Reiner Dassing <dassing(at)wettzell(dot)ifag(dot)de> writes:
> The modified "mainloop.c" I sent you is due to:
> query_start contains a value nearly infinity and then it produces
> the core dump.
Yes, but the broken query_start value is the fault of the bug in
command.c. You can get the patch from our CVS server,
http://www.ca.postgresql.org/cgi/cvsweb.cgi/pgsql/src/bin/psql/command.c.diff?r1=1.49&r2=1.49.2.1&f=c
regards, tom lane