Greetings,
* Kyotaro Horiguchi (horikyota.ntt@gmail.com) wrote:
> At Tue, 06 Sep 2022 17:10:49 -0400, Reid Thompson <reid.thompson@crunchydata.com> wrote in
> > I'm open to guidance on testing for performance degradation. I did
> > note some basic pgbench comparison numbers in the thread regarding
> > limiting backend memory allocations.
>
> Yeah.. That sounds good..
>
> (I have a patch that is stuck at benchmarking on slight possible
> degradation caused by a branch (or indirect call) on a hot path
> similary to this one. The test showed fluctuation that is not clearly
> distinguishable between noise and degradation by running the target
> functions in a busy loop..)
Just to be clear- this path is (hopefully) not *super* hot as we're only
tracking actual allocations (that is- malloc() calls), this isn't
changing anything for palloc() calls that aren't also needing to do a
malloc(), and we already try to reduce the amount of malloc() calls
we're doing by allocating more and more each time we run out in a given
context.
While I'm generally supportive of doing some benchmarking around this, I
don't think the bar is as high as it would be if we were actually
changing the cost of routine palloc() or such calls.
Thanks,
Stephen