"Blake, Geoff" <blakgeof(at)amazon(dot)com> writes:
> Have a tiny patch to add an implementation of spin_delay() for Arm64 processors to match behavior with x86's PAUSE instruction. See negligible benefit on the pgbench tpcb-like workload so at worst it appears to do no harm but should help some workloads that experience some lock contention that need to spin.
Given the very wide variety of ARM implementations out there,
I'm not sure that we want to take a patch like this on the basis of
exactly zero evidence. It could as easily be a net loss as a win.
What did you test exactly?
regards, tom lane