On Sun, 2015-08-02 at 17:11 +0200, Andrey Konovalov wrote:
> Replace ACCESS_ONCE() macro in smp_store_release() and smp_load_acquire()
> with WRITE_ONCE() and READ_ONCE() on x86, arm, arm64, ia64, metag, mips,
> powerpc, s390, sparc and asm-generic since ACCESS_ONCE does not work
> reliably on non-scalar types.
.. and there are no restrictions on the argument to smp_load_acquire(), so it
may be a non-scalar type.
Though from a quick grep it looks like no one is doing that at the moment?
> WRITE_ONCE() and READ_ONCE() were introduced in the commits 230fa253df63
> ("kernel: Provide READ_ONCE and ASSIGN_ONCE") and 43239cbe79fc ("kernel:
> Change ASSIGN_ONCE(val, x) to WRITE_ONCE(x, val)").
>
> Signed-off-by: Andrey Konovalov <andreyknvl@xxxxxxxxxx>
> ---
> Changed in v2:
> - Other archs besides x86.
>
> arch/powerpc/include/asm/barrier.h | 4 ++--
> diff --git a/arch/powerpc/include/asm/barrier.h b/arch/powerpc/include/asm/barrier.h
> index 51ccc72..0eca6ef 100644
> --- a/arch/powerpc/include/asm/barrier.h
> +++ b/arch/powerpc/include/asm/barrier.h
> @@ -76,12 +76,12 @@
> do { \
> compiletime_assert_atomic_type(*p); \
> smp_lwsync(); \
> - ACCESS_ONCE(*p) = (v); \
> + WRITE_ONCE(*p, v); \
> } while (0)
>
> #define smp_load_acquire(p) \
> ({ \
> - typeof(*p) ___p1 = ACCESS_ONCE(*p); \
> + typeof(*p) ___p1 = READ_ONCE(*p); \
> compiletime_assert_atomic_type(*p); \
> smp_lwsync(); \
> ___p1; \
Acked-by: Michael Ellerman <mpe@xxxxxxxxxxxxxx> (powerpc)
cheers
