On Tue, 14 Oct 2025 at 21:05, Richard Guo <guofenglinux@gmail.com> wrote:
> FWIW, I complained about the stray check in has_useful_pathkeys() in
> [1] last week, but you were quicker than me in making the code change
> to remove it.
I missed that. I'm confident that line does nothing but waste cycles.
A quick look at how those pathkeys are set in standard_qp_callback()
should remove any uncertainty.
David
> [1] https://postgr.es/m/CAMbWs4_zW5QU=Zk32s17p8qWY+ga-3ZUTons+y+Wopguiopm4A@mail.gmail.com