RE: [PATCH] x86: fix and improve cmpxchg_double{,_local}()

 >>> On 04.01.12 at 11:36, "David Laight" [off-list ref] wrote:

quoted

BTW atomic64_set() asm() contraints are wrong :

static inline void atomic64_set(atomic64_t *v, long long i)
{
        unsigned high = (unsigned)(i >> 32);
        unsigned low = (unsigned)i;
        asm volatile(ATOMIC64_ALTERNATIVE(set)
                     : "+b" (low), "+c" (high)
                     : "S" (v)
                     : "eax", "edx", "memory"
                     );
}


ebx/ecx registers are not modified by cmpxchg8b (or the 
atomic64_set_386 emulation). Only eax/edx can be modified.

Same would be true for atomic64_xchg() and the use of "+c" (v)
in subsequent functions (whether unnecessarily strict or too lax
varies).

Isn't it also possible to constrain the "memory"
constraint to only apply to '*v' not all of memory?
I can't remember the syntax off hand though...

Absolutely - "=m" (v->counter) would be the right way to specify
this.

Jan