Re: New WAL record to detect the checkpoint redo location

Hi,

On 2023-10-06 13:44:55 -0400, Robert Haas wrote:
> On Thu, Oct 5, 2023 at 2:34 PM Andres Freund  wrote:
> > If I add an unlikely around if (rechdr->xl_rmid == RM_XLOG_ID), the
> > performance does improve. But that "only" brings it up to 322.406. Not sure
> > what the rest is.
> 
> I don't really think this is worth worrying about. A sub-one-percent
> regression on a highly artificial test case doesn't seem like a big
> deal.

I agree. I think it's worth measuring and looking at, after all the fix might
be trivial (like the case of the unlikely for the earlier if()). But it
shouldn't block progress on significant features.

I think this "issue" might be measurable in some other, not quite as artifical
cases, like INSERT ... SELECT or such. But even then it's going to be tiny.

Greetings,

Andres Freund