postgres1(at)lelik(dot)org writes:
> Program terminated with signal SIGSEGV, Segmentation fault.
> #0 0x0000000000676a68 in make_subqueryscan (qptlist=0x33f0620, qpqual=0x0,
> scanrelid=1, subplan=0x0) at createplan.c:3259
> 3259 plan->total_cost += cpu_tuple_cost * subplan->plan_rows;
Interesting, but nobody's going to be able to find the cause with just
this stack trace to go on. Can you create a reproducible test case?
Also, can you verify that it happens on a current release not just 9.2.14?
I don't immediately see anything clearly related in the commit logs, but
it's certainly possible that we already fixed it.
regards, tom lane