Re: [PATCH] Add stdatomic.h for clang>=3.1 and gcc>=4.1

[Rich Felker <dalias@libc.org>]

On Sun, Nov 23, 2014 at 02:47:12AM +0100, Joakim Sindholt wrote:
> > > > > > I have changed it to be an atomic_bool in a struct as both GCC and Clang
> > > > > > has it in a struct. Presumably to separate it from the generic _Atomic
> > > > > > stuff.
> > > > > 
> > > > > Again, since we want to have ABI compatibility, it is not your choice
> > > > > to make. You'd simply have to stick to the choice that gcc made. So
> > > > > you have to copy the declaration of the struct, including all the
> > > > > ifdef fuzz.
> > > > 
> > > > I'd have to look at it again, but IIRC only one case of the #ifdef
> > > > mess was actually possible. The others were for hypothetical archs
> > > > without real atomics which we can't support anyway.
> > > 
> > > We should have it as a struct, if the implementations have it like
> > > that, I think:
> > > 
> > >  - It should have same alignment properties for ABI compatibility.
> > >  - It should lead to the same typename when included in C++.
> > 
> > Yes.
> > 
> > > The ifdef is a single one to switch between _Bool or unsigned char or
> > > so.
> > 
> > Yes, but I think the #ifdef always comes out one way anyway, though I
> > don't remember which one and don't have the file in front of me.
> 
> GCC 4.9:
> 
> typedef _Atomic struct
> {
> #if __GCC_ATOMIC_TEST_AND_SET_TRUEVAL == 1
>   _Bool __val;
> #else
>   unsigned char __val;
> #endif
> } atomic_flag;
> 
> Clang 3.6:
> 
> #ifdef __cplusplus
> typedef _Atomic(bool)               atomic_bool;
> #else
> typedef _Atomic(_Bool)              atomic_bool;
> #endif
> 
> typedef struct atomic_flag { atomic_bool _Value; } atomic_flag;

So yes, only the true case of the #if is needed, and that's what clang
implements. The other is for hypothetical archs which lack the ability
to CAS 0 and 1 into an object of type _Bool -- for example, perhaps
they can do a lock-free boolean test-and-set, but only with values 0
and -1, whereas _Bool needs to store 1 for true. This is just a guess
what the GCC folks might have had in mind, but in any case, it doesn't
matter, since we assume a full CAS (and require a kernel that emulates
one if the hardware does not have one).

Rich