#2671 closed defect (fixed)
twkb regression failure
Reported by: | robe | Owned by: | nicklas |
---|---|---|---|
Priority: | blocker | Milestone: | PostGIS 2.2.0 |
Component: | postgis | Version: | master |
Keywords: | Cc: |
Description
It's good we put those tests in. Our very first twkb failure.
This only seems to fail on winnie on the 32-bit test. 64-bit tests pass as far as I can tell - this is error from:
GEOS config: /projects/geos/rel-3.4.2w32/bin/geos-config GEOS version: 3.4.2 GDAL config: /projects/gdal/rel-1.10.0w32/bin/gdal-config GDAL version: 1.10.0 PostgreSQL config: /projects/postgresql/rel/pg9.2w32/bin/pg_config PostgreSQL version: PostgreSQL 9.2.4 PROJ4 version: 48 Libxml2 config: /projects/libxml/rel-libxml2-2.7.8w32/bin/xml2-config Libxml2 version: 2.7.8 JSON-C support: yes PostGIS debug level: 0 Perl: /bin/perl
twkb .. failed (diff expected obtained: /projects/postgis/tmp/2.2.0dev_pg9.2_geos3.4.2_gdal1.10.0w32/test_81_diff) ----------------------------------------------------------------------------- --- twkb_expected 2014-03-09 10:29:06 -0400 +++ /projects/postgis/tmp/2.2.0dev_pg9.2_geos3.4.2_gdal1.10.0w32/test_81_out 2014-03-09 18:30:01 -0400 @@ -13,4 +13,4 @@ MULTIPOLYGON(((1 1, 1 20, 20 20, 20 1, 1 1),(3 3,3 4, 4 4,4 3,3 3)),((-1 1, -1 20, -20 20, -20 1, -1 1),(-3 3,-3 4, -4 4,-4 3,-3 3)))|0346020202050202002626000025250005040400020200000101000205070300262500002526000503040002010000010200 GEOMETRYCOLLECTION(MULTIPOINT(1 1,2 2),POINT(78 -78),POLYGON((1 1,1 2,2 2,2 1,1 1)))|0247034402020202024198019f0143010599019e010002020000010100 MULTIPOINT(1 1,2 2)|035502060202040202 -GEOMETRYCOLLECTION(MULTIPOINT(1 1,2 2),POINT(78 -78),POLYGON((1 1,1 2,2 2,2 1,1 1)))|0387035501c8439c019b015701bd08010599019e01000202000001010058010602410000410202 +ERROR: Output WKB is not the same size as the allocated buffer.
Change History (6)
comment:1 by , 11 years ago
comment:2 by , 11 years ago
twkb .. failed (diff expected obtained: /projects/postgis/tmp/2.2.0dev_pg9.2_geos3.4.2_gdal1.10.0w32/test_81_diff) ----------------------------------------------------------------------------- --- twkb_expected 2014-03-10 17:15:42 -0400 +++ /projects/postgis/tmp/2.2.0dev_pg9.2_geos3.4.2_gdal1.10.0w32/test_81_out 2014-03-10 23:03:41 -0400 @@ -14,4 +14,4 @@ GEOMETRYCOLLECTION(POINT(1 1),LINESTRING(2 2,3 3))|024702410202420202020202 GEOMETRYCOLLECTION(MULTIPOINT(1 1,2 2),POINT(78 -78),POLYGON((1 1,1 2,2 2,2 1,1 1)))|0247034402020202024198019f0143010599019e010002020000010100 MULTIPOINT(1 1,2 2)|035502060202040202 -GEOMETRYCOLLECTION(MULTIPOINT(1 1,2 2),POINT(78 -78),POLYGON((1 1,1 2,2 2,2 1,1 1)))|0387035501c8439c019b015701bd08010599019e01000202000001010058010602410000410202 +ERROR: Output TWKB is not the same size as the allocated buffer (precalculated size:42, allocated size:31) -----------------------------------------------------------------------------
comment:3 by , 11 years ago
This is about 32 bit I think. I get the same behavior on 32 bit Linux.
There is also many queries that crashes the server on 32-bit when using ST_AsTWKBagg.
I think it is about that I have done something wrong with the MemoryContext thing.
But it gets easier to investigate when I don't have to wait for compilation in windows
comment:4 by , 11 years ago
Hopefully fixed with r12329.
I will wait for debbie and winnie to run before closing it.
comment:5 by , 11 years ago
Resolution: | → fixed |
---|---|
Status: | new → closed |
I think it looks like it is fixed. I cannot get make check work on windows, but the query giving the error seems to behave better now.
I close it,
comment:6 by , 11 years ago
Yap looks good. ran garden tests as well 9.2 32-bit and didn't crash the tests this time around.
I have added the sizes to the error messages that you get about different sizes. I also added a simpler geometry collection to the test (easier to calculate )
Could you please post the error you get now? It should contain both the calculated twkb-size and the allocated size.
I will have to get a windows environment to test on, but I don't have my job computer at home. I will get it tomorrow.
Is any of the winnie jobs on 32 bit? Winnie seems quite unhappy now, but she is still working and I don't success to get the complete console output. I think I will wait a while till she calms down.