Re: [HACKERS] initdb segfaults - latest cvs

2003-11-21 Thread Tom Lane
strk <[EMAIL PROTECTED]> writes: > It seems that the build system is missing something > (make distclean made it work) If you aren't using configure --enable-depend, you should count on doing at least "make clean", preferably "make distclean" anytime you do a CVS update. The default behavior is n

Re: [HACKERS] initdb segfaults - latest cvs

2003-11-19 Thread strk
andrew wrote: > Ok, seriously weird. This is apparently from the pg_type relation and > looks just fine on my installation. > > Have you tried "make distclean; cvs update; configure; make check" ? Tried now: 4 of 93 tests failed. .. initdb does not fault though ;) It seems that the build system

Re: [HACKERS] initdb segfaults - latest cvs

2003-11-19 Thread Andrew Dunstan
strk wrote: Running initdb: creating template1 database in /pgroot-cvs/data/base/1 ... child process was terminated by signal 11 It is working fine for me (RH9). Can you provide more details? Platform? How you are calling initdb? cheers andrew ---(end of broadcast)

[HACKERS] initdb segfaults - latest cvs

2003-11-19 Thread strk
Running initdb: creating template1 database in /pgroot-cvs/data/base/1 ... child process was terminated by signal 11 ---(end of broadcast)--- TIP 9: the planner will ignore your desire to choose an index scan if your joining column's datatype