[PATCH 1/2] Optimize generic spinlock code and use C11 like atomic macros.

[Stefan Liebler]

On 04/11/2017 02:05 PM, Andreas Schwab wrote:
> On Apr 11 2017, Stefan Liebler <stli@linux.vnet.ibm.com> wrote:
>
>> It happened in the case of __HAVE_64B_ATOMICS.
>
> We shouldn't follow a bad example.
>
> Andreas.
>

Then I'll change the check to:
/* ATOMIC_EXCHANGE_USES_CAS is non-zero if atomic_exchange operations
    are implemented based on a CAS loop; otherwise, this is zero and we 
assume
    that the atomic_exchange operations could provide better performance
    than a CAS loop.  */
#ifndef ATOMIC_EXCHANGE_USES_CAS
# error ATOMIC_EXCHANGE_USES_CAS has to be defined.
#endif

Bye
Stefan