Re: Spinlocks, yet again: analysis and proposed patches

Stephen Frost <sfrost(at)snowman(dot)net> writes:
> I suspect distributors would go for the multi-cpu setup (especially if
> a uniprocessor build is *broken* for multiprocessor) and then in a
> lot of cases you end up not actually getting any benefit. I'm afraid
> you'd also end up having to tell alot of people who complain to
> recompile, who will then complain back to their distributors, etc.

Yeah. Being in charge of Red Hat's packaging of PG, I feel that pain as
keenly as anybody ... and I *know* RH will not be interested in shipping
two different packages. If we go this way, the RH distributions will
use the --optimize-multi switch, because that's where the money is.

The bottom line here is that we will have to make some compromises:
if we want one-size-fits-all code, it will not be optimal for every
single architecture. If we don't do one-size-fits-all, then we will
pay for it in various other ways.

regards, tom lane