Re: Spinlocks, yet again: analysis and proposed patches

"Michael Paesold" <mpaesold(at)gmx(dot)at> writes:
> It seems to me the slock-no-cmpb is a win in any case. The spin-delay patch
> does not really help much on this machine. That seems to match Stephen
> Frost's results with EM64T, if I read them correctly.

Yeah, it's interesting that you both see slock-no-cmpb as saving some
cycles and the second patch as giving them back. I wonder whether the
integer-modulo-to-slow-the-loop trick is counterproductive on your
machines. You both seem to be using hardware that will recognize
rep;nop and maybe that's all that's needed.

I probably should have broken down the spindelay patch into multiple
components. But it's only a small change --- could you try simplifying
the patched line

if ((--spins % MAX_SPINS_PER_DELAY) == 0)

to

if (--spins == 0)

and see how the patch does that way?

regards, tom lane