On Sat, Oct 3, 2026 at 12:49 PM Daniel Gustafsson <[email protected]> wrote: > Zsolt also reported two small issues off-list which are solved in 0004 and > 0005. On a failed reload of non-SNI configuration we could be left with > dangling pointers to the previous config GUC values.
Hm. (Are we going to need to build an atomicity system for GUC itself? It sure seems like we should have the toolset for it...) --Jacob
