字节码格式
字节码是JVM的机器语言。JVM加载类文件时，对类中的每个方法，它都会得到一个字节码流。这些字节码流保存在JVM的方法区中。在程序运行过程中，当一个方法被调用时，它的字节码流就会被执行。根据特定JVM设计者的选择，它们可以通过解释的方式，即时编译（Just-in-time compilation）的方式或其他技术的方式被执行。
方法的字节码流就是JVM的指令（instruction）序列。每条指令包含一个单字节的操作码（opcode）和0个或多个操作数（operand）。操作码指明要执行的操作。如果JVM在执行操作前，需要更多的信息，这些信息会以0个或多个操作数的方式，紧跟在操作码的后面。
每种类型的操作码都有一个助记符（mnemonic）。类似典型的汇编语言风格，Java字节码流可以用它们的助记符和紧跟在后面的操作数来表示。例如，下面的字节码流可以分解成多个助记符的形式。
// 字节码流: 03 3b 84 00 01 1a 05 68 3b a7 ff f9
// 分解后:
iconst_0 // 03
istore_0 // 3b
iinc 0, 1 // 84 00 01
iload_0 // 1a
iconst_2 // 05
imul // 68
istore_0 // 3b
goto -7 // a7 ff f9
字节码指令集被设计的很紧凑。除了处理跳表的2条指令以外，所有的指令都以字节边界对齐。操作码的总数很少，一个字节就能搞定。这最小化了JVM加载前，通过网络传输的类文件的大小；也使得JVM可以维持很小的实现。
JVM中，所有的计算都是围绕栈（stack）而展开的。因为JVM没有存储任意数值的寄存器（register），所有的操作数在计算开始之前，都必须先压入栈中。因此，字节码指令主要是用来操作栈的。例如，在上面的字节码序列中，通过iload_0先把本地变量（local variable）入栈，然后用iconst_2把数字2入栈的方式，来计算本地变量乘以2。两个整数都入栈之后，imul指令有效的从栈中弹出它们，然后做乘法，最后把运算结果压入栈中。istore_0指令把结果从栈顶弹出，保存回本地变量。JVM被设计成基于栈，而不是寄存器的机器，这使得它在如80486寄存器架构不佳的处理器上，也能被高效的实现。
原始类型（primitive types）
JVM支持7种原始数据类型。Java程序员可以声明和使用这些数据类型的变量，而Java字节码，处理这些数据类型。下表列出了这7种原始数据类型：
类型 | 定义 | byte单字节有符号二进制补码整数short2字节有符号二进制补码整数int4字节有符号二进制补码整数long8字节有符号二进制补码整数float4字节IEEE 754单精度浮点数double8字节IEEE 754双精度浮点数char2字节无符号Unicode字符原始数据类型以操作数的方式出现在字节码流中。所有长度超过1字节的原始类型，都以大端（big-endian）的方式保存在字节码流中，这意味着高位字节出现在低位字节之前。例如，为了把常量值256（0x0100）压入栈中，你可以用sipush操作码，后跟一个短操作数。短操作数会以“01 00”的方式出现在字节码流中，因为JVM是大端的。如果JVM是小端（little-endian）的，短操作数将会是“00 01”。
// Bytecode stream: 17 01 00
// Dissassembly:
sipush 256; // 17 01 00
把常量（constants）压入栈中
很多操作码都可以把常量压入栈中。操作码以3中不同的方式指定入栈的常量值：由操作码隐式指明，作为操作数跟在操作码之后，或者从常量池（constant pool）中获取。
有些操作码本身就指明了要入栈的数据类型和常量数值。例如，iconst_1告诉JVM把整数1压入栈中。这种操作码，是为不同类型而经常入栈的数值而定义的。它们在字节码流中只占用1个字节，增进了字节码的执行效率，并减小了字节码流的大小。下表列出了int型和float型的操作码：
操作码 | 操作数 | 描述 | iconst_m1(none)pushes int -1 onto the stackiconst_0(none)pushes int 0 onto the stackiconst_1(none)pushes int 1 onto the stackiconst_2(none)pushes int 2 onto the stackiconst_3(none)pushes int 3 onto the stackiconst_4(none)pushes int 4 onto the stackiconst_5(none)pushes int 5 onto the stackfconst_0(none)pushes float 0 onto the stackfconst_1(none)pushes float 1 onto the stackfconst_2(none)pushes float 2 onto the stack下面列出的操作码处理的int型和float型都是32位的值。Java栈单元（slot）是32位宽的，因此，每次一个int数和float数入栈，它都占用一个单元。下表列出的操作码处理long型和double型。long型和double型的数值占用64位。每次一个long数或double数被压入栈中，它都占用2个栈单元。下面的表格，列出了隐含处理long型和double型的操作码
操作码 | 操作数 | 描述 | lconst_0(none)pushes long 0 onto the stacklconst_1(none)pushes long 1 onto the stackdconst_0(none)pushes double 0 onto the stackdconst_1(none)pushes double 1 onto the stack另外还有一个隐含入栈常量值的操作码，aconst_null，它把空对象（null object）的引用（reference）压入栈中。对象引用的格式取决于JVM实现。对象引用指向垃圾收集堆（garbage-collected heap）中的对象。空对象引用，意味着一个变量当前没有指向任何合法对象。aconst_null操作码用在给引用变量赋null值的时候。
操作码 | 操作数 | 描述 | aconst_null(none)pushes a null object reference onto the stack有2个操作码需要紧跟一个操作数来指明入栈的常量值。下表列出的操作码，用来把合法的byte型和short型的常量值压入栈中。byte型或short型的值在入栈之前，先被扩展成int型的值，因为栈单元是32位宽的。对byte型和short型的操作，实际上是基于它们扩展后的int型值的。
操作码 | 操作数 | 描述 | bipushbyte1expands byte1 (a byte type) to an int and pushes it onto the stacksipushbyte1, byte2expands byte1, byte2 (a short type) to an int and pushes it onto the stack有3个操作码把常量池中的常量值压入栈中。所有和类关联的常量，如final变量，都被保存在类的常量池中。把常量池中的常量压入栈中的操作码，都有一个操作数，它表示需要入栈的常量在常量池中的索引。JVM会根据索引查找常量，确定它的类型，并把它压入栈中。
在字节码流中，常量池索引（constant pool index）是一个紧跟在操作码后的无符号值。操作码lcd1和lcd2把32位的项压入栈中，如int或float。两者的区别在于lcd1只适用于1-255的常量池索引位，因为它的索引只有1个字节。（常量池0号位未被使用。）lcd2的索引有2个字节，所以它可以适用于常量池的任意位置。lcd2w也有一个2字节的索引，它被用来指示任意含有64位的long或double型数据的常量池位置。下表列出了把常量池中的常量压入栈中的操作码：
操作码 | 操作数 | 描述 | ldc1indexbyte1pushes 32-bit constant_pool entry specified by indexbyte1 onto the stackldc2indexbyte1, indexbyte2pushes 32-bit constant_pool entry specified by indexbyte1, indexbyte2 onto the stackldc2windexbyte1, indexbyte2pushes 64-bit constant_pool entry specified by indexbyte1,indexbyte2 onto the stack把局部变量（local variables）压入栈中
局部变量保存在栈帧的一个特殊区域中。栈帧是当前执行方法正在使用的栈区。每个栈帧包含3个部分：本地变量区，执行环境和操作数栈区。把本地变量入栈实际上包含了把数值从栈帧的本地变量区移动到操作数栈区。操作数栈区总是在栈的顶部，所以，把一个值压到当前栈帧的操作数栈区顶部，跟压到整个JVM栈的顶部是一个意思。
Java栈是一个先进后出（LIFO）的32位宽的栈。所有的本地变量至少占用32位，因为栈中的每个单元都是32位宽的。像long和double类型的64位的本地变量会占用2个栈单元。byte和short型的本地变量会当做int型来存储，但只拥有较小类型的合法值。例如，表示byte型的int型本地变量取值范围总是-128到127。
每个本地变量都有一个唯一索引。方法栈帧的本地变量区，可以当成是一个拥有32位宽的元素的数组，每个元素都可以用数组索引来寻址。long和double型的占用2个单元的本地变量，且用低位元素的索引寻址。例如，对一个占用2单元和3单元的double数值，会用索引2来引用。
有一些操作码可以把int和float型本地变量压入操作数栈。部分操作码，定义成隐含常用本地变量地址的引用。例如，iload_0加载处在位置0的int型本地变量。其他本地变量，通过操作码后跟一个字节的本地变量索引的方式压入栈中。iload指令就是这种操作码类型的一个例子。iload后的一个字节被解释成指向本地变量的8位无符号索引。
类似iload所用的8位无符号本地变量索引，限制了一个方法最多只能有256个本地变量。有一个单独的wide指令可以把8位索引扩展为16位索引，则使得本地变量数的上限提高到64k个。操作码wide只有1个操作数。wide和它的操作数，出现在像iload之类的有一个8位无符号本地变量索引的指令之前。JVM会把wide的操作数和iload的操作数合并为一个16位的无符号本地变量索引。
下表列出了把int和float型本地变量压入栈中的操作码：
操作码 | 操作数 | 描述 | iloadvindexpushes int from local variable position vindexiload_0(none)pushes int from local variable position zeroiload_1(none)pushes int from local variable position oneiload_2(none)pushes int from local variable position twoiload_3(none)pushes int from local variable position threefloadvindexpushes float from local variable position vindexfload_0(none)pushes float from local variable position zerofload_1(none)pushes float from local variable position onefload_2(none)pushes float from local variable position twofload_3(none)pushes float from local variable position three接下来的这张表，列出了把long和double型本地变量压入栈中的指令。这些指令把64位的数从栈帧的本地变量区移动到操作数区。
操作码 | 操作数 | 描述 | lloadvindexpushes long from local variable positions vindex and (vindex + 1)lload_0(none)pushes long from local variable positions zero and onelload_1(none)pushes long from local variable positions one and twolload_2(none)pushes long from local variable positions two and threelload_3(none)pushes long from local variable positions three and fourdloadvindexpushes double from local variable positions vindex and (vindex + 1)dload_0(none)pushes double from local variable positions zero and onedload_1(none)pushes double from local variable positions one and twodload_2(none)pushes double from local variable positions two and threedload_3(none)pushes double from local variable positions three and four最后一组操作码，把32位的对象引用从栈帧的本地变量区移动到操作数区。如下表：
操作码 | 操作数 | 描述 | aloadvindexpushes object reference from local variable position vindexaload_0(none)pushes object reference from local variable position zeroaload_1(none)pushes object reference from local variable position oneaload_2(none)pushes object reference from local variable position twoaload_3(none)pushes object reference from local variable position three弹出到本地变量
每一个将局部变量压入栈中的操作码，都有一个对应的负责弹出栈顶元素到本地变量中的操作码。这些操作码的名字可以通过替换入栈操作码名中的“load”为“store”得到。下表列出了将int和float型数值弹出操作数栈到本地变量中的操作码。这些操作码将一个32位的值从栈顶移动到本地变量中。
操作码 | 操作数 | 描述 | istorevindexpops int to local variable position vindexistore_0(none)pops int to local variable position zeroistore_1(none)pops int to local variable position oneistore_2(none)pops int to local variable position twoistore_3(none)pops int to local variable position threefstorevindexpops float to local variable position vindexfstore_0(none)pops float to local variable position zerofstore_1(none)pops float to local variable position onefstore_2(none)pops float to local variable position twofstore_3(none)pops float to local variable position three下一张表中，展示了负责将long和double类型数值出栈并存到局部变量的字节码指令，这些指令将64位的值从操作数栈顶移动到本地变量中。
操作码 | 操作数 | 描述 | lstorevindexpops long to local variable positions vindex and (vindex + 1)lstore_0(none)pops long to local variable positions zero and onelstore_1(none)pops long to local variable positions one and twolstore_2(none)pops long to local variable positions two and threelstore_3(none)pops long to local variable positions three and fourdstorevindexpops double to local variable positions vindex and (vindex + 1)dstore_0(none)pops double to local variable positions zero and onedstore_1(none)pops double to local variable positions one and twodstore_2(none)pops double to local variable positions two and threedstore_3(none)pops double to local variable positions three and four最后一组操作码，负责将32位的对象引用从操作数栈顶移动到本地变量中。
操作码 | 操作数 | 描述 | astorevindexpops object reference to local variable position vindexastore_0(none)pops object reference to local variable position zeroastore_1(none)pops object reference to local variable position oneastore_2(none)pops object reference to local variable position twoastore_3(none)pops object reference to local variable position three类型转换
JVM中有一些操作码用来将一种基本类型的数值转换成另外一种。字节码流中的转换操作码后面不跟操作数，被转换的值取自栈顶。JVM弹出栈顶的值，转换后再将结果压入栈中。下表列出了在int，long，float和double间转换的操作码。这四种类型组合的每一个可能的转换，都有一个对应的操作码。
操作码 | 操作数 | 描述 | i2l(none)converts int to longi2f(none)converts int to floati2d(none)converts int to doublel2i(none)converts long to intl2f(none)converts long to floatl2d(none)converts long to doublef2i(none)converts float to intf2l(none)converts float to longf2d(none)converts float to doubled2i(none)converts double to intd2l(none)converts double to longd2f(none)converts double to float下表列出了将int型转换为更小类型的操作码。不存在直接将long，float，double型转换为比int型小的类型的操作码。因此，像float到byte这样的转换，需要两步。第一步，f2i将float转换为int，第二步，int2byte操作码将int转换为byte。
操作码 | 操作数 | 描述 | int2byte(none)converts int to byteint2char(none)converts int to charint2short(none)converts int to short虽然存在将int转换为更小类型（byte，short，char）的操作码，但是不存在反向转换的操作码。这是因为byte，short和char型的数值在入栈之前会转换成int型。byte，short和char型数值的算术运算，首先要将这些类型的值转为int，然后执行算术运算，最后得到int型结果。也就是说，如果两个byte型的数相加，会得到一个int型的结果，如果你想要byte型的结果，你必须显式地将int类型的结果转换为byte类型的值。例如，下面的代码编译出错：
class BadArithmetic {
byte addOneAndOne() {
byte a = 1;
byte b = 1;
byte c = a + b;
return c;
}
}
javac会对上面的代码给出如下错误：
BadArithmetic.java(7): Incompatible type for declaration.
Explicit cast needed to convert int to byte.
byte c = a + b;
^
Java程序员必须显式的把a + b的结果转换为byte，这样才能通过编译。
class GoodArithmetic {
byte addOneAndOne() {
byte a = 1;
byte b = 1;
byte c = (byte) (a + b);
return c;
}
}
这样，javac会很高兴的生成GoodArithmetic.class文件，它包含如下的addOneAndOne()方法的字节码序列：
iconst_1 // Push int constant 1.
istore_1 // Pop into local variable 1, which is a: byte a = 1;
iconst_1 // Push int constant 1 again.
istore_2 // Pop into local variable 2, which is b: byte b = 1;
iload_1 // Push a (a is already stored as an int in local variable 1).
iload_2 // Push b (b is already stored as an int in local variable 2).
iadd // Perform addition. Top of stack is now (a + b), an int.
int2byte // Convert int result to byte (result still occupies 32 bits).
istore_3 // Pop into local variable 3, which is byte c: byte c = (byte) (a + b);
iload_3 // Push the value of c so it can be returned.
ireturn // Proudly return the result of the addition: return c;
本文译自：Bytecode basics
转载自：http://iofree.cc/%E5%AD%97%E8%8A%82%E7%A0%81%E5%9F%BA%E7%A1%80%EF%BC%9Ajvm%E5%AD%97%E8%8A%82%E7%A0%81%E5%88%9D%E6%8E%A2/
| Description |
aaload32 arrayref, index → valueload onto the stack a reference from an arrayaastore53 arrayref, index, value →store into a reference in an arrayaconst_null01 → nullpush a null reference onto the stackaload191: index→ objectrefload a reference onto the stack from a local variable #indexaload_02a → objectrefload a reference onto the stack from local variable 0aload_12b → objectrefload a reference onto the stack from local variable 1aload_22c → objectrefload a reference onto the stack from local variable 2aload_32d → objectrefload a reference onto the stack from local variable 3anewarraybd2: indexbyte1, indexbyte2count → arrayrefcreate a new array of references of length count and component type identified by the class referenceindex (indexbyte1 << 8 + indexbyte2) in the constant poolareturnb0 objectref → [empty]return a reference from a methodarraylengthbe arrayref → lengthget the length of an arrayastore3a1: indexobjectref →store a reference into a local variable #indexastore_04b objectref →store a reference into local variable 0astore_14c objectref →store a reference into local variable 1astore_24d objectref →store a reference into local variable 2astore_34e objectref →store a reference into local variable 3athrowbf objectref → [empty], objectrefthrows an error or exception (notice that the rest of the stack is cleared, leaving only a reference to the Throwable)baload33 arrayref, index → valueload a byte or Boolean value from an arraybastore54 arrayref, index, value →store a byte or Boolean value into an arraybipush101: byte→ valuepush a byte onto the stack as an integer valuebreakpointca reserved for breakpoints in Java debuggers; should not appear in any class filecaload34 arrayref, index → valueload a char from an arraycastore55 arrayref, index, value →store a char into an arraycheckcastc02: indexbyte1, indexbyte2objectref → objectrefchecks whether an objectref is of a certain type, the class reference of which is in the constant pool at index (indexbyte1 << 8 + indexbyte2)d2f90 value → resultconvert a double to a floatd2i8e value → resultconvert a double to an intd2l8f value → resultconvert a double to a longdadd63 value1, value2 → resultadd two doublesdaload31 arrayref, index → valueload a double from an arraydastore52 arrayref, index, value →store a double into an arraydcmpg98 value1, value2 → resultcompare two doublesdcmpl97 value1, value2 → resultcompare two doublesdconst_00e → 0.0push the constant 0.0 onto the stackdconst_10f → 1.0push the constant 1.0 onto the stackddiv6f value1, value2 → resultdivide two doublesdload181: index→ valueload a double value from a local variable #indexdload_026 → valueload a double from local variable 0dload_127 → valueload a double from local variable 1dload_228 → valueload a double from local variable 2dload_329 → valueload a double from local variable 3dmul6b value1, value2 → resultmultiply two doublesdneg77 value → resultnegate a doubledrem73 value1, value2 → resultget the remainder from a division between two doublesdreturnaf value → [empty]return a double from a methoddstore391: indexvalue →store a double value into a local variable #indexdstore_047 value →store a double into local variable 0dstore_148 value →store a double into local variable 1dstore_249 value →store a double into local variable 2dstore_34a value →store a double into local variable 3dsub67 value1, value2 → resultsubtract a double from anotherdup59 value → value, valueduplicate the value on top of the stackdup_x15a value2, value1 → value1, value2, value1insert a copy of the top value into the stack two values from the top. value1 and value2 must not be of the type double or long.dup_x25b value3, value2, value1 → value1, value3, value2, value1insert a copy of the top value into the stack two (if value2 is double or long it takes up the entry of value3, too) or three values (if value2 is neither double nor long) from the topdup25c {value2, value1} → {value2, value1}, {value2, value1}duplicate top two stack words (two values, if value1 is not double nor long; a single value, if value1 is double or long)dup2_x15d value3, {value2, value1} → {value2, value1}, value3, {value2, value1}duplicate two words and insert beneath third word (see explanation above)dup2_x25e {value4, value3}, {value2, value1} → {value2, value1}, {value4, value3}, {value2, value1}duplicate two words and insert beneath fourth wordf2d8d value → resultconvert a float to a doublef2i8b value → resultconvert a float to an intf2l8c value → resultconvert a float to a longfadd62 value1, value2 → resultadd two floatsfaload30 arrayref, index → valueload a float from an arrayfastore51 arrayref, index, value →store a float in an arrayfcmpg96 value1, value2 → resultcompare two floatsfcmpl95 value1, value2 → resultcompare two floatsfconst_00b → 0.0fpush 0.0f on the stackfconst_10c → 1.0fpush 1.0f on the stackfconst_20d → 2.0fpush 2.0f on the stackfdiv6e value1, value2 → resultdivide two floatsfload171: index→ valueload a float value from a local variable #indexfload_022 → valueload a float value from local variable 0fload_123 → valueload a float value from local variable 1fload_224 → valueload a float value from local variable 2fload_325 → valueload a float value from local variable 3fmul6a value1, value2 → resultmultiply two floatsfneg76 value → resultnegate a floatfrem72 value1, value2 → resultget the remainder from a division between two floatsfreturnae value → [empty]return a floatfstore381: indexvalue →store a float value into a local variable #indexfstore_043 value →store a float value into local variable 0fstore_144 value →store a float value into local variable 1fstore_245 value →store a float value into local variable 2fstore_346 value →store a float value into local variable 3fsub66 value1, value2 → resultsubtract two floatsgetfieldb42: index1, index2objectref → valueget a field value of an object objectref, where the field is identified by field reference in the constant pool index (index1 << 8 + index2)getstaticb22: index1, index2→ valueget a static field value of a class, where the field is identified by field reference in the constant pool index (index1 << 8 + index2)gotoa72: branchbyte1, branchbyte2[no change]goes to another instruction at branchoffset (signed short constructed from unsigned bytes branchbyte1 << 8 + branchbyte2)goto_wc84: branchbyte1, branchbyte2, branchbyte3, branchbyte4[no change]goes to another instruction at branchoffset (signed int constructed from unsigned bytes branchbyte1 << 24 + branchbyte2 << 16 + branchbyte3 << 8 + branchbyte4)i2b91 value → resultconvert an int into a bytei2c92 value → resultconvert an int into a characteri2d87 value → resultconvert an int into a doublei2f86 value → resultconvert an int into a floati2l85 value → resultconvert an int into a longi2s93 value → resultconvert an int into a shortiadd60 value1, value2 → resultadd two intsiaload2e arrayref, index → valueload an int from an arrayiand7e value1, value2 → resultperform a bitwise and on two integersiastore4f arrayref, index, value →store an int into an arrayiconst_m102 → -1load the int value -1 onto the stackiconst_003 → 0load the int value 0 onto the stackiconst_104 → 1load the int value 1 onto the stackiconst_205 → 2load the int value 2 onto the stackiconst_306 → 3load the int value 3 onto the stackiconst_407 → 4load the int value 4 onto the stackiconst_508 → 5load the int value 5 onto the stackidiv6c value1, value2 → resultdivide two integersif_acmpeqa52: branchbyte1, branchbyte2value1, value2 →if references are equal, branch to instruction at branchoffset (signed short constructed from unsigned bytes branchbyte1 << 8 + branchbyte2)if_acmpnea62: branchbyte1, branchbyte2value1, value2 →if references are not equal, branch to instruction at branchoffset (signed short constructed from unsigned bytes branchbyte1 << 8 + branchbyte2)if_icmpeq9f2: branchbyte1, branchbyte2value1, value2 →if ints are equal, branch to instruction at branchoffset (signed short constructed from unsigned bytesbranchbyte1 << 8 + branchbyte2)if_icmpgea22: branchbyte1, branchbyte2value1, value2 →if value1 is greater than or equal to value2, branch to instruction at branchoffset (signed short constructed from unsigned bytes branchbyte1 << 8 + branchbyte2)if_icmpgta32: branchbyte1, branchbyte2value1, value2 →if value1 is greater than value2, branch to instruction at branchoffset (signed short constructed from unsigned bytes branchbyte1 << 8 + branchbyte2)if_icmplea42: branchbyte1, branchbyte2value1, value2 →if value1 is less than or equal to value2, branch to instruction at branchoffset (signed short constructed from unsigned bytes branchbyte1 << 8 + branchbyte2)if_icmplta12: branchbyte1, branchbyte2value1, value2 →if value1 is less than value2, branch to instruction at branchoffset (signed short constructed from unsigned bytes branchbyte1 << 8 + branchbyte2)if_icmpnea02: branchbyte1, branchbyte2value1, value2 →if ints are not equal, branch to instruction at branchoffset (signed short constructed from unsigned bytes branchbyte1 << 8 + branchbyte2)ifeq992: branchbyte1, branchbyte2value →if value is 0, branch to instruction at branchoffset (signed short constructed from unsigned bytesbranchbyte1 << 8 + branchbyte2)ifge9c2: branchbyte1, branchbyte2value →if value is greater than or equal to 0, branch to instruction at branchoffset (signed short constructed from unsigned bytes branchbyte1 << 8 + branchbyte2)ifgt9d2: branchbyte1, branchbyte2value →if value is greater than 0, branch to instruction at branchoffset (signed short constructed from unsigned bytes branchbyte1 << 8 + branchbyte2)ifle9e2: branchbyte1, branchbyte2value →if value is less than or equal to 0, branch to instruction at branchoffset (signed short constructed from unsigned bytes branchbyte1 << 8 + branchbyte2)iflt9b2: branchbyte1, branchbyte2value →if value is less than 0, branch to instruction at branchoffset (signed short constructed from unsigned bytes branchbyte1 << 8 + branchbyte2)ifne9a2: branchbyte1, branchbyte2value →if value is not 0, branch to instruction at branchoffset (signed short constructed from unsigned bytesbranchbyte1 << 8 + branchbyte2)ifnonnullc72: branchbyte1, branchbyte2value →if value is not null, branch to instruction at branchoffset (signed short constructed from unsigned bytes branchbyte1 << 8 + branchbyte2)ifnullc62: branchbyte1, branchbyte2value →if value is null, branch to instruction at branchoffset (signed short constructed from unsigned bytesbranchbyte1 << 8 + branchbyte2)iinc842: index, const[No change]increment local variable #index by signed byte constiload151: index→ valueload an int value from a local variable #indexiload_01a → valueload an int value from local variable 0iload_11b → valueload an int value from local variable 1iload_21c → valueload an int value from local variable 2iload_31d → valueload an int value from local variable 3impdep1fe reserved for implementation-dependent operations within debuggers; should not appear in any class fileimpdep2ff reserved for implementation-dependent operations within debuggers; should not appear in any class fileimul68 value1, value2 → resultmultiply two integersineg74 value → resultnegate intinstanceofc12: indexbyte1, indexbyte2objectref → resultdetermines if an object objectref is of a given type, identified by class reference index in constant pool (indexbyte1 << 8 + indexbyte2)invokedynamicba4: indexbyte1, indexbyte2, 0, 0[arg1, [arg2 ...]] →invokes a dynamic method identified by method reference index in constant pool (indexbyte1 << 8 + indexbyte2)invokeinterfaceb94: indexbyte1, indexbyte2, count, 0objectref, [arg1, arg2, ...] →invokes an interface method on object objectref, where the interface method is identified by method reference index in constant pool (indexbyte1 << 8 + indexbyte2)invokespecialb72: indexbyte1, indexbyte2objectref, [arg1, arg2, ...] →invoke instance method on object objectref, where the method is identified by method reference indexin constant pool (indexbyte1 << 8 + indexbyte2)invokestaticb82: indexbyte1, indexbyte2[arg1, arg2, ...] →invoke a static method, where the method is identified by method reference index in constant pool (indexbyte1 << 8 + indexbyte2)invokevirtualb62: indexbyte1, indexbyte2objectref, [arg1, arg2, ...] →invoke virtual method on object objectref, where the method is identified by method reference index in constant pool (indexbyte1 << 8 + indexbyte2)ior80 value1, value2 → resultbitwise int orirem70 value1, value2 → resultlogical int remainderireturnac value → [empty]return an integer from a methodishl78 value1, value2 → resultint shift leftishr7a value1, value2 → resultint arithmetic shift rightistore361: indexvalue →store int value into variable #indexistore_03b value →store int value into variable 0istore_13c value →store int value into variable 1istore_23d value →store int value into variable 2istore_33e value →store int value into variable 3isub64 value1, value2 → resultint subtractiushr7c value1, value2 → resultint logical shift rightixor82 value1, value2 → resultint xorjsra82: branchbyte1, branchbyte2→ addressjump to subroutine at branchoffset (signed short constructed from unsigned bytes branchbyte1 << 8 + branchbyte2) and place the return address on the stackjsr_wc94: branchbyte1, branchbyte2, branchbyte3, branchbyte4→ addressjump to subroutine at branchoffset (signed int constructed from unsigned bytes branchbyte1 << 24 + branchbyte2 << 16 + branchbyte3 << 8 + branchbyte4) and place the return address on the stackl2d8a value → resultconvert a long to a doublel2f89 value → resultconvert a long to a floatl2i88 value → resultconvert a long to a intladd61 value1, value2 → resultadd two longslaload2f arrayref, index → valueload a long from an arrayland7f value1, value2 → resultbitwise and of two longslastore50 arrayref, index, value →store a long to an arraylcmp94 value1, value2 → resultcompare two longs valueslconst_009 → 0Lpush the long 0 onto the stacklconst_10a → 1Lpush the long 1 onto the stackldc121: index→ valuepush a constant #index from a constant pool (String, int or float) onto the stackldc_w132: indexbyte1, indexbyte2→ valuepush a constant #index from a constant pool (String, int or float) onto the stack (wide index is constructed as indexbyte1 << 8 + indexbyte2)ldc2_w142: indexbyte1, indexbyte2→ valuepush a constant #index from a constant pool (double or long) onto the stack (wide index is constructed as indexbyte1 << 8 + indexbyte2)ldiv6d value1, value2 → resultdivide two longslload161: index→ valueload a long value from a local variable #indexlload_01e → valueload a long value from a local variable 0lload_11f → valueload a long value from a local variable 1lload_220 → valueload a long value from a local variable 2lload_321 → valueload a long value from a local variable 3lmul69 value1, value2 → resultmultiply two longslneg75 value → resultnegate a longlookupswitchab4+: <0-3 bytes padding>, defaultbyte1, defaultbyte2, defaultbyte3, defaultbyte4, npairs1, npairs2, npairs3, npairs4, match-offset pairs...key →a target address is looked up from a table using a key and execution continues from the instruction at that addresslor81 value1, value2 → resultbitwise or of two longslrem71 value1, value2 → resultremainder of division of two longslreturnad value → [empty]return a long valuelshl79 value1, value2 → resultbitwise shift left of a long value1 by value2 positionslshr7b value1, value2 → resultbitwise shift right of a long value1 by value2 positionslstore371: indexvalue →store a long value in a local variable #indexlstore_03f value →store a long value in a local variable 0lstore_140 value →store a long value in a local variable 1lstore_241 value →store a long value in a local variable 2lstore_342 value →store a long value in a local variable 3lsub65 value1, value2 → resultsubtract two longslushr7d value1, value2 → resultbitwise shift right of a long value1 by value2 positions, unsignedlxor83 value1, value2 → resultbitwise exclusive or of two longsmonitorenterc2 objectref →enter monitor for object ("grab the lock" - start of synchronized() section)monitorexitc3 objectref →exit monitor for object ("release the lock" - end of synchronized() section)multianewarrayc53: indexbyte1, indexbyte2, dimensionscount1, [count2,...] → arrayrefcreate a new array of dimensions dimensions with elements of type identified by class reference in constant pool index (indexbyte1 << 8 + indexbyte2); the sizes of each dimension is identified bycount1, [count2, etc.]newbb2: indexbyte1, indexbyte2→ objectrefcreate new object of type identified by class reference in constant pool index (indexbyte1 << 8 + indexbyte2)newarraybc1: atypecount → arrayrefcreate new array with count elements of primitive type identified by atypenop00 [No change]perform no operationpop57 value →discard the top value on the stackpop258 {value2, value1} →discard the top two values on the stack (or one value, if it is a double or long)putfieldb52: indexbyte1, indexbyte2objectref, value →set field to value in an object objectref, where the field is identified by a field reference index in constant pool (indexbyte1 << 8 + indexbyte2)putstaticb32: indexbyte1, indexbyte2value →set static field to value in a class, where the field is identified by a field reference index in constant pool (indexbyte1 << 8 + indexbyte2)reta91: index[No change]continue execution from address taken from a local variable #index (the asymmetry with jsr is intentional)returnb1 → [empty]return void from methodsaload35 arrayref, index → valueload short from arraysastore56 arrayref, index, value →store short to arraysipush112: byte1, byte2→ valuepush a short onto the stackswap5f value2, value1 → value1, value2swaps two top words on the stack (note that value1 and value2 must not be double or long)tableswitchaa4+: [0-3 bytes padding], defaultbyte1, defaultbyte2, defaultbyte3, defaultbyte4, lowbyte1, lowbyte2, lowbyte3, lowbyte4, highbyte1, highbyte2, highbyte3, highbyte4, jump offsets...index →continue execution from an address in the table at offset indexwidec43/5: opcode, indexbyte1, indexbyte2
or
iinc, indexbyte1, indexbyte2, countbyte1, countbyte2[same as for corresponding instructions]execute opcode, where opcode is either iload, fload, aload, lload, dload, istore, fstore, astore, lstore, dstore, or ret, but assume the index is 16 bit; or execute iinc, where the index is 16 bits and the constant to increment by is a signed 16 bit short(no name)cb-fd these values are currently unassigned for opcodes and are reserved for future use参考：http://blog.csdn.net/ygc87/article/details/12953421
0 0
