regression failures under windows 9.2 64-bit release

[robe]

I have to verify if same issue on my regular 32-bit chain but on mingw64 my 9.2 is showing this, but might be something amiss with my compilation or something since I recently built 9.2 release.

 loader/Point .............. ok 
 loader/PointM .............. ok 
 loader/PointZ .............. ok 
 loader/MultiPoint .............. ok 
 loader/MultiPointM .............. ok 
 loader/MultiPointZ .............. ok 
 loader/Arc .............. ok 
 loader/ArcM .............. ok 
 loader/ArcZ .............. ok 
 loader/Polygon .............. ok 
 loader/PolygonM .............. ok 
 loader/PolygonZ .............. ok 
 loader/TSTPolygon ......... ok 
 loader/TSIPolygon ......... ok 
 loader/TSTIPolygon ......... ok 
 loader/PointWithSchema ..... ok 
 loader/NoTransPoint ......... ok 
 loader/NotReallyMultiPoint ......... ok 
 loader/MultiToSinglePoint ......... ok 
 loader/ReprojectPts ........ ok 
 loader/ReprojectPtsGeog ........ ok 
 loader/Latin1 .... ok 
 binary .. ok 
 regress .. failed (diff expected obtained: /tmp/pgis_reg/test_24_diff)
 regress_index .. failed (diff expected obtained: /tmp/pgis_reg/test_25_diff)
 regress_index_nulls .. failed (diff expected obtained: /tmp/pgis_reg/test_26_diff)
 lwgeom_regress .. ok 
 regress_lrs .. ok 
 removepoint .. ok 
 setpoint .. ok 
 simplify .. ok 
 snaptogrid .. ok 
 summary .. ok 
 affine .. ok 
 empty .. ok 
 measures .. ok 
 legacy .. failed (diff expected obtained: /tmp/pgis_reg/test_37_diff)
 long_xact .. ok 
 ctors .. ok 
 sql-mm-serialize .. ok 
 sql-mm-circularstring .. ok 
 sql-mm-compoundcurve .. ok 
 sql-mm-curvepoly .. ok 
 sql-mm-general .. ok 
 sql-mm-multicurve .. ok 
 sql-mm-multisurface .. ok 
 polyhedralsurface .. ok 
 polygonize .. ok 
 postgis_type_name .. ok 
 geography .. ok 
 out_geometry .. ok 
 out_geography .. ok 
 in_gml .. ok 
 in_kml .. ok 
 iscollection .. ok 
 regress_ogc .. ok 
 regress_ogc_cover .. ok 
 regress_ogc_prep .. ok 
 regress_bdpoly .. ok 
 regress_proj .. ok 
 regress_management .. failed (diff expected obtained: /tmp/pgis_reg/test_61_diff)
 dump .. ok 
 dumppoints .. ok 
 wmsservers .. ok 
 wkt .. ok 
 wkb .. ok 
 tickets .. ok 
 typmod .. failed (diff expected obtained: /tmp/pgis_reg/test_68_diff)
 remove_repeated_points .. ok 
 split .. ok 
 relate .. ok 
 bestsrid .. ok 
 concave_hull .. ok 
 hausdorff .. ok 
 regress_buffer_params .. ok 
 offsetcurve .. failed (diff expected obtained: /tmp/pgis_reg/test_76_diff)
 relatematch .. ok 
 isvaliddetail .. ok 
 sharedpaths .. ok 
 snap .. ok 
 node .. ok 
 unaryunion .. ok 
 clean .. ok 
 relate_bnr .. ok 
 delaunaytriangles .. ok 
 in_geojson .. ok 
 uninstall .

Winnie 9.1 x64 bit is passing tests fine.

I've ruled out 9.2 as primary culprit since Debbie is passing with flying colors and she's running 9.2 release version too.

[robe]

I think I may have been missing iconv or something path settings. My 9.2 32-bit is going thru okay after I copied. Still need to check my 64-bit, but I'm guessing it is the same issue

[robe]

Okay something else wrong here. Winnie is doing fine on regressing 32-bit 9.2 but coughing with regressing the 9.2 64bit. idon't recall this issue with the rc1s.

Inspection of failures, are crashes of the backend. I'm having similar issue with building the 2.0.1 64-bit for 9.2 on my windows 7 64-bit. The annoying thing is if I run the statements manually that cause the crash, they don't crash.

[robe]

Okay seems to be some sort of memory leak or something.

An example that crashes the service (only my 64-bit the 32-bit seems fine)

If I run this statement 3-4 times on the 3rd to 4th time it crashes the service

select '165', ST_AsEWKT('POLYGON((0 0 0, 1 0 0, 1 1 0, 0 1 0, 0 0.1 1))');

I don't recall having this issue when testing Pg9.2rc1, so this may be a new development.

[robe]

Seems to be something to do with error handling as it only crashes for the error handler regress ones.

Why I don't have issue with 9.1x64bit EDB is a bit puzzling.

[robe]

I think this is a winnie only issue. I think my windows 7 mingw is fine for 9.2. So I'm going to push this since it's probably a config issue and not code.

[robe]

this hasn't happened in a while even on winnie.