Re: segfault with incremental sort

luis.roberto@siscobra.com.br writes:
> I also managed to find a coredump from systemd, if you want,I can send a link... it's 290MB, though.

Not much use in that, nobody else could do anything without the
matching debug symbols either.

> I can see messages like this, using journalctl -xe:

> Stack trace of thread 548078:
> #0 0x0000000000665ee4 ExecSubPlan (/usr/pgsql-13/bin/postgres)

Hm ... suggestive, but not really enough to debug it.

Can you build a self-contained test case?

            regards, tom lane