Em sex., 11 de abr. de 2025 às 02:09, Michael Paquier <michael@paquier.xyz> escreveu:
On Thu, Apr 10, 2025 at 03:10:02PM -0300, Ranier Vilela wrote: > While it is arguable that this is a false warning, there is a benefit in > moving the initialization of the string buffer, silencing the warnings that > are presented in this case. > > 1. pg_overexplain.c > 2. ruleutils.c
These code paths are far from being critical and the two ones in ruleutils.c are older, even if it is a practice that had better be discouraged particularly as initStringInfo() can allocate some memory for nothing. So it could bloat the current memory context if these code paths are repeatedly taken.
Yeah, it's a bit annoying to do unnecessary work.
Plus a small gain, by delaying memory allocation until when it is actually needed.
FWIW, I'm with these changes to delay these initializations as you are proposing.
Thanks.
The RMT has a say about such changes post feature-freeze, though, even if the one in pg_overexplain.c is new to v18.
