Re: locked reads for atomics

On Fri, Feb 23, 2024 at 10:25:00AM -0800, Jeff Davis wrote:
> To be clear:
> 
>   x = pg_atomic_[read|write]_membarrier_u64(&v);
> 
> is semantically equivalent to:
> 
>   pg_memory_barrier();
>   x = pg_atomic_[read|write]_u64(&v);
>   pg_memory_barrier();
> 
> ?
> 
> If so, that does seem more convenient.

I think that's about right.  The upthread feedback from Andres [0] provides
some additional context.

[0] https://postgr.es/m/20231110231150.fjm77gup2i7xu6hc%40alap3.anarazel.de

-- 
Nathan Bossart
Amazon Web Services: https://aws.amazon.com