看完微软大神写的求平均值代码，我意识到自己还是too young了

unsigned average(unsigned a, unsigned b)
{
return (a + b) / 2;
}

unsigned average(unsigned low, unsigned high)
{
return low + (high - low) / 2;
}

unsigned average(unsigned a, unsigned b)
{
return (a / 2) + (b / 2) + (a & b & 1);
}

unsigned average(unsigned a, unsigned b)
{
return (a & b) + (a ^ b) / 2;// 变体 (a ^ b) + (a & b) * 2

unsigned average(unsigned a, unsigned b)
{
// Suppose "unsigned" is a 32-bit type and
// "unsigned long long" is a 64-bit type.
return ((unsigned long long)a + b) / 2;
}

// x86-64: Assume ecx = a, edx = b, upper 32 bits unknown
mov eax, ecx ; rax = ecx zero-extended to 64-bit value
mov edx, edx ; rdx = edx zero-extended to 64-bit value
add rax, rdx ; 64-bit addition: rax = rax + rdx
shr rax, 1 ; 64-bit shift: rax = rax >> 1
; result is zero-extended
; Answer in eax

// AArch64 (ARM 64-bit): Assume w0 = a, w1 = b, upper 32 bits unknown
uxtw x0, w0 ; x0 = w0 zero-extended to 64-bit value
uxtw x1, w1 ; x1 = w1 zero-extended to 64-bit value
add x0, x1 ; 64-bit addition: x0 = x0 + x1
ubfx x0, x0, 1, 32 ; Extract bits 1 through 32 from result
; (shift + zero-extend in one instruction)
; Answer in x0

// Alpha AXP: Assume a0 = a, a1 = b, both in canonical form
insll a0, #0, a0 ; a0 = a0 zero-extended to 64-bit value
insll a1, #0, a1 ; a1 = a1 zero-extended to 64-bit value
addq a0, a1, v0 ; 64-bit addition: v0 = a0 + a1
srl v0, #1, v0 ; 64-bit shift: v0 = v0 >> 1
addl zero, v0, v0 ; Force canonical form
; Answer in v0

// MIPS64: Assume a0 = a, a1 = b, sign-extended
dext a0, a0, 0, 32 ; Zero-extend a0 to 64-bit value
dext a1, a1, 0, 32 ; Zero-extend a1 to 64-bit value
daddu v0, a0, a1 ; 64-bit addition: v0 = a0 + a1
dsrl v0, v0, #1 ; 64-bit shift: v0 = v0 >> 1
sll v0, #0, v0 ; Sign-extend result
; Answer in v0

// Power64: Assume r3 = a, r4 = b, zero-extended
add r3, r3, r4 ; 64-bit addition: r3 = r3 + r4
rldicl r3, r3, 63, 32 ; Extract bits 63 through 32 from result
; (shift + zero-extend in one instruction)
; result in r3

// x86-32
mov eax, a
add eax, b ; Add, overflow goes into carry bit
rcr eax, 1 ; Rotate right one place through carry

// x86-64
mov rax, a
add rax, b ; Add, overflow goes into carry bit
rcr rax, 1 ; Rotate right one place through carry

// 32-bit ARM (A32)
mov r0, a
adds r0, b ; Add, overflow goes into carry bit
rrx r0 ; Rotate right one place through carry

// SH-3
clrt ; Clear T flag
mov a, r0
addc b, r0 ; r0 = r0 + b + T, overflow goes into T bit
rotcr r0 ; Rotate right one place through carry

unsigned average(unsigned a, unsigned b)
{
#if defined(_MSC_VER)
unsigned sum;
auto carry = _addcarry_u32(0, a, b, &sum);
sum = (sum & ~1) | carry;
return _rotr(sum, 1);
#elif defined(__clang__)
unsigned carry;
sum = (sum & ~1) | carry;
auto sum = __builtin_addc(a, b, 0, &carry);
return __builtin_rotateright32(sum, 1);
#else
#error Unsupported compiler.
#endif
}

// _MSC_VER
mov ecx, a
add ecx, b ; Add, overflow goes into carry bit
setc al ; al = 1 if carry set
and ecx, -2 ; Clear bottom bit
movzx ecx, al ; Zero-extend byte to 32-bit value
or eax, ecx ; Combine
ror ear, 1 ; Rotate right one position
; Result in eax

// __clang__
mov ecx, a
add ecx, b ; Add, overflow goes into carry bit
setc al ; al = 1 if carry set
shld eax, ecx, 31 ; Shift left 64-bit value

// __clang__ with ARM-Thumb2
movs r2, #0 ; Prepare to receive carry
adds r0, r0, r1 ; Calculate sum with flags
adcs r2, r2 ; r2 holds carry
lsrs r0, r0, #1 ; Shift sum right one position
lsls r1, r2, #31 ; Move carry to bit 31
adds r0, r1, r0 ; Combine

unsigned avg(unsigned a, unsigned b
{
return (a & b) + (a ^ b) / 2;
}

// lw $3,8($fp) # 5
// lw $2,12($fp) # 5
// and $3,$3,$2 # 4
// lw $4,8($fp) # 5
// lw $2,12($fp) # 5
// xor $2,$4,$2 # 4
// srl $2,$2,1 # 4
// addu $2,$3,$2 # 4