[postgis-users] Postgres 8.1beta2 crash with 3D (and beyond) geometries..

[Alex Mayrhofer]

Hi,

I'm trying to use geometries with dimensions > 2 on one of my test machines. 
   Table creation was fine (i was using a 4D POINT geometry), however, when 
PostGIS tries to parse a POINT geometry with dimension > 2 (tried with 
3d/4d, just POINTs), Postgres crashes:

first is fine, second makes postgis crash:
----
geotest=# select geomfromtext('POINT( 0 0)');
                 geomfromtext
--------------------------------------------
  000000000100000000000000000000000000000000
(1 row)

geotest=# select geomfromtext('POINT( 0 0 0)');
WARNING:  OGC WKT expected, EWKT provided - use GeomFromEWKT() for this
CONTEXT:  SQL function "geomfromtext" statement 1
server closed the connection unexpectedly
         This probably means the server terminated abnormally
         before or while processing the request.
The connection to the server was lost. Attempting reset: Failed.
!>
!> \q
----

This can be reproduced at any time.
I'm running the following configuration:

----
geotest=# select postgis_full_Version();
                                          postgis_full_version 

------------------------------------------------------------------------------------------------------
  POSTGIS="1.0.4" GEOS="2.1.3" PROJ="Rel. 4.4.9, 29 Oct 2004" USE_STATS 
DBPROC="0.3.0" RELPROC="0.3.0"
(1 row)

geotest=# select version();
                                               version 

----------------------------------------------------------------------------------------------------
  PostgreSQL 8.1beta2 on sparc-unknown-linux-gnu, compiled by GCC cc (GCC) 
3.3.5 (Debian 1:3.3.5-13)
(1 row)
----

Everything seems to run fine for 2D geometries - this was my first attempt 
at 3d geometries on this box.

The system is an UltraSparc machine, Postgres as well as PostGIS are running 
in 32bit userland. A gdb backtrace of a crashed postmaster shows a SIGBUS, 
and a corrupted stack.

Can i track this down somehow in a "standalone mode" of PostGIS only? I'd 
rather not like to recompile Postgres to include debugging stuff..

I'd appreciate any comments on how to proceed here - still hoping i've 
triggered a well-known bug...

thanks

Alex Mayrhofer