1. Note that double and long values occupy two registers (e.g. the value addressed by vy is located in vy and vy+1 registers)
2. The offset can be positive or negative and it is calculated from the offset of the starting byte of the instruction. The offset is always interpreted in words (2 bytes per 1 offset value increment/decrement). Negative offset is stored in two's complement format. The current position is the offset of the starting byte of the instruction.
3. Compare operations return positive value if the first operand is greater than the second operand, 0 if they are equal and negative value if the first operand is smaller than the second operand.
4. Not seen in the wild, interpolated from Dalvik bytecode list.
5. The invocation parameter list encoding is somewhat weird. Starting if parameter number > 4 and parameter number % 4 == 1, the 5th (9th, etc.) parameter is encoded on the 4 lowest bit of the byte immediately following the instruction. Curiously, this encoding is not used in case of 1 parameter, in this case an entire 16 bit word is added after the method index of which only 4 bit is used to encode the single parameter while the lowest 4 bit of the byte following the instruction byte is left unused.
6. This is an unsafe instruction and occurs only in ODEX files.
7. In Smali, there are different types of instructions used to define different types of constants, including const/16 and const-wide/16.

Source

Dalvik opcodes

Practice

.method public static func(I)I
    .parameter n, "n"

    .prologue
    const/4 v0, 0x0         
    if-ne v0, p0, :cond_0   
    const/4 v0, 0x1         
    return v0

    :cond_0
    sub-int v1, p0, 0x1     
    invoke-static {v1}, LMyClass;->func(I)I   
    mul-int/2addr v0, p0    
    return v0
.end method

Convert this code to java.

• Solution