Test 1.4 against Windows

[pramsey]

Build / regress against PgSQL 8.1, 8.2, 8.3, CVS.

[mcayland]

I've done a build against 8.3 on Windows and it generally works, although the autotools output seems a little strange and the old CR/LF issue with the regression tests has re-appeared with the SVN move.

ATB,

Mark.

[robe]

For what its worth I've been testing on 8.3 and 8.4 under windows, but I for the life of me can't get make check to work on windows so been just doing tests with real data and diffing against my 1.3 output.

Is make check doable under Windows?

[robe]

nevermind my wining. .Forgot to install CUnit so now I can run make check under windows. However running against the 8.4 enterprisedb beta2 and GEOS 3.1.1 and Postgis 1.4 after Paul's cleanup Back end crash with make check

regress_log just lists a BEGIN ..M

[mcayland]

Hmmm. Does that mean it doesn't even get as far as loading the functions into the regression database?!

[robe]

Yap it manages to create the postgis_reg db. pgpllang language exists too, though that may be defaulted in 8.4 so might not mean it created it and then it promptly kills the backend. There are so many variables that changed -- all 3 (geos, Paul doing his cleanup, and me running make check for first time) , so I'm going to isolate it unless you are seeing the same thing.

Anyrate seems to be the make check process rather than anything wrong with the install since I can run spatial queries fine. Maybe its a weird character or something that windows doesn't like. I also have not retested on my OpenSUSE boxes.

[pramsey]

Any progress here? Really want to push an RC...

[robe]

Mark -- any chance you can do a quick make check on yours. I don't have 8.3 loaded on here to verify if make check works on that and probably won't be able to get to that for a couple days.

Paul -- if you really want to push out an RC, I'm all for it. Don't let this get in your way. Like I said this is the first time I attempted a make check under MingW and I've been running a bunch of tests of 8.4 with GEOS 3.1.1 (ming compiled) and some on real data. As far as I can tell it performs fine. Perhaps if you call it a beta instead of an RC you'll feel better :).

[mcayland]

I think Regina has now managed this?

ATB,

Mark.

[robe]

Mark,

Still having issues with make check, but I think I got a bit further. Managed to get CUnit to work so I get this

CUnit - A Unit testing framework for C - Version 2.1-0  http://cunit.sourceforge.net/

Suite: PostGIS Computational Geometry Suite

Test: test_lw_segment_side() ... passed Test: test_lw_segment_intersects() ... passed Test: test_lwline_crossing_short_lines() ... passed Test: test_lwline_crossing_long_lines() ... passed Test: test_lwpoint_set_ordinate() ... passed Test: test_lwpoint_get_ordinate() ... passed Test: test_lwpoint_interpolate() ... passed Test: test_lwline_clip() ... passed Test: test_lwline_clip_big() ... passed Test: test_lwmline_clip() ... passed Test: test_geohash_point() ... passed Test: test_geohash_precision() ... passed Test: test_geohash() ... passed

Suite: PostGIS Measures Suite

Test: test_mindistance2d_recursive_tolerance() ... passed

--Run Summary: Type Total Ran Passed Failed

suites 2 2 n/a 0 tests 14 14 14 0 asserts 84 84 84 0

So I think its just runnning the regular PostgreSQL tests that are the problem now -- give stack trace. That could be becuase I'm trying to test my enterprisedb install against this. Anyrate the stack error gives this

MSYS-1.0.11 Build:2009-05-23 19:33 Exception: STATUS_ACCESS_VIOLATION at eip=71084D85 eax=636BF77F ebx=0023ED04 ecx=0023F119 edx=63696666 esi=0023F119 edi=0023ED04 ebp=0023ECA4 esp=0023EC9C program=C:\msys\1.0\bin\make.exe cs=001B ds=0023 es=0023 fs=003B gs=0000 ss=0023 Stack trace: Frame Function Args 0023ECA4 71084D85 (0023F119, 63696666, 7C810EB6, 71023DB2) 0023ECD4 71084C2E (0023ED04, 63696666, 0000003B, 0023EFEC) 0023F104 71081EE3 (63696666, 664F2F65, 65636966, 0040002F)

605735 [main] make 3000 handle_exceptions: Error while dumping state (probably corrupted stack).

Also still can't get a compile against 8.2.

[robe]

Mark, It fails for us on the loader, but I think there is something wrong with our libiconv2 install since it keeps complaining about not being able to find it. I tested on my messy isntall which doesn't complain about libiconv, and it gives similar errors too. Is this what you were talking about to confirm.

TMPDIR is /tmp/pgis_reg_308
PATH is "c:/POSTGRES84//bin":/c/postgres84/bin:.:/usr/local/bin:/mingw/bin:/bin:/c/WINDOWS/system32:/c/WINDOWS:/c/WINDOWS/System32/Wbem:/c/Program Files/GnuWin32/bin:/mi
n

 PostgreSQL 8.4.0 on i686-pc-mingw32, compiled by GCC gcc.exe (GCC) 3.4.5 (mingw-vista special r3), 32-bit
 Postgis 1.4.0rc2 - 2009-07-17 13:55:37
   GEOS: 3.1.1-CAPI-1.6.0
   PROJ: Rel. 4.6.1, 21 August 2008

Running tests

 loader/Point. failed (running shp2pgsql: /tmp/pgis_reg_308/loader.err)
 loader/PointM. failed (running shp2pgsql: /tmp/pgis_reg_308/loader.err)
 loader/PointZ. failed (running shp2pgsql: /tmp/pgis_reg_308/loader.err)
 loader/MultiPoint. failed (running shp2pgsql: /tmp/pgis_reg_308/loader.err)
 loader/MultiPointM. failed (running shp2pgsql: /tmp/pgis_reg_308/loader.err)
 loader/MultiPointZ. failed (running shp2pgsql: /tmp/pgis_reg_308/loader.err)
 loader/Arc. failed (running shp2pgsql: /tmp/pgis_reg_308/loader.err)
 loader/ArcM. failed (running shp2pgsql: /tmp/pgis_reg_308/loader.err)
 loader/ArcZ. failed (running shp2pgsql: /tmp/pgis_reg_308/loader.err)
 loader/Polygon. failed (running shp2pgsql: /tmp/pgis_reg_308/loader.err)
 loader/PolygonM. failed (running shp2pgsql: /tmp/pgis_reg_308/loader.err)
 loader/PolygonZ. failed (running shp2pgsql: /tmp/pgis_reg_308/loader.err)
 regress. failed (diff expected obtained: /tmp/pgis_reg_308/test_13_diff)
 regress_index. ok
 regress_index_nulls. ok
 lwgeom_regress. ok
 regress_lrs. ok
 removepoint. ok
 setpoint. ok
 simplify. ok
 snaptogrid. ok
 affine. ok
 wkt. failed (diff expected obtained: /tmp/pgis_reg_308/test_23_diff)
 measures. ok
 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
 geojson. ok
 gml. ok
 svg. ok
 kml. ok
 regress_ogc. failed (diff expected obtained: /tmp/pgis_reg_308/test_38_diff)
 regress_bdpoly. ok
 regress_proj. ok
 regress_ogc_cover. ok
 regress_ogc_prep. ok

Run tests: 42
Failed: 15
make[1]: Leaving directory `/c/thesrc/postgis-1.4.0rc2/regress'

[mcayland]

What do the loader diffs look like? On my Win32 virtual, it looks as if there is random corruption in the Z/M coordinates :(

Incidentally all of the other non-loader tests look fine for me, although I did have to run dos2unix on all of the input .sql and _expected output files.

Oh and one more thing: can you launch shp2pgsql/pgsql2shp directly from the command line? If you try and launch each with no parameters, you should be able to see the help text - if there is a .DLL missing then you'll get a suitable popup.

ATB,

Mark.

[robe]

On mine as far as I can tell -- there is an additional whitespace ^{M character added (which I can't see with a regular editor I can seein vim). Well actually I'm looking at my other install (my messed up one). I get more failures on this one, but they look like that extra white space tacked at the end. I'll attach for you to verify. So I assume this is just an artifact of MingW and not necessarily a true error.}

Below is output of this make check

make -C liblwgeom make[1]: Entering directory `/D/SVNProjects/Postgis/branches/1.4/liblwgeom' make[1]: Nothing to be done for `all'. make[1]: Leaving directory `/D/SVNProjects/Postgis/branches/1.4/liblwgeom' make -C postgis make[1]: Entering directory `/D/SVNProjects/Postgis/branches/1.4/postgis' make[1]: Nothing to be done for `all'. make[1]: Leaving directory `/D/SVNProjects/Postgis/branches/1.4/postgis' make -C liblwgeom/cunit check make[1]: Entering directory `/D/SVNProjects/Postgis/branches/1.4/liblwgeom/cunit'

CUnit - A Unit testing framework for C - Version 2.1-0  http://cunit.sourceforge.net/

Suite: PostGIS Computational Geometry Suite

Test: test_lw_segment_side() ... passed Test: test_lw_segment_intersects() ... passed Test: test_lwline_crossing_short_lines() ... passed Test: test_lwline_crossing_long_lines() ... passed Test: test_lwpoint_set_ordinate() ... passed Test: test_lwpoint_get_ordinate() ... passed Test: test_lwpoint_interpolate() ... passed Test: test_lwline_clip() ... passed Test: test_lwline_clip_big() ... passed Test: test_lwmline_clip() ... passed Test: test_geohash_point() ... passed Test: test_geohash_precision() ... passed Test: test_geohash() ... passed

Suite: PostGIS Measures Suite

Test: test_mindistance2d_recursive_tolerance() ... passed

--Run Summary: Type Total Ran Passed Failed

suites 2 2 n/a 0 tests 14 14 14 0 asserts 84 84 84 0

make[1]: Leaving directory `/D/SVNProjects/Postgis/branches/1.4/liblwgeom/cunit' make -C regress check make[1]: Entering directory `/D/SVNProjects/Postgis/branches/1.4/regress' make -C ../postgis make[2]: Entering directory `/D/SVNProjects/Postgis/branches/1.4/postgis' make[2]: Nothing to be done for `all'. make[2]: Leaving directory `/D/SVNProjects/Postgis/branches/1.4/postgis' cp ../postgis/postgis.sql . Creating spatial db postgis_reg TMPDIR is /tmp/pgis_reg_3848 PATH is /c/thesrc/pg84/REL-84~1.0/bin:.:/usr/local/bin:/mingw/bin:/bin:/c/program files/imagemagick-6.3.4-q16:/c/WINDOWS/system32:/c/WINDOWS:/c/WINDOWS/System32/Wbem:/c/Program Files/Microsoft? SQL Server/80/Tools/Binn/:/c/Program Files/Microsoft? SQL Server/90/DTS/Binn/:/c/Program Files/Microsoft? SQL Server/90/Tools/binn/:/c/Program Files/Microsoft? SQL Server/90/Tools/Binn/VSShell/Common7/IDE/:/c/Program Files/Microsoft? Visual Studio 8/Common7/IDE/PrivateAssemblies/:/c/Program Files/QuickTime?/QTSystem/:/c/Program Files/jEdit:/c/Program Files/Support? Tools/:/c/Program Files/pgAdmin III/1.8/:/c/Program Files/HTMLDOC:/c/Program Files/Smart? Projects/IsoBuster:/c/WINDOWS/system32/WindowsPowerShell/v1.0:/c/Program Files/Microsoft? SQL Server/100/Tools/Binn/:/c/Program Files/Microsoft? SQL Server/100/DTS/Binn/:/c/Program Files/Microsoft? SQL Server/100/Tools/Binn/VSShell/Common7/IDE/:/c/Program Files/doxygen/bin:/c/Program Files/WinMerge:/c/opt/gtk/bin:/c/MS4W/Apache/cgi-bin:/mingw/bin:/c/Program Files/Microsoft? Office/Office?

PostgreSQL 8.4beta1 on i686-pc-mingw32, compiled by GCC gcc.exe (GCC) 3.4.5 (mingw special), 32-bit Postgis 1.4.0SVN - 2009-07-17 13:24:11

GEOS: 3.1.1-CAPI-1.6.0 PROJ: Rel. 4.6.1, 21 August 2008

Running tests

loader/Point... failed (wkb insert: diff expected obtained: /tmp/pgis_reg_3848/test_1_diff) loader/PointM... failed (wkb insert: diff expected obtained: /tmp/pgis_reg_3848/test_2_diff) loader/PointZ... failed (wkb insert: diff expected obtained: /tmp/pgis_reg_3848/test_3_diff) loader/MultiPoint... failed (wkb insert: diff expected obtained: /tmp/pgis_reg_3848/test_4_diff) loader/MultiPointM... failed (wkb insert: diff expected obtained: /tmp/pgis_reg_3848/test_5_diff) loader/MultiPointZ... failed (wkb insert: diff expected obtained: /tmp/pgis_reg_3848/test_6_diff) loader/Arc... failed (wkb insert: diff expected obtained: /tmp/pgis_reg_3848/test_7_diff) loader/ArcM... failed (wkb insert: diff expected obtained: /tmp/pgis_reg_3848/test_8_diff) loader/ArcZ.......... ok loader/Polygon... failed (wkb insert: diff expected obtained: /tmp/pgis_reg_3848/test_10_diff) loader/PolygonM... failed (wkb insert: diff expected obtained: /tmp/pgis_reg_3848/test_11_diff) loader/PolygonZ... failed (wkb insert: diff expected obtained: /tmp/pgis_reg_3848/test_12_diff) regress. failed (diff expected obtained: /tmp/pgis_reg_3848/test_13_diff) regress_index. failed (diff expected obtained: /tmp/pgis_reg_3848/test_14_diff) regress_index_nulls. ok lwgeom_regress. failed (diff expected obtained: /tmp/pgis_reg_3848/test_16_diff) regress_lrs. failed (diff expected obtained: /tmp/pgis_reg_3848/test_17_diff) removepoint. failed (diff expected obtained: /tmp/pgis_reg_3848/test_18_diff) setpoint. failed (diff expected obtained: /tmp/pgis_reg_3848/test_19_diff) simplify. failed (diff expected obtained: /tmp/pgis_reg_3848/test_20_diff) snaptogrid. failed (diff expected obtained: /tmp/pgis_reg_3848/test_21_diff) affine. failed (diff expected obtained: /tmp/pgis_reg_3848/test_22_diff) wkt. failed (diff expected obtained: /tmp/pgis_reg_3848/test_23_diff) measures. failed (diff expected obtained: /tmp/pgis_reg_3848/test_24_diff) long_xact. failed (diff expected obtained: /tmp/pgis_reg_3848/test_25_diff) ctors. failed (diff expected obtained: /tmp/pgis_reg_3848/test_26_diff) 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 geojson. ok gml. ok svg. ok kml. ok regress_ogc. failed (diff expected obtained: /tmp/pgis_reg_3848/test_38_diff) regress_bdpoly. failed (diff expected obtained: /tmp/pgis_reg_3848/test_39_diff) regress_proj. failed (diff expected obtained: /tmp/pgis_reg_3848/test_40_diff) regress_ogc_cover. ok regress_ogc_prep. ok

Run tests: 42 Failed: 27 make[1]: Leaving directory `/D/SVNProjects/Postgis/branches/1.4/regress'

[robe]

Sorry forgot to put pretty code bracket in. {{{make -C liblwgeom make[1]: Entering directory `/D/SVNProjects/Postgis/branches/1.4/liblwgeom' make[1]: Nothing to be done for `all'. make[1]: Leaving directory `/D/SVNProjects/Postgis/branches/1.4/liblwgeom' make -C postgis make[1]: Entering directory `/D/SVNProjects/Postgis/branches/1.4/postgis' make[1]: Nothing to be done for `all'. make[1]: Leaving directory `/D/SVNProjects/Postgis/branches/1.4/postgis' make -C liblwgeom/cunit check make[1]: Entering directory `/D/SVNProjects/Postgis/branches/1.4/liblwgeom/cunit'

CUnit - A Unit testing framework for C - Version 2.1-0  http://cunit.sourceforge.net/

Suite: PostGIS Computational Geometry Suite

Test: test_lw_segment_side() ... passed Test: test_lw_segment_intersects() ... passed Test: test_lwline_crossing_short_lines() ... passed Test: test_lwline_crossing_long_lines() ... passed Test: test_lwpoint_set_ordinate() ... passed Test: test_lwpoint_get_ordinate() ... passed Test: test_lwpoint_interpolate() ... passed Test: test_lwline_clip() ... passed Test: test_lwline_clip_big() ... passed Test: test_lwmline_clip() ... passed Test: test_geohash_point() ... passed Test: test_geohash_precision() ... passed Test: test_geohash() ... passed

Suite: PostGIS Measures Suite

Test: test_mindistance2d_recursive_tolerance() ... passed

--Run Summary: Type Total Ran Passed Failed

suites 2 2 n/a 0 tests 14 14 14 0 asserts 84 84 84 0

make[1]: Leaving directory `/D/SVNProjects/Postgis/branches/1.4/liblwgeom/cunit' make -C regress check make[1]: Entering directory `/D/SVNProjects/Postgis/branches/1.4/regress' make -C ../postgis make[2]: Entering directory `/D/SVNProjects/Postgis/branches/1.4/postgis' make[2]: Nothing to be done for `all'. make[2]: Leaving directory `/D/SVNProjects/Postgis/branches/1.4/postgis' cp ../postgis/postgis.sql . Creating spatial db postgis_reg TMPDIR is /tmp/pgis_reg_3848 PATH is /c/thesrc/pg84/REL-84~1.0/bin:.:/usr/local/bin:/mingw/bin:/bin:/c/program files/imagemagick-6.3.4-q16:/c/WINDOWS/system32:/c/WINDOWS:/c/WINDOWS/System32/Wbem:/c/Program Files/Microsoft? SQL Server/80/Tools/Binn/:/c/Program Files/Microsoft? SQL Server/90/DTS/Binn/:/c/Program Files/Microsoft? SQL Server/90/Tools/binn/:/c/Program Files/Microsoft? SQL Server/90/Tools/Binn/VSShell/Common7/IDE/:/c/Program Files/Microsoft? Visual Studio 8/Common7/IDE/PrivateAssemblies/:/c/Program Files/QuickTime?/QTSystem/:/c/Program Files/jEdit:/c/Program Files/Support? Tools/:/c/Program Files/pgAdmin III/1.8/:/c/Program Files/HTMLDOC:/c/Program Files/Smart? Projects/IsoBuster:/c/WINDOWS/system32/WindowsPowerShell/v1.0:/c/Program Files/Microsoft? SQL Server/100/Tools/Binn/:/c/Program Files/Microsoft? SQL Server/100/DTS/Binn/:/c/Program Files/Microsoft? SQL Server/100/Tools/Binn/VSShell/Common7/IDE/:/c/Program Files/doxygen/bin:/c/Program Files/WinMerge:/c/opt/gtk/bin:/c/MS4W/Apache/cgi-bin:/mingw/bin:/c/Program Files/Microsoft? Office/Office?

PostgreSQL 8.4beta1 on i686-pc-mingw32, compiled by GCC gcc.exe (GCC) 3.4.5 (mingw special), 32-bit Postgis 1.4.0SVN - 2009-07-17 13:24:11

GEOS: 3.1.1-CAPI-1.6.0 PROJ: Rel. 4.6.1, 21 August 2008

Running tests

loader/Point... failed (wkb insert: diff expected obtained: /tmp/pgis_reg_3848/test_1_diff) loader/PointM... failed (wkb insert: diff expected obtained: /tmp/pgis_reg_3848/test_2_diff) loader/PointZ... failed (wkb insert: diff expected obtained: /tmp/pgis_reg_3848/test_3_diff) loader/MultiPoint... failed (wkb insert: diff expected obtained: /tmp/pgis_reg_3848/test_4_diff) loader/MultiPointM... failed (wkb insert: diff expected obtained: /tmp/pgis_reg_3848/test_5_diff) loader/MultiPointZ... failed (wkb insert: diff expected obtained: /tmp/pgis_reg_3848/test_6_diff) loader/Arc... failed (wkb insert: diff expected obtained: /tmp/pgis_reg_3848/test_7_diff) loader/ArcM... failed (wkb insert: diff expected obtained: /tmp/pgis_reg_3848/test_8_diff) loader/ArcZ.......... ok loader/Polygon... failed (wkb insert: diff expected obtained: /tmp/pgis_reg_3848/test_10_diff) loader/PolygonM... failed (wkb insert: diff expected obtained: /tmp/pgis_reg_3848/test_11_diff) loader/PolygonZ... failed (wkb insert: diff expected obtained: /tmp/pgis_reg_3848/test_12_diff) regress. failed (diff expected obtained: /tmp/pgis_reg_3848/test_13_diff) regress_index. failed (diff expected obtained: /tmp/pgis_reg_3848/test_14_diff) regress_index_nulls. ok lwgeom_regress. failed (diff expected obtained: /tmp/pgis_reg_3848/test_16_diff) regress_lrs. failed (diff expected obtained: /tmp/pgis_reg_3848/test_17_diff) removepoint. failed (diff expected obtained: /tmp/pgis_reg_3848/test_18_diff) setpoint. failed (diff expected obtained: /tmp/pgis_reg_3848/test_19_diff) simplify. failed (diff expected obtained: /tmp/pgis_reg_3848/test_20_diff) snaptogrid. failed (diff expected obtained: /tmp/pgis_reg_3848/test_21_diff) affine. failed (diff expected obtained: /tmp/pgis_reg_3848/test_22_diff) wkt. failed (diff expected obtained: /tmp/pgis_reg_3848/test_23_diff) measures. failed (diff expected obtained: /tmp/pgis_reg_3848/test_24_diff) long_xact. failed (diff expected obtained: /tmp/pgis_reg_3848/test_25_diff) ctors. failed (diff expected obtained: /tmp/pgis_reg_3848/test_26_diff) 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 geojson. ok gml. ok svg. ok kml. ok regress_ogc. failed (diff expected obtained: /tmp/pgis_reg_3848/test_38_diff) regress_bdpoly. failed (diff expected obtained: /tmp/pgis_reg_3848/test_39_diff) regress_proj. failed (diff expected obtained: /tmp/pgis_reg_3848/test_40_diff) regress_ogc_cover. ok regress_ogc_prep. ok

Run tests: 42 Failed: 27 make[1]: Leaving directory `/D/SVNProjects/Postgis/branches/1.4/regress' }}}

[robe]

make check against 8.4beta2 with log

[robe]

Sorry didn't see your last note about having to run dos2unix. For some reason I don't have dos2unix. I thought I did before. Anyrate I'll try again later after I have run that.

[robe]

Mark we tested on our new install again. There is nothing wrong with our libiconv. The problem seems to be that the loader files can't see it. If we copy the libpq.dll and libiconv-2.dll that we compiled into the loader folder. The error goes away.

This is what our test looks like now against PostgreSQL 8.4.0

 PostgreSQL 8.4.0 on i686-pc-mingw32, compiled by GCC gcc.exe (GCC) 3.4.5 (mingw-vista special r3), 32-bit
 Postgis 1.4.0rc2 - 2009-07-17 13:55:37
   GEOS: 3.1.1-CAPI-1.6.0
   PROJ: Rel. 4.6.1, 21 August 2008

Running tests

 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
 regress. failed (diff expected obtained: /tmp/pgis_reg_692/test_13_diff)
 regress_index. ok
 regress_index_nulls. ok
 lwgeom_regress. ok
 regress_lrs. ok
 removepoint. ok
 setpoint. ok
 simplify. ok
 snaptogrid. ok
 affine. ok
 wkt. failed (diff expected obtained: /tmp/pgis_reg_692/test_23_diff)
 measures. ok
 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
 geojson. ok
 gml. ok
 svg. ok
 kml. ok
 regress_ogc. failed (diff expected obtained: /tmp/pgis_reg_692/test_38_diff)
 regress_bdpoly. ok
 regress_proj. ok
 regress_ogc_cover. ok
 regress_ogc_prep. ok

Run tests: 42
Failed: 3

I'll attach the diff files in a bit.

[robe]

make check regress log against 8.4.0 mingw

[mcayland]

Right, so that is an 8.4-ism. I think I've fixed this in 1.4 branch/trunk, and I've also applied the various regress/Makefile.in patches to help get the regression tests to run. Please can you test and see what happens?

ATB,

Mark.

[robe]

You mean about the M thing. that was probably an 8.4betism. If you look we were testing on both 8.4 (one was an 8.4beta and the other was 8.4 release).

Anyrate we were also testing one against an RC patched and one against SVN. So we'll try the SVN if you have all the patches in place and see what we get.

[robe]

Mark,

Sorry for being so slow on this. This week as you can imagine is really bad. I was able to test the PostgreSQL 8.4.0 (I upgraded from beta) on non-clean box (that one runs GCC 4.4)

After running the dos2unix -- all tests passed with flying colors.

Still have yet to test 8.3 and 8.4 8.3 running into stupid issues recompiling it (not even getting to the PostGIS part). Was trying to use my originally compiled 8.3, but was getting errors such as

"more" not defined when trying to do PostGIS make check. Which is why I decided to try to recompile PostgreSQL 8.3. this is all on the mixed up install. Have yet to retry on clean mingW install.

[robe]

Correction. I mean test 8.3 and 8.2

[mcayland]

Yeah, I get that irritating "more" error when trying "make check" on my running PostgreSQL 8.3 install :( I think trying on a MingW-compiled PostgreSQL 8.3 may yield slightly more success...

ATB,

Mark.

[robe]

I was using a MingW compiled 8.3. I was actually able to eventually compile it again under MingW after deleting my install director. It seems that somehow our make check is messing up the path or something in MingW is messing up the path. I couldn't find more anywhere, but noticed that if I launched

my 8.3 psql from within MinGW and did a

help, I got the same more error. So I think its the psql that is somehow triggering it. why it doesn't trigger it for my MingW installed PostgreSQL 8.4 I have no clue.

But running my 8.3 psql outside of MingW seems to work fine.

Note that I don't believe MingW includes more. All articles I read seem to suggest using less instead of more (perhaps maybe 8.4 is using less?). But copying my more.com from windows (and even renaming it to more.exe and putting in my msys/1.0/bin doesn't seem to resolve the situation).

[mcayland]

Hmmm okay. I'd try raising this on the pgsql mailing lists just in case something got patched for 8.4. Incidentally if it's a PATH issue, my top 2 candidates for locations to copy more.com would be the PostgreSQL bin directory, followed by the regression test directory.

HTH,

Mark.

[robe]

I think I tried bin of postgresql already. But haven't tried regress

[nicklas]

I did a successful make check with PostgreSQL 8.2, postgis 1.4.0rc2 I couldn't find any log with the testresults so I copied from MinGW to a textdocument. I also attached the folder created in temp directory.

platform: Windows XP SP3

/Nicklas

[nicklas]

Now I also successed with postgreSQL8.3 Attached same documents as 8.2 above.

/Nicklas

[robe]

I was able to make check successfully against 8.3 after copying more.com into the regress folder and renaming it to more.exe. See below:

I verified that Nick's tests look authentic for both 8.2 and 8.3.

Nick did you have to copy more.exe into regress folder? If not, can you do us a favor and test against the 1.4 SVN. I noticed your tests are probably against a patched rc2 and I believe Mark had made additional changes only in SVN, so I'm a bit concerned he introduced something that is causing the more issue.

It is also quite possible that its something strange with the newer GCC and Mark may be running the newer GCC as well.

 PostgreSQL 8.3.7 on i686-pc-mingw32, compiled by GCC gcc.exe (GCC) 4.4.0
 Postgis 1.4.0SVN - 2009-07-21 13:44:14
   GEOS: 3.1.1-CAPI-1.6.0
   PROJ: Rel. 4.6.1, 21 August 2008

Running tests

 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
 regress. ok
 regress_index. ok
 regress_index_nulls. ok
 lwgeom_regress. ok
 regress_lrs. ok
 removepoint. ok
 setpoint. ok
 simplify. ok
 snaptogrid. ok
 affine. ok
 wkt. ok
 measures. ok
 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
 geojson. ok
 gml. ok
 svg. ok
 kml. ok
 regress_ogc. ok
 regress_bdpoly. ok
 regress_proj. ok
 regress_ogc_cover. ok
 regress_ogc_prep. ok

Run tests: 42
Failed: 0

[nicklas]

No, I didn't copy more.exe. but I had to put libpq.dll in the bin-foder. I used this version  http://www.postgis.org/download/postgis-1.4.0rc2.tar.gz

I also noted that I seems to have an old version ov libiconv. I think I have version 1.9.2.1 I say "think" because I thought I could do libiconv -v but had no luck. But the files in minGW says 1.9.2.1 The rest I think is up to date.

/Nicklas

[nicklas]

and I forgot to mention: I edited the path in makefile, took away the ":" and wrote the whole name for truncated fildernames.

/Nicklas

[nicklas]

I'm talking about the makefile in regression-folder

/Nicklas

[robe]

We have successfully make checked under windows for 8.2,8.3,8.4

Can we have a 1.4 release

+1