Обсуждение: Status of 64 bit atomics

<div dir="ltr"><p dir="ltr"><br /> On May 27, 2016 5:01 PM, "John Gorman" <<a href="mailto:johngorman2@gmail.com"
target="_blank">johngorman2@gmail.com</a>>wrote:<br /> ><br /> > Hi All<br /> ><br /> > Someone recently
toldme that the postgresql atomics library was incomplete<br /> > for 64 bit operations such
as pg_atomic_fetch_add_u64()and should not be used.<br /> ><br /> > Can someone definitively confirm whether it
isokay to rely on the 64 bit atomics<br /> > or whether it is better to protect 64 bit operations with a
spinlock?<br/> ><br /> > Thanks!<br /> > John<p dir="ltr">Golang has asm implementations for these even on
32bitplatforms (see <a
href="https://github.com/golang/go/tree/master/src/sync/atomic">https://github.com/golang/go/tree/master/src/sync/atomic</a>).<p
dir="ltr">Couldn'twe borrow them? Or even better, fall back to spin lock on these, but transparently. </div> 

Hi,

On 2016-05-27 11:00:42 -0400, John Gorman wrote:
> Someone recently told me that the postgresql atomics library was incomplete
> for 64 bit operations such as pg_atomic_fetch_add_u64() and should not be
> used.

There's currently no fallback for 32bit platforms without 64bit atomics
support.  I posted a patch adding that fallback, which I plan to commit
soon after the start of the 9.7 development window opens.


> Can someone definitively confirm whether it is okay to rely on the 64
> bit atomics or whether it is better to protect 64 bit operations with
> a spinlock?

For current versions 64bit atomics are working, but you'll get
compilation errors if the platform doesn't have 64bit atomics
support. That's not actually many CPUs these days; most prominent are
probably older arm CPUs.

Andres