21s</td> - <td>const/16 vAA, #+BBBB</td> - <td><code>A:</code> destination register (8 bits)<br/> - <code>B:</code> signed int (16 bits)</td> -</tr> -<tr> - <td>14 31i</td> - <td>const vAA, #+BBBBBBBB</td> - <td><code>A:</code> destination register (8 bits)<br/> - <code>B:</code> arbitrary 32-bit constant</td> -</tr> -<tr> - <td>15 21h</td> - <td>const/high16 vAA, #+BBBB0000</td> - <td><code>A:</code> destination register (8 bits)<br/> - <code>B:</code> signed int (16 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - A must be a valid register index in the current stackframe. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - First, an adjusted value B' is determined as follows: - <ul> - <li> - If we are executing the /high16 variant, then B is left-shifted by 16 - bits, that is, B'=B << 0x10 - <li> - Otherwise, if B is a 4 bit or 16 bit constant, it is sign-extended to 32 - bits, that is, B'=sign-extended(B). - </li> - <li> - Otherwise, B'=B. - </li> - </ul> - <li> - Then, the adjusted value B' is moved into the register A, that is, vA'=B' - </li> - <li> - If v(A-1) is the lower half of a register pair, v(A-1)' becomes undefined. - </li> - <li> - If v(A+1) is the upper half of a register pair, v(A+1)' becomes undefined. - </li> -</ul> - -<h2>Exceptions</h2> - -<p> -None. -</p> - -</body> -</html> diff --git a/docs/opcodes/opcode-16-const-wide.html b/docs/opcodes/opcode-16-const-wide.html deleted file mode 100644 index 6197e3556..000000000 --- a/docs/opcodes/opcode-16-const-wide.html +++ /dev/null @@ -1,110 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>const-wide</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>const-wide</h1> - -<h2>Purpose</h2> - -<p> -Move the given literal value (sign-extended to 64 bits) into the specified -register-pair. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>16 21s</td> - <td>const-wide/16 vAA, #+BBBB</td> - <td><code>A:</code> destination register (8 bits)<br/> - <code>B:</code> signed int (16 bits)</td> -</tr> -<tr> - <td>17 31i</td> - <td>const-wide/32 vAA, #+BBBBBBBB</td> - <td><code>A:</code> destination register (8 bits)<br/> - <code>B:</code> signed int (32 bits)</td> -</tr> -<tr> - <td>18 51l</td> - <td>const-wide vAA, #+BBBBBBBBBBBBBBBB</td> - <td><code>A:</code> destination register (8 bits)<br/> - <code>B:</code> arbitrary double-width (64-bit) constant</td> -</tr> -<tr> - <td>19 21h</td> - <td>const-wide/high16 vAA, #+BBBB000000000000</td> - <td><code>A:</code> destination register (8 bits)<br/> - <code>B:</code> signed int (16 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - A must be a valid register index in the current stack frame. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - First, an adjusted value B' is determined as follows: - <ul> - <li> - If we are executing the /high16 variant, then B is left-shifted by 40 - bits, that is, B'=B << 0x28 - <li> - Otherwise, if B is a 16 bit or 32 bit constant, it is sign-extended to - 64 bits, that is, B'=sign-extended(B). - </li> - <li> - Otherwise, B'=B. - </li> - </ul> - <li> - The immediate value B is moved into the register pair (vA, v(A+1)), that is, - <ul> - <li> - vA' = B << 0x20 - </li> - <li> - v(A+1)' = B & 0xffffffff - </li> - </ul> - </li> - <li> - If v(A-1) is the lower half of a register pair, v(A-1)' becomes undefined. - </li> - <li> - If v(A+2) is the upper half of a register pair, v(A+2)' becomes undefined. - </li> -</ul> - -<h2>Exceptions</h2> - -<p> -None. -</p> - -</body> -</html> diff --git a/docs/opcodes/opcode-1a-const-string.html b/docs/opcodes/opcode-1a-const-string.html deleted file mode 100644 index d10c11584..000000000 --- a/docs/opcodes/opcode-1a-const-string.html +++ /dev/null @@ -1,85 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>const-string</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>const-string</h1> - -<h2>Purpose</h2> - -<p> -Move a reference to the string specified by the given index into the specified -register. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>1a 21c</td> - <td>const-string vAA, string@BBBB</td> - <td><code>A:</code> destination register (8 bits)<br/> - <code>B:</code> string index</td> -</tr> -<tr> - <td>1b 31c</td> - <td>const-string/jumbo vAA, string@BBBBBBBB</td> - <td><code>A:</code> destination register (8 bits)<br/> - <code>B:</code> string index</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - A must be a valid register index in the current stack frame. - </li> - <li> - B must be a valid index into the string constant pool. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - A new java.lang.String object S is allocated on the heap and filled with the - contents of string pool entry B. - </li> - <li> - A reference to an internalized version of the new object is moved into - register vA, that is, the instruction behaves as if vA' = S.intern() was - called. - </li> - <li> - If v(A-1) is the lower half of a register pair, v(A-1)' becomes undefined. - </li> - <li> - If v(A+1) is the upper half of a register pair, v(A+1)' becomes undefined. - </li> -</ul> - -<h2>Exceptions</h2> - -<p> -None. -</p> - -</body> -</html> diff --git a/docs/opcodes/opcode-1b-const-class.html b/docs/opcodes/opcode-1b-const-class.html deleted file mode 100644 index f40b986b4..000000000 --- a/docs/opcodes/opcode-1b-const-class.html +++ /dev/null @@ -1,92 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>const-class</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>const-class</h1> - -<h2>Purpose</h2> - -<p> -Move a reference to the class specified by the given index into the specified -register. In the case where the indicated type is primitive, this will store a -reference to the primitive type's degenerate class. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>1c 21c</td> - <td>const-class vAA, type@BBBB</td> - <td><code>A:</code> destination register (8 bits)<br/> - <code>B:</code> type index</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - A must be a valid register index in the current stack frame. - </li> - <li> - B must be a valid index into the type constant pool. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - An attempt is made to get a reference to the class C the name of which is - contained in type pool entry B. - </li> - <li> - If B refers to a primitive type, the corresponding degenerate class is used - instead. - </li> - <li> - If C has not been loaded and resolved before, it is being loaded and - resolved. All exceptions that are possible during class loading can occur at - this point. - </li> - <li> - A reference to C is moved into register vA, that is, vA' = C. - </li> - <li> - If v(A-1) is the lower half of a register pair, v(A-1)' becomes undefined. - </li> - <li> - If v(A+1) is the upper half of a register pair, v(A+1)' becomes undefined. - </li> -</ul> - -<h2>Exceptions</h2> - -<ul> - <li> - ClassNotFoundException is thrown if the class does not exist at all. - </li> - <li> - VerifyError is thrown if the class does exist, but could not be verified. - </li> -</ul> - -</body> -</html> diff --git a/docs/opcodes/opcode-1d-monitor-enter.html b/docs/opcodes/opcode-1d-monitor-enter.html deleted file mode 100644 index 28c10f414..000000000 --- a/docs/opcodes/opcode-1d-monitor-enter.html +++ /dev/null @@ -1,93 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>monitor-enter</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>monitor-enter</h1> - -<h2>Purpose</h2> - -<p> -Acquire the monitor for the indicated object. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>1d 11x</td> - <td>monitor-enter vAA</td> - <td><code>A:</code> reference-bearing register (8 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - A must be a valid register index for the current stack frame. - </li> - <li> - Register vA must contain a reference to an object. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - An attempt is made for the current thread to acquire the monitor of the - indicated object. Various results are possible: - <ul> - <li> - If the monitor is not owned by any thread at this point, then the - current thread becomes owner of the monitor. The entry count of the - indicated object is set to 1. - </li> - <li> - Otherwise, if the monitor is owned by the same thread that attempts the - acquiration, then the entry count of the indicated object is increased - by 1. - </li> - <li> - Otherwise the monitor is owned by a different thread. The current thread - sleeps until the monitor of the object is released. Once that happens, a - new attempt to acquire the monitor is made, as described here. There is - no guarantee that the second attempt (or any subsequent attempt) will be - successful. - </li> - </ul> - </li> -</ul> - -<h2>Exceptions</h2> - -<ul> - <li> - NullPointerException if vA is null. - </li> - <li> - IllegalMonitorStateException if the entry count exceeds an - (implementation-dependent) upper bound for recursive monitor entries. Note - that it is unlikely this bound is ever hit, since for most implementations - the call stack will overflow before. - </li> -</ul> - -</body> -</html> diff --git a/docs/opcodes/opcode-1e-monitor-exit.html b/docs/opcodes/opcode-1e-monitor-exit.html deleted file mode 100644 index cd7b16506..000000000 --- a/docs/opcodes/opcode-1e-monitor-exit.html +++ /dev/null @@ -1,101 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>monitor-exit</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>monitor-exit</h1> - -<h2>Purpose</h2> - -<p> -Release the monitor for the indicated object. -</p> -<p> -Note: If this instruction needs to throw an exception, it must do so as if the -pc has already advanced past the instruction. It may be useful to think of this -as the instruction successfully executing (in a sense), and the exception -getting thrown after the instruction but before the next one gets a chance to -run. This definition makes it possible for a method to use a monitor cleanup -catch-all (e.g., finally) block as the monitor cleanup for that block itself, -as a way to handle the arbitrary exceptions that might get thrown due to the -historical implementation of Thread.stop(), while still managing to have proper -monitor hygiene. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>1e 11x</td> - <td>monitor-exit vAA</td> - <td><code>A:</code> reference-bearing register (8 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - A must be a valid register index for the current stack frame. - </li> - <li> - Register vA must contain a reference to an object. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - An attempt is made for the current thread to release the monitor of the - indicated object. - </li> - <li> - If the current thread is the owner, the following happens: - <ul> - <li> - The monitor's entry count is decreased by one. - </li> - <li> - If the entry count has reached zero, the monitor is released. Other - threads waiting for the same monitor have a chance to acquire it. - </li> - </ul> - </li> - <li> - Any exception that gets thrown by this instruction bears the PC of the - instruction following the monitor-exit. That is, from the point of view of - an exception handler it cannot be distinguished from the same type of - exception being thrown immediately after the monitor-exit instruction. - </li> -</ul> - -<h2>Exceptions</h2> - -<ul> - <li> - NullPointerException is thrown if vA is null. - </li> - <li> - IllegalMonitorStateException is thrown if the current thread is not the - owner of that monitor. - </li> -</ul> - -</body> -</html> diff --git a/docs/opcodes/opcode-1f-check-cast.html b/docs/opcodes/opcode-1f-check-cast.html deleted file mode 100644 index 8eedd2dd7..000000000 --- a/docs/opcodes/opcode-1f-check-cast.html +++ /dev/null @@ -1,95 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>check-cast</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>check-cast</h1> - -<h2>Purpose</h2> - -<p> -Throw if the reference in the given register cannot be cast to the indicated -type. The type must be a reference type (not a primitive type). -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>1f 21c</td> - <td>check-cast vAA, type@BBBB</td> - <td><code>A:</code> reference-bearing register (8 bits)<br/> - <code>B:</code> type index (16 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - A must be a valid register index in the current stack frame. - </li> - <li> - Register vA must contain a reference value. - </li> - <li> - B must be a valid index into the type pool. - </li> - <li> - Type pool entry B must contain a valid type descriptor for a reference type. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - An attempt is made to get a reference to the class C the name of which is - contained in type pool entry B. - </li> - <li> - If C has not been loaded and resolved before, it is being loaded and - resolved. All exceptions that are possible during class loading can occur at - this point. - </li> - <li> - The run-time type of the object reference vA is compared against C. - <ul> - <li> - If vA is null, the instruction succeeds (without further effects). - </li> - <li> - If vA is assignment compatible with C according to the usual rules of - the Java programming language, the instruction succeeds (without further - effects). - </li> - </ul> - </li> -</ul> - -<h2>Exceptions</h2> - -<ul> - <li> - ClassCastException is thrown if vA is either not null or not assignment - compatible with C. - </li> -</ul> - -</body> -</html> diff --git a/docs/opcodes/opcode-20-instance-of.html b/docs/opcodes/opcode-20-instance-of.html deleted file mode 100644 index 88076d84d..000000000 --- a/docs/opcodes/opcode-20-instance-of.html +++ /dev/null @@ -1,100 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>instance-of</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>instance-of</h1> - -<h2>Purpose</h2> - -<p> -Store in the given destination register 1 if the indicated reference is an -instance of the given type, or 0 if not. The type must be a reference type (not -a primitive type). -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>20 22c</td> - <td>instance-of vA, vB, type@CCCC</td> - <td><code>A:</code> destination register (4 bits)<br/> - <code>B:</code> reference-bearing register (4 bits)<br/> - <code>C:</code> type index (16 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - Both A and B must be a valid register indices for the current stack frame. - </li> - <li> - Register vB must contain a reference value. - </li> - <li> - C must be a valid index into the type constant pool. - </li> - <li> - Type constant pool entry C must contain a valid type descriptor for a - reference type. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - An attempt is made to get a reference to the class K the name of which is - contained in type pool entry C. - </li> - <li> - If K has not been loaded and resolved before, it is being loaded and - resolved. All exceptions that are possible during class loading can occur at - this point. - </li> - <li> - The run-time type of the object reference vB is compared against K. The - register vA reflects the result: - <ul> - <li> - vA' = 1 if (and only if) vB is not null and vB is assignment compatible - with K according to the usual rules of the Java programming language. - </li> - <li> - vA' = 0 otherwise - </li> - </ul> - </li> - <li> - If v(A-1) is the lower half of a register pair, v(A-1)' becomes undefined. - </li> - <li> - If v(A+1) is the upper half of a register pair, v(A+1)' becomes undefined. - </li> - -<h2>Exceptions</h2> - -<p> -None. -</p> - -</body> -</html> diff --git a/docs/opcodes/opcode-21-array-length.html b/docs/opcodes/opcode-21-array-length.html deleted file mode 100644 index 8072a7cb4..000000000 --- a/docs/opcodes/opcode-21-array-length.html +++ /dev/null @@ -1,78 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>array-length</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>array-length</h1> - -<h2>Purpose</h2> - -<p> -Store in the given destination register the length of the indicated array, -in entries. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>21 12x</td> - <td>array-length vA, vB</td> - <td><code>A:</code> destination register (4 bits)<br/> - <code>B:</code> array reference-bearing register (4 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - Both A and B must be valid register indices for the current stack frame. - </li> - <li> - Register vB must contain a reference to an array. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - The length of the array referenced by vB is stored in vA, that is - vA' = length(vB). - </li> - <li> - If register v(A-1) is the lower half of a register pair, register v(A-1)' - becomes undefined. - </li> - <li> - If register v(A+1) is the upper half of a register pair, register v(A+1)' - becomes undefined. - </li> -</ul> - -<h2>Exceptions</h2> - -<ul> - <li> - NullPointerException is thrown if the value of register vB is null. - </li> -</ul> - -</body> -</html> diff --git a/docs/opcodes/opcode-22-new-instance.html b/docs/opcodes/opcode-22-new-instance.html deleted file mode 100644 index bdcfc3efe..000000000 --- a/docs/opcodes/opcode-22-new-instance.html +++ /dev/null @@ -1,95 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>new-instance</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>new-instance</h1> - -<h2>Purpose</h2> - -<p> -Construct a new instance of the indicated type, storing a reference to it in the -destination. The type must refer to a non-array class. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>22 21c</td> - <td>new-instance vAA, type@BBBB</td> - <td><code>A:</code> destination register (8 bits)<br/> - <code>B:</code> type index</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - A must be a valid register index for the current stack frame. - </li> - <li> - B must be a valid index into the type pool. - </li> - <li> - Type constant pool entry B must contain a valid type descriptor for a - non-array class. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - An attempt is made to get a reference to the class C the name of which is - contained in type pool entry B. - </li> - <li> - If C has not been loaded and resolved before, it is being loaded and - resolved. All exceptions that are possible during class loading can occur at - this point. - </li> - <li> - An attempt is made to create a new instance I of C. All exceptions that are - possible during instantiation can occur at this point. - </li> - <li> - A reference to the new instance is stored in register vA, that is vA' = I. - </li> - <li> - If v(A-1) is the lower part of a register pair, v(A-1)' becomes undefined. - </li> - <li> - If v(A+1) is the upper part of a register pair, v(A+1)' becomes undefined. - </li> -</ul> - -<h2>Exceptions</h2> - -<ul> - <li> - All exceptions that are possible during class loading can occur. - </li> - <li> - All exceptions that are possible during instantiation can occur. - </li> -</ul> - -</body> -</html> diff --git a/docs/opcodes/opcode-23-new-array.html b/docs/opcodes/opcode-23-new-array.html deleted file mode 100644 index 29327e921..000000000 --- a/docs/opcodes/opcode-23-new-array.html +++ /dev/null @@ -1,108 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>new-array</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>new-array</h1> - -<h2>Purpose</h2> - -<p> -Construct a new array of the indicated type and size. The type must be an array -type. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>23 22c</td> - <td>new-array vA, vB, type@CCCC</td> - <td><code>A:</code> destination register (8 bits)<br/> - <code>B:</code> size register<br/> - <code>C:</code> type index</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - A and B must be valid register indices for the current stack frame. - </li> - <li> - Register vB must not contain a reference value. - </li> - <li> - Register vB must not be part of a register pair. - </li> - <li> - C must be a valid index into the type pool. - </li> - <li> - Type constant pool entry C must contain a valid array type descriptor. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - An attempt is made to get a reference to the class K the name of which is - contained in type pool entry C. - </li> - <li> - If K has not been loaded and resolved before, it is being loaded and - resolved. All exceptions that are possible during class loading can occur at - this point. - </li> - <li> - An attempt is made to create a new instance I of K and length B. All - exceptions that are possible during instantiation can occur at this point. - </li> - <li> - All elements of the new array are initialized to null (for object arrays) or - 0 (for numeric arrays) or false (for boolean arrays). - </li> - <li> - A reference to the new array is moved to register vA, that is, vA' = I. - </li> - <li> - If v(A-1) is the lower half of a register pair, v(A-1)' becomes undefined. - </li> - <li> - If v(A+1) is the upper half of a register pair, v(A+1)' becomes undefined. - </li> -</ul> - -<h2>Exceptions</h2> - -<ul> - <li> - NegativeArraySizeException if vB < 0 - </li> - <li> - All exceptions that are possible during class loading can occur. - </li> - <li> - All exceptions that are possible during instantiation can occur. - </li> -</ul> - -</body> -</html> diff --git a/docs/opcodes/opcode-24-filled-new-array.html b/docs/opcodes/opcode-24-filled-new-array.html deleted file mode 100644 index 1dfa0896f..000000000 --- a/docs/opcodes/opcode-24-filled-new-array.html +++ /dev/null @@ -1,144 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>filled-new-array</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>filled-new-array</h1> - -<h2>Purpose</h2> - -<p> -Construct an array of the given type and size, filling it with the supplied -contents. The type must be an array type. The array's contents must be -single-word (that is, no arrays of long or double). The constructed instance is -stored as a "result" in the same way that the method invocation instructions -store their results, so the constructed instance must be moved to a register -with a subsequent move-result-object instruction (if it is to be used). -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>24 35c</td> - <td>filled-new-array {vD, vE, vF, vG, vA}, type@CCCC</td> - <td><code>B:</code> array size and argument word count (4 bits)<br/> - <code>C:</code> type index (16 bits)<br/> - <code>D..G, A:</code> argument registers (4 bits each)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - The value B must not be greater than 5. - </li> - <li> - If B > 0, then D must be a valid register index in the current stack frame. - </li> - <li> - If B > 1, then E must be a valid register index in the current stack frame. - </li> - <li> - If B > 2, then F must be a valid register index in the current stack frame. - </li> - <li> - If B > 3, then G must be a valid register index in the current stack frame. - </li> - <li> - If B > 4, then A must be a valid register index in the current stack frame. - </li> - <li> - C must be a valid index into the type pool. - </li> - <li> - The type denoted by C must be a valid array type descriptor. - </li> - <li> - The element size of the type denoted by C must be no larger than 32 bits. - </li> - <li> - If the element type is a primitive type, then all actual arguments - (vD .. vA, depending on B) must be primitive, too. - </li> - <li> - If the element type is a reference type, then all actual arguments - (vD .. vA, depending on B) must be references, too. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - An attempt is made to get a reference to the type T the name of which is - contained in type pool entry C. - </li> - <li> - If T is a reference type and it has not been loaded and resolved before, it - is being loaded and resolved. All exceptions that are possible during class - loading can occur at this point. - </li> - <li> - An attempt is made to create a new array R of type T and length B. All - exceptions that are possible during instantiation can occur at this point. - </li> - <li> - The elements of R are filled according to the following rules: - <ul> - <li> - If B > 0 then R[0] = vD - </li> - <li> - If B > 1 then R[1] = vE - </li> - <li> - If B > 2 then R[2] = vF - </li> - <li> - If B > 3 then R[3] = vG - </li> - <li> - If B > 4 then R[4] = vA - </li> - </ul> - </li> - <li> - No reference to R is stored in any register. Instead, R can be accessed by a - move-result-object instruction immediately following this filled-new-array - instruction. - </li> -</ul> - -<h2>Exceptions</h2> - -<ul> - <li> - NegativeArraySizeException if vB < 0 - </li> - <li> - All exceptions that are possible during class loading can occur. - </li> - <li> - All exceptions that are possible during instantiation can occur. - </li> -</ul> - -</body> -</html> diff --git a/docs/opcodes/opcode-25-filled-new-array-range.html b/docs/opcodes/opcode-25-filled-new-array-range.html deleted file mode 100644 index 2ee750562..000000000 --- a/docs/opcodes/opcode-25-filled-new-array-range.html +++ /dev/null @@ -1,125 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>filled-new-array/range</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>filled-new-array/range</h1> - -<h2>Purpose</h2> - -<p> -Construct an array of the given type and size, filling it with the supplied -contents. Clarifications and restrictions are the same as filled-new-array, -described above. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>25 3rc</td> - <td>filled-new-array/range {vCCCC .. vNNNN}, type@BBBB</td> - <td><code>A:</code> array size and argument word count (8 bits)<br/> - <code>B:</code> type index (16 bits)<br/> - <code>C:</code> first argument register (16 bits)<br/> - <code>N = A + C - 1</code></td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - vN must be a valid register index in the current stack frame (this means - everything below vN is valid, too). - </li> - <li> - For all values I in the interval [C .. N] the following must hold: - <ul> - <li> - vI must not be part of a register pair - </li> - <li> - If the array type is a simple type, vI must be a simple type, too. - </li> - <li> - If the array type is a reference type, vI must be a reference type, too. - </li> - </ul> - </li> - <li> - B must be a valid index into the type pool. - </li> - <li> - The type denoted by B must be an array type. - </li> - <li> - The element size of the type denoted by B must be no larger than 32 bits. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - An attempt is made to get a reference to the type T the name of which is - contained in type pool entry B. - </li> - <li> - If T has not been loaded and resolved before, it is being loaded and - resolved. All exceptions that are possible during class loading can occur at - this point. - </li> - <li> - An attempt is made to create a new instance J of type T and length vA. All - exceptions that are possible during instantiation can occur at this point. - </li> - <li> - The elements of R are filled according to the following rules: - <ul> - <li> - J[0] = vC - </li> - <li> - J[1] = v(C+1) - </li> - <li> - ... - </li> - <li> - J[vA] = vN - </li> - </ul> - </li> - <li> - No reference to J is stored in any register. Instead, J can be accessed by a - move-result-object instruction immediately following this filled-new-array - instruction. - </li> -</ul> - -<h2>Exceptions</h2> - -<ul> - <li> - NegativeArraySizeException if vA < 0 - </li> -</ul> - -</body> -</html> diff --git a/docs/opcodes/opcode-26-fill-array-data.html b/docs/opcodes/opcode-26-fill-array-data.html deleted file mode 100644 index 77b45ae0c..000000000 --- a/docs/opcodes/opcode-26-fill-array-data.html +++ /dev/null @@ -1,96 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>fill-array-data</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>fill-array-data</h1> - -<h2>Purpose</h2> - -<p> -Fill the given array with the indicated data. The reference must be to an array -of primitives, and the data table must match it in type and size. -</p> -<p> -Note: The address of the table is guaranteed to be even (that is, 4-byte -aligned). If the code size of the method is otherwise odd, then an extra code -unit is inserted between the main code and the table whose value is the same as -a nop. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> - <td>26 31t</td> - <td>fill-array-data vAA, +BBBBBBBB <i>(with supplemental data as specified - below in "<code>fill-array-data</code> Format")</i></td> - <td><code>A:</code> array reference (8 bits)<br/> - <code>B:</code> signed "branch" offset to table data (32 bits)</td> - </td> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - A must be a valid register index in the current stack frame. - </li> - <li> - vA must be a reference-bearing register (according to data flow) and contain - an array-reference. - </li> - <li> - B must be branch offset in the same method. - </li> - <li> - The target address (PC+B) must be 4-byte aligned. - </li> - <li> - The target address must hold the pseudo-opcode 0x300. - </li> - <li> - The table entry size must match the size of the data type of the array. - </li> - <li> - The table size must be equal or smaller than the array length. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - The contents of the array referenced by vA are filled with the table data, - starting from array index 0 and in the given order. - </li> - <li> - If there are less elements in the table than the array provides space for, - the remaining array elements stay untouched. - </li> -</ul> - -<h2>Exceptions</h2> - -<ul> - <li> - NullPointerException if vA is null. - </li> -</ul> - -</body> -</html> diff --git a/docs/opcodes/opcode-27-throw.html b/docs/opcodes/opcode-27-throw.html deleted file mode 100644 index 1a0eb0937..000000000 --- a/docs/opcodes/opcode-27-throw.html +++ /dev/null @@ -1,80 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>throw</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>throw</h1> - -<h2>Purpose</h2> - -<p> -Throw the indicated exception. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>27 11x</td> - <td>throw vAA</td> - <td><code>A:</code> exception-bearing register (8 bits)<br/></td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - A must be a valid register index in the current stack frame. - </li> - <li> - Register vA must be a reference-bearing register - </li> - <li> - Register vA must be assignment-compatible with java.lang.Throwable according - to the usual rules of the Java programming language. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - Throws the given exception vA, resulting in a search for a matching handler - according to the usual rules of the Java programming language. - </li> - <li> - If no matching handler is found for the current thread, the thread - terminates, possibly notifying its uncaught exception handler or thread - group before. - </li> -</ul> - -<h2>Exceptions</h2> - -<ul> - <li> - NullPointerException if vA is null. - </li> - <li> - Otherwise, the indicated exception. - </li> -</ul> - -</body> -</html> diff --git a/docs/opcodes/opcode-28-goto.html b/docs/opcodes/opcode-28-goto.html deleted file mode 100644 index fec294c1b..000000000 --- a/docs/opcodes/opcode-28-goto.html +++ /dev/null @@ -1,73 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>goto</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>goto</h1> - -<h2>Purpose</h2> - -<p> -Unconditionally jump to the indicated instruction. -</p> -<p> -Note: The branch offset may not be 0. (A spin loop may be legally constructed -either with goto/32 or by including a nop as a target before the branch.) -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>28 10t</td> - <td>goto +AA</td> - <td><code>A:</code> signed branch offset (8 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - A must point to a valid bytecode instruction inside the current method. - </li> - <li> - A must not be 0. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - The (otherwise invisible) program counter PC is set to the address of the - instruction plus the given offset, that is, PC' = PC(goto) + A. - </li> - <li> - Executions resumes at PC'. - </li> -</ul> - -<h2>Exceptions</h2> - -<p> -None. -</p> - -</body> -</html> diff --git a/docs/opcodes/opcode-29-goto-16.html b/docs/opcodes/opcode-29-goto-16.html deleted file mode 100644 index 791456b60..000000000 --- a/docs/opcodes/opcode-29-goto-16.html +++ /dev/null @@ -1,73 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>goto/16</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>goto/16</h1> - -<h2>Purpose</h2> - -<p> -Unconditionally jump to the indicated instruction. -</p> -<p> -Note: The branch offset may not be 0. (A spin loop may be legally constructed -either with goto/32 or by including a nop as a target before the branch.) -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>29 20t</td> - <td>goto/16 +AAAA</td> - <td><code>A:</code> signed branch offset (16 bits)<br/></td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - A must point to a valid bytecode instruction inside the current method. - </li> - <li> - A must not be 0. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - The (otherwise invisible) program counter PC is set to the address of the - instruction plus the given offset, that is, PC' = PC(goto) + A. - </li> - <li> - Executions resumes at PC'. - </li> -</ul> - -<h2>Exceptions</h2> - -<p> -None. -</p> - -</body> -</html> diff --git a/docs/opcodes/opcode-2a-goto-32.html b/docs/opcodes/opcode-2a-goto-32.html deleted file mode 100644 index b98dd8565..000000000 --- a/docs/opcodes/opcode-2a-goto-32.html +++ /dev/null @@ -1,66 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>goto/32</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>goto/32</h1> - -<h2>Purpose</h2> - -<p> -Unconditionally jump to the indicated instruction. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>2a 30t</td> - <td>goto/32 +AAAAAAAA</td> - <td><code>A:</code> signed branch offset (32 bits)<br/></td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - A must point to a valid bytecode instruction inside the current method. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - The (otherwise invisible) program counter PC is set to the address of the - instruction plus the given offset, that is, PC' = PC(goto) + A. - </li> - <li> - Executions resumes at PC'. - </li> -</ul> - -<h2>Exceptions</h2> - -<p> -None. -</p> - -</body> -</html> diff --git a/docs/opcodes/opcode-2b-packed-switch.html b/docs/opcodes/opcode-2b-packed-switch.html deleted file mode 100644 index b2d525138..000000000 --- a/docs/opcodes/opcode-2b-packed-switch.html +++ /dev/null @@ -1,103 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>packed-switch</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>packed-switch</h1> - -<h2>Purpose</h2> - -<p> -Jump to a new instruction based on the value in the given register, using a -table of offsets corresponding to each value in a particular integral range, or -fall through to the next instruction if there is no match. -</p> -<p> -Note: The address of the table is guaranteed to be even (that is, 4-byte -aligned). If the code size of the method is otherwise odd, then an extra code -unit is inserted between the main code and the table whose value is the same as -a nop. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>2b 31t</td> - <td>packed-switch vAA, +BBBBBBBB <i>(with supplemental data as - specified below in "<code>packed-switch</code> Format")</i></td> - <td><code>A:</code> register to test<br/> - <code>B:</code> signed "branch" offset to table data (32 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - A must be a valid register index in the current stack frame. - </li> - <li> - Let PC be the address of the packed-switch instruction in the code array of - the current method. Then T = PC + B with the following properties: - <ul> - <li> - T must be 4-byte-aligned. - </li - <li> - T must be in the same method. - </li> - <li> - T must point to a packed-switch data table. - </li> - </ul> - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - The value of vA is used as an index into the given table data. - </li> - <li> - If vA is in the range of the table, that is, if vA >= table.first_key and - vA < first_key + size, then the jump target is determined as follows: - <ul> - <li> - PC' = PC + table.targets[vA - table.firstKey]. - </li> - <li> - Execution resumes at this address. - </li> - </ul> - </li> - <li> - Otherwise execution continues at the instruction following the packed-switch - statement. - </li> -</ul> - -<h2>Exceptions</h2> - -<p> -None. -</p> - -</body> -</html> diff --git a/docs/opcodes/opcode-2c-sparse-switch.html b/docs/opcodes/opcode-2c-sparse-switch.html deleted file mode 100644 index 9d81eda45..000000000 --- a/docs/opcodes/opcode-2c-sparse-switch.html +++ /dev/null @@ -1,110 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>sparse-switch</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>sparse-switch</h1> - -<h2>Purpose</h2> - -<p> -Jump to a new instruction based on the value in the given register, using an -ordered table of value-offset pairs, or fall through to the next instruction if -there is no match. -</p> -<p> -Note: The address of the table is guaranteed to be even (that is, 4-byte -aligned). If the code size of the method is otherwise odd, then an extra code -unit is inserted between the main code and the table whose value is the same as -a nop. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>2c 31t</td> - <td>sparse-switch vAA, +BBBBBBBB <i>(with supplemental data as - specified below in "<code>sparse-switch</code> Format")</i></td> - <td><code>A:</code> register to test<br/> - <code>B:</code> signed "branch" offset to table data (32 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - A must be a valid register index in the current stack frame. - </li> - <li> - Let PC be the address of the packed-switch instruction in the code array of - the current method. Then T = PC + B with the following properties: - <ul> - <li> - T must be 4-byte-aligned. - </li> - <li> - T must be in the same method. - </li> - <li> - T must point to a sparse-switch data table. - </li> - </ul> - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - The value of vA is used as a lookup key inside the sparse table data. - </li> - <li> - If there exists an I with 0 <= I < table.size such that table.keys[I] = vA, - then the jump target is determined as follows: - <ul> - <li> - PC' = PC + table.targets[I]. - </li> - <li> - Execution will resume at this address. - </li> - </ul> - </li> - <li> - Otherwise execution continues at the instruction following the sparse-switch - statement. - </li> -</ul> - -<h2>Exceptions</h2> - -<p> -None. -</p> - -<h2>Notes</h2> - -<p> -The low-to-high ordering of the keys allows the VM to employ binary search for -the lookup, resulting in O(log table.size) comparisons. -</p> - -</body> -</html> diff --git a/docs/opcodes/opcode-2d-cmp-kind.html b/docs/opcodes/opcode-2d-cmp-kind.html deleted file mode 100644 index f55a006c9..000000000 --- a/docs/opcodes/opcode-2d-cmp-kind.html +++ /dev/null @@ -1,121 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>cmp<kind></title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>cmp<kind></h1> - -<h2>Purpose</h2> - -<p> -Perform the indicated floating point or long comparison, storing 0 if the two -arguments are equal, 1 if the second argument is larger, or -1 if the first -argument is larger. The "bias" listed for the floating point operations -indicates how NaN comparisons are treated: "Gt bias" instructions return 1 for -NaN comparisons, and "lt bias" instructions return -1. -</p> -<p> -For example, to check to see if floating point a < b, then it is advisable to -use cmpg-float; a result of -1 indicates that the test was true, and the other -values indicate it was false either due to a valid comparison or because one -or the other values was NaN. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>2d..31 23x</td> - <td>cmp<i>kind</i> vAA, vBB, vCC<br/> - 2d: cmpl-float <i>(lt bias)</i><br/> - 2e: cmpg-float <i>(gt bias)</i><br/> - 2f: cmpl-double <i>(lt bias)</i><br/> - 30: cmpg-double <i>(gt bias)</i><br/> - 31: cmp-long - </td> - <td><code>A:</code> destination register (8 bits)<br/> - <code>B:</code> first source register or pair<br/> - <code>C:</code> second source register or pair</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - A, B and C must be valid register indices in the current stack frame. - </li> - <li> - For the two -float variants, both vB and vC must be of type float. - </li> - <li> - For the two -double variants, both vB and vC must be the lower part of a - register pair holding a double value. - </li> - <li> - For the -long variant, both both vB and vC must be the lower part of a - register pair holding a long value. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - The values of registers vB and vC are compared. The result, which is stored - in vA, is one of the following three: - <ul> - <li> - If vB < vC, then vA'=-1. - </li> - <li> - If vB == vC, then vA'=0. - </li> - <li> - If vC > vC, then vA'=1. - </li> - </ul> - </li> - <li> - For the -float and -double variants, an addition "bias" specifies what - happens if one or both of the arguments are NaN: - <ul> - <li> - A "lt bias" results in vA'=-1. - </li> - <li> - A "gt bias" results in vA'=1. - </li> - </ul> - </li> - <li> - If v(A-1) is the lower half of a register pair, v(A-1)' becomes undefined. - </li> - <li> - If v(A+1) is the upper half of a register pair, v(A+1)' becomes undefined. - </li> -</ul> - -<h2>Exceptions</h2> - -<p> -None. -</p> - -</body> -</html> diff --git a/docs/opcodes/opcode-32-if-test.html b/docs/opcodes/opcode-32-if-test.html deleted file mode 100644 index ee394f6d1..000000000 --- a/docs/opcodes/opcode-32-if-test.html +++ /dev/null @@ -1,101 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>if-test</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>if-test</h1> - -<h2>Purpose</h2> - -<p> -Branch to the given destination if the given two registers' values compare as -specified. -</p> -<p> -Note: The branch offset may not be 0. (A spin loop may be legally constructed -either by branching around a backward goto or by including a nop as a target -before the branch.) -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>32..37 22t</td> - <td>if-<i>test</i> vA, vB, +CCCC<br/> - 32: if-eq<br/> - 33: if-ne<br/> - 34: if-lt<br/> - 35: if-ge<br/> - 36: if-gt<br/> - 37: if-le<br/> - </td> - <td><code>A:</code> first register to test (4 bits)<br/> - <code>B:</code> second register to test (4 bits)<br/> - <code>C:</code> signed branch offset (16 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - A and B must be valid register indices for the current stack frame. - </li> - <li> - Registers vA and vB must not contain a reference value. - </li> - <li> - Registers vA and vB must not be part of a register pair. - </li> - <li> - Registers vA and vB must not contain a floating point value (???). - </li> - C must of a signed offset that, when added to the PC of the instruction, - points to a valid bytecode instruction inside the same method. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - The values of registers vA and vB are compared according to the <test> - condition. Two results are possible: - <ul> - <li> - The condition holds. The value of C is used as a signed offset to the - address of the if-<test> instruction. Execution continues at the - resulting address. - </li> - <li> - The condition does not hold. Execution continues at the instruction - following the if-<test> instruction. - </li> - </ul> - </li> -</ul> - -<h2>Exceptions</h2> - -<p> -None. -</p> - -</body> -</html> diff --git a/docs/opcodes/opcode-38-if-testz.html b/docs/opcodes/opcode-38-if-testz.html deleted file mode 100644 index 060bbdb60..000000000 --- a/docs/opcodes/opcode-38-if-testz.html +++ /dev/null @@ -1,100 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>if-<test>z</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>if-<test>z</h1> - -<h2>Purpose</h2> - -<p> -Branch to the given destination if the given register's value compares with 0 -as specified. -</p> -<p> - Note: The branch offset may not be 0. (A spin loop may be legally constructed - either by branching around a backward goto or by including a nop as a target - before the branch.) -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>38..3d 21t</td> - <td>if-<i>test</i>z vAA, +BBBB<br/> - 38: if-eqz<br/> - 39: if-nez<br/> - 3a: if-ltz<br/> - 3b: if-gez<br/> - 3c: if-gtz<br/> - 3d: if-lez<br/> - </td> - <td><code>A:</code> register to test (8 bits)<br/> - <code>B:</code> signed branch offset (16 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - A must be a valid register index for the current stackframe. - </li> - <li> - Register vA must not contain a reference value. - </li> - <li> - Register vA must not be part of a register pair. - </li> - <li> - Register vA must not contain a floating point value (???). - </li> - <li> - B must not be 0. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - The value of register vA is compared to zero according to the <test> - condition. Two results are possible: - <ul> - <li> - The condition holds. The value of B is used as a signed offset to the - address of the if-<test>z instruction. Execution continues at the - resulting address. - </li> - <li> - The condition does not hold. Execution continues at the instruction - following the if-<test>z instruction. - </li> - </ul> - </li> -</ul> - -<h2>Exceptions</h2> - -<p> -None. -</p> - -</body> -</html> diff --git a/docs/opcodes/opcode-44-aget.html b/docs/opcodes/opcode-44-aget.html deleted file mode 100644 index 6e8836f3b..000000000 --- a/docs/opcodes/opcode-44-aget.html +++ /dev/null @@ -1,113 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>aget<kind></title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>aget<kind></h1> - -<h2>Purpose</h2> - -<p> -Perform the identified array operation at the identified index of the given -array, storing into the value register. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>44..51 23x</td> - <td><i>arrayop</i> vAA, vBB, vCC<br/> - 44: aget<br/> - 45: aget-wide<br/> - 46: aget-object<br/> - 47: aget-boolean<br/> - 48: aget-byte<br/> - 49: aget-char<br/> - 4a: aget-short<br/> - </td> - <td><code>A:</code> dest value register or pair; (8 bits)<br/> - <code>B:</code> array register (8 bits)<br/> - <code>C:</code> index register (8 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - A, B and C must be valid register indices in the current stackframe. - </li> - <li> - For the aget-wide variant, also A+1 must be a valid register index in the - current stackframe. - </li> - <li> - Register vB must contain an array reference. The component type of the - array must match the variant of the instruction. - </li> - <li> - Register vC must contain an integer value. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - For all but the -wide variant, the array element at the given index is moved - into register vA, that is, vA'=array[index]. - </li> - <li> - For the -wide variant, the array element at the given index is moved into - registers vA and v(A+1) as follows: - <ul> - <li> - vA'=array[index] >> 0x20 - </li> - <li> - v(A+1)'=array[index] & 0xffffffff; - </li> - </ul> - </li> - <li> - If v(A-1) is the lower half of a register pair, v(A-1)' becomes undefined. - </li> - <li> - For all but the -wide variant, if v(A+1) is the upper half of a register - pair, v(A+1)' becomes undefined. - </li> - <li> - For the -wide variant, if v(A+2) is the upper half of a register pair, - v(A+2)' becomes undefined. - </li> -</ul> - -<h2>Exceptions</h2> - -<ul> - <li> - NullPointerException if vB=null. - </li> - <li> - ArrayIndexOutOfBoundsException if vC < 0 or vC >= array.length. - </li> -</ul> - -</body> -</html> diff --git a/docs/opcodes/opcode-4b-aput.html b/docs/opcodes/opcode-4b-aput.html deleted file mode 100644 index 089c1cad9..000000000 --- a/docs/opcodes/opcode-4b-aput.html +++ /dev/null @@ -1,98 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>aput<kind></title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>aput<kind></h1> - -<h2>Purpose</h2> - -<p> -Waste cycles. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>4b..51 23x</td> - <td><i>arrayop</i> vAA, vBB, vCC<br/> - 4b: aput<br/> - 4c: aput-wide<br/> - 4d: aput-object<br/> - 4e: aput-boolean<br/> - 4f: aput-byte<br/> - 50: aput-char<br/> - 51: aput-short - </td> - <td><code>A:</code> source value register or pair; (8 bits)<br/> - <code>B:</code> array register (8 bits)<br/> - <code>C:</code> index register (8 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - A, B and C must be valid register indices in the current stack frame. - </li> - <li> - For the aget-wide variant, also A+1 must be a valid register index in the - current stack frame. - </li> - <li> - Register vB must contain an array reference. The component type of the array - must match the variant of the instruction. - </li> - <li> - Register vC must contain an integer value. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - For all but the -wide variant, the value of register vA is move into the - array element at the given index, that is, array[index]'=vA. - </li> - <li> - For the -wide variant, the registers vA and v(A+1) are moved into the array - element at the given index as follows: - <ul> - <li> - array[index]' = vA << 0x20 | v(A+1) - </li> - </ul> - </li> -</ul> - -<h2>Exceptions</h2> - -<ul> - <li> - NullPointerException if vB=null. - </li> - <li> - ArrayIndexOutOfBoundsException if vC < 0 or vC >= array.length. - </li> -</ul> - -</body> -</html> diff --git a/docs/opcodes/opcode-52-iget.html b/docs/opcodes/opcode-52-iget.html deleted file mode 100644 index 837b511ce..000000000 --- a/docs/opcodes/opcode-52-iget.html +++ /dev/null @@ -1,109 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>iget<kind></title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>iget<kind></h1> - -<h2>Purpose</h2> - -<p> -Perform the identified object instance field operation with the identified -field, loading or storing into the value register. -</p> -<p> -Note: These opcodes are reasonable candidates for static linking, altering the -field argument to be a more direct offset. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>52..58 22c</td> - <td>i<i>instanceop</i> vA, vB, field@CCCC<br/> - 52: iget<br/> - 53: iget-wide<br/> - 54: iget-object<br/> - 55: iget-boolean<br/> - 56: iget-byte<br/> - 57: iget-char<br/> - 58: iget-short<br/> - </td> - <td><code>A:</code> dest value register or pair; (4 bits)<br/> - <code>B:</code> object register (4 bits)<br/> - <code>C:</code> instance field reference index (16 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - A and B must be valid register indices in the current stackframe. - </li> - <li> - For the -wide variant, also A+1 must be a valid register index in the - current stackframe. - </li> - <li> - Register vB must contain an object reference. - </li> - <li> - C must be a valid index into the field reference pool. - </li> - <li> - The field must be an instance field. The type of the field denoted by C must - match the variant of the instruction. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - The value of the given instance field is read from the given object and - moved into the given register vA, that is, vA'=<object>.<field>. - </li> - <li> - If v(A-1) is the lower half of a register pair, v(A-1)' becomes undefined. - </li> - <li>For all but the -wide variant, if v(A+1) is the upper half of a register - pair, v(A+1)' becomes undefined. - </li> - <li> - For the -wide variant, if v(A+2) is the upper half of a register pair, - v(A+2)' becomes undefined. - </li> -</ul> - -<h2>Exceptions</h2> - -<ul> - <li> - NullPointerException if object is null. - </li> - <li> - IllegalAccessException if <object>.<field> is not visible from - the current context according to the usual visibility and access rules of - the Java programming language. - </li> -</ul> - -</body> -</html> diff --git a/docs/opcodes/opcode-59-iput.html b/docs/opcodes/opcode-59-iput.html deleted file mode 100644 index 22a3479b5..000000000 --- a/docs/opcodes/opcode-59-iput.html +++ /dev/null @@ -1,112 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>iget<kind></title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>iget<kind></h1> - -<h2>Purpose</h2> - -<p> -Perform the identified object instance field operation with the identified -field, loading or storing into the value register. -</p> -<p> -Note: These opcodes are reasonable candidates for static linking, altering the -field argument to be a more direct offset. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>59..5f 22c</td> - <td>i<i>instanceop</i> vA, vB, field@CCCC<br/> - 59: iput<br/> - 5a: iput-wide<br/> - 5b: iput-object<br/> - 5c: iput-boolean<br/> - 5d: iput-byte<br/> - 5e: iput-char<br/> - 5f: iput-short - </td> - <td><code>A:</code> source value register or pair; (4 bits)<br/> - <code>B:</code> object register (4 bits)<br/> - <code>C:</code> instance field reference index (16 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - A and B must be valid register indices in the current stack frame. - </li> - <li> - For the -wide variant, also A+1 must be a valid register index in the - current stack frame. - </li> - <li> - Register vB must contain an object reference. - </li> - <li> - C must be a valid index into the field reference pool. - </li> - <li> - The field must be an instance field. The type of the field denoted by C must - match the variant of the instruction. - </li> - <li> - For the -object variant, the instance referenced by register vA must be - assignment-compatible to the type of the field. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - For all but the -wide variant, the value of register vA is move into the - field, that is, <object>.<field>'=vA. - </li> - <li> - For the -wide variant, the registers vA and v(A+1) are moved into the - field as follows: - <ul> - <li> - <object>.<field>' = vA << 0x20 | v(A+1) - </li> - </ul> - </li> -</ul> - -<h2>Exceptions</h2> - -<ul> - <li> - NullPointerException if vB=null. - </li> - <li> - IllegalAccessException if <object>.<field> is not visible from - the current context according to the usual visibility and access rules of - the Java programming language, or final. - </li> -</ul> - -</body> -</html> diff --git a/docs/opcodes/opcode-60-sget.html b/docs/opcodes/opcode-60-sget.html deleted file mode 100644 index 820886ef1..000000000 --- a/docs/opcodes/opcode-60-sget.html +++ /dev/null @@ -1,106 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>sget<kind></title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>sget<kind></h1> - -<h2>Purpose</h2> - -<p> -Perform the identified object static field operation with the identified static -field, loading or storing into the value register. -</p> -<p> -Note: These opcodes are reasonable candidates for static linking, altering the -field argument to be a more direct offset. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>60..6d 21c</td> - <td>s<i>staticop</i> vAA, field@BBBB<br/> - 60: sget<br/> - 61: sget-wide<br/> - 62: sget-object<br/> - 63: sget-boolean<br/> - 64: sget-byte<br/> - 65: sget-char<br/> - 66: sget-short<br/> - </td> - <td><code>A:</code> dest value register or pair; (8 bits)<br/> - <code>B:</code> static field reference index (16 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - A must be a valid register index in the current stackframe. - </li> - <li> - For the -wide variant, also A+1 must be a valid register index in the - current stackframe. - </li> - <li> - B must be a valid index into the field reference pool. - </li> - <li> - The field denoted by B must be static. The type of the field denoted by B - must match the variant of the instruction. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - The value of the given instance field is read from the given object and - moved into the given register vA, that is, vA'=<class>.<field>. - </li> - <li> - If v(A-1) is the lower half of a register pair, v(A-1)' becomes undefined. - </li> - <li> - For all but the -wide variant, if v(A+1) is the upper half of a register - pair, v(A+1)' becomes undefined. - </li> - <li> - For the -wide variant, if v(A+2) is the upper half of a register pair, - v(A+2)' becomes undefined. - </li> -</ul> - -<h2>Exceptions</h2> - -<ul> - <li> - NullPointerException if object is null. - </li> - <li> - IllegalAccessException if <object>.<field> is not visible from - the current context according to the usual visibility and access rules of - the Java programming language. - </li> -</ul> - -</body> -</html> diff --git a/docs/opcodes/opcode-67-sput.html b/docs/opcodes/opcode-67-sput.html deleted file mode 100644 index b4d88bba3..000000000 --- a/docs/opcodes/opcode-67-sput.html +++ /dev/null @@ -1,108 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>sput<kind></title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>sput<kind></h1> - -<h2>Purpose</h2> - -<p> -Perform the identified object static field operation with the identified static -field, loading or storing into the value register. -</p> -<p> -Note: These opcodes are reasonable candidates for static linking, altering the -field argument to be a more direct offset. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>67..6d 21c</td> - <td>s<i>staticop</i> vAA, field@BBBB<br/> - 67: sput<br/> - 68: sput-wide<br/> - 69: sput-object<br/> - 6a: sput-boolean<br/> - 6b: sput-byte<br/> - 6c: sput-char<br/> - 6d: sput-short - </td> - <td><code>A:</code> source value register or pair; (8 bits)<br/> - <code>B:</code> static field reference index (16 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - A must be a valid register index in the current stackframe. - </li> - <li> - For the -wide variant, also A+1 must be a valid register index in the - current stackframe. - </li> - <li> - B must be a valid index into the field reference pool. - </li> - <li> - The field must be static. The type of the field denoted by C must match the - variant of the instruction. - </li> - <li> - For the -object variant, the instance referenced by register vA must be - assignment-compatible to the type of the field. - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - For all but the -wide variant, the value of register vA is move into the - field, that is, <class>.<field>'=vA. - </li> - <li> - For the -wide variant, the registers vA and v(A+1) are moved into the field - as follows: - <ul> - <li> - <class>.<field>' = vA << 0x20 | v(A+1) - </li> - </ul> - </li> -</ul> - -<h2>Exceptions</h2> - -<ul> - <li> - NullPointerException if vB=null. - </li> - <li> - IllegalAccessException if <object>.<field> is not visible from - the current context according to the usual visibility and access rules of - the Java programming language, or final. - </li> -</ul> - -</body> -</html> diff --git a/docs/opcodes/opcode-7b-unop.html b/docs/opcodes/opcode-7b-unop.html deleted file mode 100644 index 8b0609210..000000000 --- a/docs/opcodes/opcode-7b-unop.html +++ /dev/null @@ -1,108 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>unop</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>unop</h1> - -<h2>Purpose</h2> - -<p> -Perform the identified unary operation on the source register, storing the -result in the destination register. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>7b..8f 12x</td> - <td><i>unop</i> vA, vB<br/> - 7b: neg-int<br/> - 7c: not-int<br/> - 7d: neg-long<br/> - 7e: not-long<br/> - 7f: neg-float<br/> - 80: neg-double<br/> - 81: int-to-long<br/> - 82: int-to-float<br/> - 83: int-to-double<br/> - 84: long-to-int<br/> - 85: long-to-float<br/> - 86: long-to-double<br/> - 87: float-to-int<br/> - 88: float-to-long<br/> - 89: float-to-double<br/> - 8a: double-to-int<br/> - 8b: double-to-long<br/> - 8c: double-to-float<br/> - 8d: int-to-byte<br/> - 8e: int-to-char<br/> - 8f: int-to-short - </td> - <td><code>A:</code> destination register or pair (4 bits)<br/> - <code>B:</code> source register or pair (4 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - Both A and B must be valid register indices for the current stackframe. - </li> - <li> - If the input type of <unop> is double or long, also B+1 must be a - valid register index in the current stackframe. - </li> - <li> - If the output type of <unop> is double or long, also A+1 must be a - valid register index in the current stackframe. - </li> - <li> - The type of register vB must match the source type of the instruction (this - probably needs more detail). - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - The given operation <unop> is performed according to the semantics - specified in table XXX. - </li> - <li> - The result is stored in register vA, that is, vA'=<unop> vB. - </li> - <li> - It gets a bit messy if we want to describe all the combinations of input and - output with and without pairs here. Probably it's better to split it up. - </li> -</ul> - -<h2>Exceptions</h2> - -<ul> - <li> - ArithmeticException if an arithmetic error occurs during the instruction. - </li> -</ul> - -</body> -</html> diff --git a/docs/opcodes/opcode-90-binop.html b/docs/opcodes/opcode-90-binop.html deleted file mode 100644 index cdc08a8f4..000000000 --- a/docs/opcodes/opcode-90-binop.html +++ /dev/null @@ -1,120 +0,0 @@ -<!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 & 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 <binop> is performed according to the semantics - specified in table XXX. - </li> - <li> - The result is stored in register vA, that is, vA'=<biop> 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> diff --git a/docs/opcodes/opcode-b0-binop-2addr.html b/docs/opcodes/opcode-b0-binop-2addr.html deleted file mode 100644 index b3374f4a0..000000000 --- a/docs/opcodes/opcode-b0-binop-2addr.html +++ /dev/null @@ -1,120 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>binop/2addr</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>binop/2addr</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 & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>b0..cf 12x</td> - <td><i>binop</i>/2addr vA, vB<br/> - b0: add-int/2addr<br/> - b1: sub-int/2addr<br/> - b2: mul-int/2addr<br/> - b3: div-int/2addr<br/> - b4: rem-int/2addr<br/> - b5: and-int/2addr<br/> - b6: or-int/2addr<br/> - b7: xor-int/2addr<br/> - b8: shl-int/2addr<br/> - b9: shr-int/2addr<br/> - ba: ushr-int/2addr<br/> - bb: add-long/2addr<br/> - bc: sub-long/2addr<br/> - bd: mul-long/2addr<br/> - be: div-long/2addr<br/> - bf: rem-long/2addr<br/> - c0: and-long/2addr<br/> - c1: or-long/2addr<br/> - c2: xor-long/2addr<br/> - c3: shl-long/2addr<br/> - c4: shr-long/2addr<br/> - c5: ushr-long/2addr<br/> - c6: add-float/2addr<br/> - c7: sub-float/2addr<br/> - c8: mul-float/2addr<br/> - c9: div-float/2addr<br/> - ca: rem-float/2addr<br/> - cb: add-double/2addr<br/> - cc: sub-double/2addr<br/> - cd: mul-double/2addr<br/> - ce: div-double/2addr<br/> - cf: rem-double/2addr - </td> - <td><code>A:</code> destination and first source register or pair - (4 bits)<br/> - <code>B:</code> second source register or pair (4 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - Both A and B must be valid register indices in the current stackframe. - </li> - <li> - For the -long and -double variants, also A+1 and B+1 must be valid register - indices. - </li> - <li> - Registers vA and vB 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 <binop> is performed according to the semantics - specified in table XXX. - </li> - <li> - The result is stored in register vA, that is, vA'=vA <binop> 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> diff --git a/docs/opcodes/opcode-d0-binop-lit16.html b/docs/opcodes/opcode-d0-binop-lit16.html deleted file mode 100644 index f9d3327b9..000000000 --- a/docs/opcodes/opcode-d0-binop-lit16.html +++ /dev/null @@ -1,94 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>binop/lit16</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>binop/lit16</h1> - -<h2>Purpose</h2> - -<p> -Perform the indicated binary op on the indicated register (first argument) and -literal value (second argument), storing the result in the destination register. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>d0..d7 22s</td> - <td><i>binop</i>/lit16 vA, vB, #+CCCC<br/> - d0: add-int/lit16<br/> - d1: rsub-int (reverse subtract)<br/> - d2: mul-int/lit16<br/> - d3: div-int/lit16<br/> - d4: rem-int/lit16<br/> - d5: and-int/lit16<br/> - d6: or-int/lit16<br/> - d7: xor-int/lit16 - </td> - <td><code>A:</code> destination register (4 bits)<br/> - <code>B:</code> source register (4 bits)<br/> - <code>C:</code> signed int constant (16 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - Both A and B must be valid register indices in the current stackframe. - </li> - <li> - Registers vA and vB must be defined. They must both contain integer values. - </li> - <li> - C is an immediate, signed integer constant taken from the instruction stream - (actually this means there are no special requirements for C at all). - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - The given operation <binop> is performed according to the semantics - specified in table XXX. - </li> - <li> - Argument C is sign-extended to 32 bits before. - </li> - <li> - The result is stored in register vA, that is, vA'=vB <binop> vC. - </li> - <li> - As usual, neighboring registers might get undefined, if vA was 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> diff --git a/docs/opcodes/opcode-d8-binop-lit8.html b/docs/opcodes/opcode-d8-binop-lit8.html deleted file mode 100644 index 26005e9fc..000000000 --- a/docs/opcodes/opcode-d8-binop-lit8.html +++ /dev/null @@ -1,97 +0,0 @@ -<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> - -<html> - -<head> -<title>binop/lit8</title> -<link rel=stylesheet href="opcode.css"> -</head> - -<body> - -<h1>binop/lit8</h1> - -<h2>Purpose</h2> - -<p> -Perform the indicated binary op on the indicated register (first argument) and -literal value (second argument), storing the result in the destination register. -</p> - -<h2>Details</h2> - -<table class="instruc"> -<thead> -<tr> - <th>Op & Format</th> - <th>Mnemonic / Syntax</th> - <th>Arguments</th> -</tr> -</thead> -<tbody> -<tr> - <td>d8..e2 22b</td> - <td><i>binop</i>/lit8 vAA, vBB, #+CC<br/> - d8: add-int/lit8<br/> - d9: rsub-int/lit8<br/> - da: mul-int/lit8<br/> - db: div-int/lit8<br/> - dc: rem-int/lit8<br/> - dd: and-int/lit8<br/> - de: or-int/lit8<br/> - df: xor-int/lit8<br/> - e0: shl-int/lit8<br/> - e1: shr-int/lit8<br/> - e2: ushr-int/lit8 - </td> - <td><code>A:</code> destination register (8 bits)<br/> - <code>B:</code> source register (8 bits)<br/> - <code>C:</code> signed int constant (8 bits)</td> -</tr> -</tbody> -</table> - -<h2>Constraints</h2> - -<ul> - <li> - Both A and B must be valid register indices in the current stackframe. - </li> - <li> - Registers vA and vB must be defined. They must both contain integer values. - </li> - <li> - C is an immediate, signed integer constant taken from the instruction stream - (actually this means there are no special requirements for C at all). - </li> -</ul> - -<h2>Behavior</h2> - -<ul> - <li> - The given operation <binop> is performed according to the semantics - specified in table XXX. - </li> - <li> - Argument C is sign-extended to 32 bits before. - </li> - <li> - The result is stored in register vA, that is, vA'=vB <binop> vC. - </li> - <li> - As usual, neighboring registers might get undefined, if vA was 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> diff --git a/docs/opcodes/opcode.css b/docs/opcodes/opcode.css deleted file mode 100644 index c3c130424..000000000 --- a/docs/opcodes/opcode.css +++ /dev/null @@ -1,166 +0,0 @@ -h1 { - font-family: serif; - color: #222266; -} - -h2 { - font-family: serif; - border-top-style: solid; - border-top-width: 2px; - border-color: #ccccdd; - padding-top: 12px; - margin-top: 48px; - margin-bottom: 2px; - color: #222266; -} - -@media print { - table { - font-size: 8pt; - } -} - -@media screen { - table { - font-size: 10pt; - } -} - - -/* general for all tables */ - -table { - border-collapse: collapse; - margin-top: 12px; -} - -table th { - font-family: sans-serif; - background: #aabbff; - text-align: left; -} - -table td { - font-family: sans-serif; - border-top-style: solid; - border-bottom-style: solid; - border-width: 1px; - border-color: #aaaaff; - padding-top: 4px; - padding-bottom: 4px; - padding-left: 4px; - padding-right: 6px; - background: #eeeeff; -} - -table td p { - margin-top: 4pt; - margin-bottom: 0pt; -} - - - -/* opcodes table */ - -table.instruc { - margin-top: 24px; - margin-bottom: 24px; - margin-left: 48px; - margin-right: 48px; -} - -table.instruc td { - font-family: sans-serif; - border-top-style: solid; - border-bottom-style: solid; - border-width: 1px; - padding-top: 4px; - padding-bottom: 4px; - padding-left: 2px; - padding-right: 2px; -} - -table.instruc td:first-child { - font-family: monospace; - font-size: 90%; - vertical-align: top; - width: 12%; -} - -table.instruc td:first-child + td { - font-family: monospace; - font-size: 90%; - vertical-align: top; - width: 23%; -} - -table.instruc td:first-child + td i { - font-family: sans-serif; - font-size: 90%; -} - -table.instruc td:first-child + td + td { - vertical-align: top; - width: 28%; -} - -table.instruc td:first-child + td + td + td { - vertical-align: top; - width: 37%; -} - - -/* supplemental opcode format table */ - -table.supplement { - margin-top: 24px; - margin-bottom: 24px; - margin-left: 48px; - margin-right: 48px; -} - -table.supplement td:first-child { - font-family: monospace; - vertical-align: top; - width: 20%; -} - -table.supplement td:first-child + td { - font-family: monospace; - vertical-align: top; - width: 20%; -} - -table.supplement td:first-child + td + td { - font-family: sans-serif; - vertical-align: top; - width: 60%; -} - - -/* math details table */ - -table.math { - margin-top: 24px; - margin-bottom: 24px; - margin-left: 48px; - margin-right: 48px; -} - -table.math td:first-child { - font-family: monospace; - vertical-align: top; - width: 10%; -} - -table.math td:first-child + td { - font-family: monospace; - vertical-align: top; - width: 30%; -} - -table.math td:first-child + td + td { - font-family: sans-serif; - vertical-align: top; - width: 60%; -} |