Hi all,
Since the "fix" for varchars in "SELECT INTO" I'm getting a core dump on
initdb when initializing the database.
I've tracked this down to the "REVOKE ALL ON pg_user FROM public" step.
Looking at the (not compiled with -g) core I see.
(gdb seems to be broken on S/Linux as I can't get a backtrace)
GDB 4.16 (sparc-unknown-linux), Copyright 1996 Free Software Foundation, Inc...
Core was generated by `postgres -F -Q -D/usr/local/pgsql/data template1'.
Program terminated with signal 11, Segmentation fault.
Reading symbols from /lib/libdl.so.1.8.3...done.
Reading symbols from /lib/libm.so.5.0.6...done.
Reading symbols from /usr/lib/libreadline.so.2.0...done.
Reading symbols from /lib/libtermcap.so.2.0.8...done.
Reading symbols from /lib/libc.so.5.3.12...done.
Reading symbols from /lib/ld-linux.so.1...done.
#0 0xc7f38 in OrderedElemGetBase ()
(gdb) bt
#0 0xc7f38 in OrderedElemGetBase ()
(gdb)
OrderedElemGetBase() seems to have something to so with extracting stuff :-)
from structures so I wonder if somehow the new...
#define USE_ATTTYPMOD(typeid) ((typeid) == BPCHAROID || (typeid) ==
VARCHAROID)
in include/catalog/pg_type.h is somehow breaking some alignment rules
on SPARC architecture?
I'm currently trying to build on SPARC-Solaris2.6 but it's a major struggle
due to the changes in "dynloader" and "tas". (configure no longer works)
Keith.