最近看自旋锁的实现，自选锁的循环查找锁的主要实现类似如下，该实现使用到了内嵌的汇编（摘自sanos内核，源代码有2处实现，一处使用intel汇编，是没有问题的，另一处使用内嵌汇编语法，源代码中为cmpxchgl %2, %0，是错误的，应该是cmpxchgl %0, %2）
内嵌汇编有个固定格式，如下：
asm ( assembler
template /* 汇编语句 */
: output operands /* 输出 */
: input operands /* 输入 */
: list of clobbered registers
);
cmpxchgl的描述如下：
Compares the value in the AL, AX, EAX, or RAX register with the first operand (destination operand). If the twovalues are equal, the second operand (source operand) is loaded into the destination operand. Otherwise, thedestination operand is loaded into the AL, AX, EAX or RAX register. RAX register is available only in 64-bit mode.
(* Accumulator = AL, AX, EAX, or RAX depending on whether a byte, word, doubleword, or quadword comparison is being performed *)
TEMP ← DEST
IF accumulator = TEMP
THEN
ZF ← 1;
DEST ← SRC;
ELSE
ZF ← 0;
accumulator ← TEMP;
DEST ← TEMP;
FI;
TEMP ← DEST
IF accumulator = TEMP
THEN
ZF ← 1;
DEST ← SRC;
ELSE
ZF ← 0;
accumulator ← TEMP;
DEST ← TEMP;
FI;
- cmpxchgl %0, %2为汇编语句，表示对第3个和第1个入参进行操作，即cmpxchgl *dest，exchange；
- "=m" (*dest), "=a" (old)为输出部分，将m内存的内容存到*dest中，将a寄存器内容存到old；
- "r" (exchange), "m" (*dest), "a" (comperand)); 为输入部分，将exchange放入r寄存器，将*dest放入m，将comperand放入a寄存器；
使用C语言翻译如下：
int atomic_compare_and_exchange(int *dest, int exchange, int comperand)
{
int old = comperand;
if ( comperand== *dest)
{
*dest = exchange;
}
else
{
old = *dest;
}
return old;
}